Home | Back
commit 4f8294aa9382fbd15346b3c1991069f3a1d7d2a0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 11 12:15:17 2010 +0300

    Preparing for 4.8.1

commit 6ff9124a9fae440c53dd92d092f475882d7851ab
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 11 08:17:12 2010 +0300

    If there are no hardlinks, dont bother with s-bit and caps removal
    (cherry picked from commit 89be57ad9239c9ada0cba94a5003876b456d46bf)

commit 5c75ee0f7a99a69af17c9be23b1cd64f77a4380c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 3 16:22:09 2010 +0300

    Fix "empty reply from server" curl-syndrome with url retrieval (RhBug:598988)
    - We were adding a trailing newline to urlhelper command line passed
      to execvp(), with the newline ending up in the URL passed to curl.
      Many servers dont seem to mind the extra newline, but some do. Oh well...
    (cherry picked from commit 15c0c4e7a1291915a4ea4115e94d2b747d5091ac)

commit e61d6b1cea27c1cd9aa3143032d96d2e6cdc9830
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 3 11:04:12 2010 +0300

    Remove POSIX file capabilities from executables on erase/rename too
    - Just like suid/sgid bits, hardlinks to executables with capabilities
      set can "leak" permissions to old, potentially vulnerable versions
      of binaries. Related to RhBug:598775.
    (cherry picked from commit 4d172a194addc49851e558ea390d3045894e3230)

commit ff3e5e5744e0c28e46dc28413738e20b5f807cce
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 2 09:15:48 2010 +0300

    Strip s-bits on upgrade too (RhBug:598775)
    - Previously only actual package removal would clear s-bits from
      s[ug]id files. This needs to be done on upgrade too to prevent
      leaving vulnerable hardlinks of s[ug]id binaries around.
    - backported from commit ca2d6b2b484f1501eafdde02e1688409340d2383

commit 13c2f430931aa30d735e635e32bab8dfd28dee06
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 1 11:58:42 2010 +0300

    Permit DOS line-endings in PGP armors (RhBug:532992)
    - RFC-4880 doesn't requires unix-style line-endings, we shouldn't either.
      This is probably still oversly strict as RFC-4880 appears to permit
      any whitespace to follow armor headers but ... shrug.
    (cherry picked from commit cadab3afa55388f3082a3468d582896caabfd55c)

commit ffca3787a6f334cfbc6dd478337bf8c4e9c48a49
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 19 15:33:52 2010 +0300

    Use pgpValString() for :pgpsig header formatting (RhBug:587755)
    - instead of copy-slopping (and forgetting to do so whenever something
      changes) just use the values we know from pgpVals..
    (cherry picked from commit e63293b0ea1c6a3a4be6aca328ee9fae4c3bfcdb)

commit ae66316c2cf9a49a804369fc1629eda426fa0ecd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 19 15:32:54 2010 +0300

    Add public pgpValString() function + enum for the various types
    - The equivalent used to be possible in 4.4.x era as the value tables
      were all exported, but this way we need to export far less
      and the implementation details stay internal.
    (cherry picked from commit 2be413810cd78b0a503699ceca3f8e1505f1b425)

commit 6b38fba9bc6ea85e36bb2140cce1a895cfa28537
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 19 10:12:43 2010 +0300

    Handle non-existent dependency sets in python (RhBug:593553)
    - rpmdsNew() returns NULL if the requested dependency type doesn't
      exist in the header. The C-side API can handle NULL to all rpmds
      "methods" and this is how librpm deals with non-existent sets rather
      than waste memory on for empty ds structures. However the python side
      wasn't expecting NULL for legal requests (but not setting error either)
      and thus blowing up with SystemError exception.
    - Raise TypeError on illegal arguments to rpm.ds constructor, and present
      non-existent dependency sets as empty rpm.ds objects to python. This
      lets python callers use iteration over ds items regardless of whether
      the dependency actually exists or not. The alternative of returning
      None (or raising exceptions) would break existing code for no
      particularly good reason.
    (cherry picked from commit 0e0e332b466a9784620c483faa374067381e96ce)

commit 0757c1cf5e3e76399e7bf05bc1efcb21c676406c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 18 10:39:22 2010 +0300

    Fix return from chroot() on verify (RhBug:590588)
    - a couple of important steps in chroot() in and out sequence missing,
      causing "No such file or directory" whining on return from chroot()
      unless cwd happened to be /
    - backported from commit 2b7884ce2914c48514023dbe61dc7a126964f438

commit 3f2573bb68156d55522c05ade5ea65381b418416
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 6 12:28:23 2010 +0300

    Fix minor header iterator leak in rpmReSign()
    - fix dumb double-initialization error introduced somewhere between
      4.4 and 4.6, also while-loop is more natural here than for-loop
    (cherry picked from commit 834c8383476fe54c65fe07b69c01cd3d35f5a488)

commit 7cf5cde7737eab947dba86568279421718169bfd
Author: Pavol Rusnak <stick@gk2.sk>
Date:   Wed Mar 17 16:02:46 2010 +0100

    Don't process desktop files without Type=Application and Exec= line
    (cherry picked from commit 3b96dfb29ab75a4f7f37713a93ec39460c158ef0)

commit da7b8f42229b48d906b678d5757fc7000278413e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 11 09:22:38 2010 +0300

    Make rpm version available to python through __version__, add __version_info__
    - backported from commits ed74161b7e5b90ba886926ced14ba3aca018a4f4 and
      21962ab6142cd1e54f1087a545244de506d9abf1

commit 8b5592d00514902956842fc441d1a6a86e1a414b
Author: Giulio Eulisse <Giulio.Eulisse@cern.ch>
Date:   Tue Apr 20 11:42:10 2010 +0200

    Fixes missing __fxstat64 symbol on mac.
    (cherry picked from commit d131c3393a2a43d87047bffdd73cf9ae3c8146c2)

commit 9a824d57c3679b241cb362f98119d5ecfcf2d0d2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 14 12:24:37 2010 +0300

    Add a basic reinstall test to test-suite
    - this currently fails due to commit e16695e932e294ec600e673d38c41bde20931204,
      on reinstall element header vs db header mismatch, adding this as
      a reminder to fix it...
    (cherry picked from commit a0ed26b894e301d479c8092c9c7fa16f2d6c3c2c)

commit 650c924cca32bfde680119b3ac2ee0d2d5b95bd4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 13 08:35:45 2010 +0300

    Use correct tag for NOPATCH
    - dumb copy-paste error from commit a317eaadc3f7b9b4ec373719e7d60fb8b98e2c21
    (cherry picked from commit a761daa2f8a0edb66733444f49f032f3a13035b1)

commit 18ca8246757ebdd6046dea3cb866adaf7cbd692f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 13 08:32:49 2010 +0300

    RPMTAG_NOSOURCE and NOPATCH aren't internal, they end up in nosrc headers
    (cherry picked from commit 538ddff69cacbd0db346cac3c10171424ae9b229)

commit f9c7db311302d3c03d524b133a372b8a05410d64
Author: Michael Schroeder <mls@suse.de>
Date:   Mon Apr 12 12:12:46 2010 +0200

    Gnu tar sometimes prints a "tar: Record size = 16" message, breaking
    rpmbuild -ta.
    (cherry picked from commit b2efc89d7db4f3a31dfd7f91028c5a9f9643334f)

commit 5526b5555b9110dbf69d68f31c3de0d8330941a3
Author: Michael Schroeder <mls@suse.de>
Date:   Mon Apr 12 12:11:56 2010 +0200

    Adapt the Makefile as rpmdb_svc can no longer be built.
    (cherry picked from commit dfa98e7a064d9ea28b416b52f2673202799bcb9f)

commit 5e1bc8c1d7b49058c9665ea52febd4cd4c14a12a
Author: Michael Schroeder <mls@suse.de>
Date:   Mon Apr 12 12:11:05 2010 +0200

    New gcc versions comlpain about undefined behaviour...
    (cherry picked from commit 18c3e1a37c482811b7b3af1415ef67aac527b869)

commit 5ae3e055ace91915cebf3ee1187a09aa39dba56a
Author: Michael Schroeder <mls@suse.de>
Date:   Mon Apr 12 12:10:20 2010 +0200

    readLineFromOFI may modify the fileStack, thus we have to re-set
    ofi after calling it
    (cherry picked from commit daec6ebf58331776279f389738af4788b865033d)

commit 8b1600432b8162d3e7a98277e000991274e602b4
Author: Michael Schroeder <mls@suse.de>
Date:   Mon Apr 12 12:09:04 2010 +0200

    Do not load keyring if signature checking is disabled.
    (cherry picked from commit cad147070e5513312d851f44998012e8f0cdf1e3)

commit a56897e571aea53b8cf78c7b180e58e127a98587
Author: Michael Schroeder <mls@suse.de>
Date:   Mon Apr 12 12:07:01 2010 +0200

    Seems like a change was made to make %attr(-) go back to the
    defattr setting. Unfortunatelly this broke %defattr(-).
    (cherry picked from commit 64e7f2aeb4ce01fccacde021fb22c85083284efb)

commit a650764cee83f53a292305b027a17d29573679ad
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 1 10:03:21 2010 +0300

    On header iteration format queries, dump the contents as-is
    - This makes eg --xml query of packages give out the true contents
      of the package, raw i18n strings and all, which is what you probably
      want when examining the contents on this level. Regular queries still
      run with all bells 'n whistless enabled.
    (cherry picked from commit c397a09be3efd087afeb25a5b01715eb8c889cdb)

commit a4d9f84f954ad24dbec8ad0682fbd565c96912c4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 1 09:15:06 2010 +0300

    Disable header i18n lookup in headerCopyTags() (RhBug:578299)
    - without HEADERGET_RAW here, only one of the translations in header
      (whatever happens to be "current" locale) gets copied
    - backported from commit 3ac99bd8c21436d8c70d462b64bf63453415a689

commit 95f75846448cc2dbc0c0a3f03508c3ec15e5fecc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 29 07:20:33 2010 +0300

    Report all enabled dependency bits in deptype format extension
    - dont make assumptions about which bits can be enabled simultaneously,
      just dump 'em all
    (cherry picked from commit 750b534942f0191f2d14d2974541ecaaa13820a5)

commit 01518483497bf759cf356a3451d848ce058bd88d
Author: Ville Skyttä <ville.skytta@iki.fi>
Date:   Fri Mar 19 20:09:15 2010 +0200

    Document deprecation of mi.count() and ds.Count().
    (cherry picked from commit e50e3d14e5149f5c7b30c8a0208e7e8bd80ee24a)

commit 2ded15368cfcce4206ddaa5ebd85ddfff2ac533b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 24 10:24:55 2010 +0200

    Add RPMBUILD_ISFOO constants to python rpmb module (ticket #123)
    (cherry picked from commit ee98a373cf79d3248122ae147d98ac461d59c9b7)

commit 9a93c9364d15cb569c6c9492be9ed992d28a3fc3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 24 09:53:25 2010 +0200

    Add __bool__() / __nonzero__() method to python rpmmi objects (ticket #153)
    - Objects supporting __len__() use (len > 0) for boolean representation,
      which normally makes sense but as the match iterator count is often
      zero despite the iterator actually existing and returning something,
      and breaks existing code (rpmlint at least)
    - Adding a __bool__() (known as __nonzero__() in Python < 3) method
      returning true for non-NULL iterator fixes this and gives more
      meaningful answers than pre 4.8.0 which simply always returned True
    (cherry picked from commit 40f788a7bf3741f9c613ff302d4e1b0ceec2658c)

commit 783dd6413aaf9a2fe3a951028c33884fdd16aec1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Mar 19 10:49:17 2010 +0200

    Verify the various version comparison results in testsuite
    (cherry picked from commit 09423eb47039a6f830592cc326d40bcd71650530)

commit 20d2fb8c91fcfe8572edf5e3b0e72a14fb4f9ec3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Mar 18 12:34:44 2010 +0200

    Add some basic dependency check tests to test-suite
    (cherry picked from commit d2d7ca7d43be0491f2f43d4867d4adbefbf922de)

commit 4d93fc4b3259c276524ce22ace9bc3385db6eb81
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Mar 18 12:23:38 2010 +0200

    Differentiate between conflicts in db versus install set
    - this appears to have been broken since its introduction in 2002,
      commit d89ce5854d39a9e896759fb96e6e303c67573bb6...
    - backported from commit a609aa7dbcf4738f3b47a63d3062dcadf58ef2fa

commit 0a08c8330e59bd3c0f65b7add021f303a3004f72
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 15 12:00:55 2010 +0200

    Avoid eating empty lines in spec %prep section (RhBug:573339)
    - In spec %prep context empty lines don't usually matter but they can
      be significant in eg here-documents.
    - Fixes regression from commit 94ff22b129aeb31c38848231e40f87aa4a5613a1
    (cherry picked from commit 12802c36c9a3b7260d9f788afc826b1cc5ee05e2)

commit 00ee15c7e33fec29df7abbb21f2e0f1d306e85d0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 15 11:54:55 2010 +0200

    Add an enhanced argvSplitString() function for splitting strings to argv's
    - Returns the newly created argv instead of useless "this always returns 0"
    - By default make a "real" split, including empty strings
    - Flags argument allows controlling behavior, for now only flag is to
      preserve argvSplit() behavior but leaves room for future enhancements
      such as quoted splitting etc
    (cherry picked from commit 35052b96232810cbf0d91a4f1d1d3ff25a142fd0)

commit 09843ba5c18fd3d3a4e7dcf1b71f3321b05c295e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 15 09:45:49 2010 +0200

    Tell python about RPMTRANS_FLAG_NOCONTEXTS
    (cherry picked from commit 8c5332984e32d27d28f9a440947b070af0d14c45)

commit c765a276124a2e209b475de0b753f233336b3f62
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Feb 17 12:58:40 2010 +0200

    Plug a silly memleak in processPackageFiles()
    (cherry picked from commit f3ca4fb48171b4bd0549fe67f302517135c465ff)

commit 51f8e2e71b0845fd6fa211fb05aa9191162a7f0a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Feb 4 15:55:16 2010 +0200

    Handle erased packages wrt disk-space problem reporting (RhBug:561160)
    - erased packages in the transaction cause space needs to shrink, which
      commit 2222e81e1cf74bbf78acb9969592ab06e2289929 failed to take into
      account, causing false disk space errors from unaffected packages
    - adjust obneeded and oineeded when the requirements shrink
    - new disk space problems can only occur when the requirements grow,
      change the pmtsCheckDSIProblems() check to reflect that (although not
      strictly necessary)
    (cherry picked from commit 4e4aec28c544f2884cb76be37d508f376db5cb7c)

commit 7734fb25ce7c6cb32961a8f980925f9c6e159c4d
Author: Florian Festi <ffesti@redhat.com>
Date:   Fri Feb 5 12:43:21 2010 +0100

    Remove no longer neede rpmts->filesystemCount and->filesystems
    (cherry picked from commit 27b216694a9e19f932f24e6fceb6c114aeae5fb9)

commit aac1b65645f730baa352de70f01b886f2b6de38b
Author: Florian Festi <ffesti@redhat.com>
Date:   Wed Feb 3 15:42:52 2010 +0100

    Fix stupid off by one error
    (cherry picked from commit ee35635243f1e282d04066014af6349227cd6498)

commit 13d8543caef7fbab171b2928839c282b87124120
Author: Florian Festi <ffesti@redhat.com>
Date:   Wed Feb 3 11:45:21 2010 +0100

    Avoid loading information for all filesystems
    
    This patch is heavily based on the "lazystatfs" patch by Michael Schroeder from
    Novell/SUSE. Thanks!
    (cherry picked from commit 647a809641a15cf0fcbc02ada85adcd493f2dc3b)

commit 01022b00058b6a3dad8ffa14bb143d834ccf88c3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 27 15:06:05 2010 +0200

    Fix include path of luaext subdir
    (cherry picked from commit 53e3f65b8e2f3b8883a8410c7732062155b87957)

commit 6329ce70b6505034f5e034944a7497192f07b716
Author: Ville Skyttä <ville.skytta@iki.fi>
Date:   Sat Jan 23 14:50:17 2010 +0200

    Document --conflicts in rpm(8).
    (cherry picked from commit d390e2544660e028bf42441c8cec57b4b92c018e)

commit d307da51f5df39a7392f245151ea820cbf0454a5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jan 5 10:45:23 2010 +0200

    Sanitize signature checking formatResult() a bit
    - whole lotta unnecessary wankery here (what have I been thinking)
    - avoid using the non-standard strndup(), we checked for it in configure
      but didn't provide an alternative so checking was just bogus
    - backported from commit 0edc291e50f9b6023a3e8e560d590e1dd2149a7f

commit 1ba072128dde446f1c51ad48c12e92ad654531db
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jan 21 14:14:28 2010 +0200

    Really fix verification error code to not depend on verbosity (RhBug:557101)
    - earlier fix from commit 6649c23cac5823d0c7415df5cfd66e5aac5eebd2 was
      broken as it could reset return code to zero after detecting errors
      earlier
    (cherry picked from commit 04ad4b1155ec4b082fbac425a575a0ba05337ab1)

commit 467da6d7936d24cf17c1a3a993ef4e3c9b177261
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jan 21 13:45:34 2010 +0200

    Fix double-free segfault on url retrieve to read-only media (RhBug:557118)
    (cherry picked from commit ce1d2e87157f4a7ebc4f4d6f26024ac597400367)

commit 00711e52f70f8273a3bae39e41e6b3c903b6b73b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jan 8 10:36:21 2010 +0200

    Preparing for rpm 4.8.0

commit 76a82dd314f355404b343aef5cc75210387fabb6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 17 11:35:11 2009 +0200

    Hardwire lead major version to 3 (RhBug:546663)
    - besides the logic being backwards wrt intended --nodirtokens linkage,
      LSB wants this to be 3 and .. it just makes no difference whatsover
      anyway
    (cherry picked from commit 0bac1d421ccdcc719e2e60bcfb4859d0f602510c)

commit 8f5a991259647ba07f694e0029b04311242ee44b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 17 10:38:08 2009 +0200

    Move python ts element key refcounting back to C-side of things
    - it's really the C-side that needs this, doing it in C avoids nasty
      surprises when subclassing rpm.ts
    - partial revert of commit 92d18d1d7139b37065ea742cfe0e8cee5e9c3500
    (cherry picked from commit edac0476a8ea3fbd1d86df2667a1b62d831b2653)

commit ba2a1ecc8c01926c170dfa06de3d8215da03f93b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 17 08:43:31 2009 +0200

    Permit unexpanded macros to slip through spec tag sanity checks (RhBug:547997)
    - For now, let %{} characters through our valid character set checks
      on specs. This isn't right really, but unexpanded, unrelated macros
      in spec can now cause failure to parse out buildrequires which would
      allow those macros to be expanded. This needs a better fix.
    (cherry picked from commit 507f21f6bb4bf7029a0bca255cfe4aae3361f358)

commit e66a51d84b10569943df278ede5f79d8a3bbfc93
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 14 15:23:28 2009 +0200

    Eliminate transaction disk space checking functions from the API
    - The DSI stuff runs from deep inside rpmtsRun() with no chance for
      API users to do anything with it. Bury into transaction.c and
      make static to let us change the way it works without API constraints...
    (cherry picked from commit 91eb212654b80d00202432da36e9746567d4f30b)

commit 371e23929502fe8c5a10fae094e582814a5b9a52
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Dec 9 15:11:39 2009 +0200

    Fix memleak in python ts.hdrFromFdno()
    - the python object holds reference to header here, decref the header
      so it gets freed eventually
    (cherry picked from commit 17eb18d0b1b259dbeb269148e6571ae49df9259f)

commit d90c1ce10c33fd219ed1840e93e0bdad10982193
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Dec 9 14:42:03 2009 +0200

    Fix a few list-related memleaks in python bindings
    - PyList_Append() bumps the object reference count, callers need to
      explicitly decref them... oops :)
    (cherry picked from commit 60b66dc7d9c728d6b779d3d772ac36b9134a2f56)

commit a2d6018a6144dd73a300fc33b30db14203dcf34a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 8 16:22:42 2009 +0200

    Avoid exposing libmagic dependency in public headers
    (cherry picked from commit 50d1b52edb202fbf9329cea0ecec285c0a73414f)

commit 3ffc0373c4db9824a72a6356d04d862e51f42be0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 8 15:58:51 2009 +0200

    Avoid noise from python bytecompile on non-python pkgs (RhBug:539635)
    - If a pattern doesn't match, we'd prefer an empty list instead of
      unexpanded garbage thank you... nullglob shell option is just what the
      doctor ordered but unfortunately its a bash-specific thing. Otoh there
      are number of other bash-specific scripts in rpm so probably not
      much of an issue...
    (cherry picked from commit d6bda5a1665d8d04526aea299781570ee7486b2e)

commit 139c031d2c54fe9e692765dae05da89d7ecfa8cc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 7 13:51:48 2009 +0200

    Preparing for 4.8.0-beta1

commit 7a2395941e3eb81740f6b7b5d417991ef9b42fa4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 7 13:35:53 2009 +0200

    Bump soname for all our libraries
    - while the "really public" API of librpmio has only been enhanced,
      librpm and librpmbuild use some supposedly internal bits which *have*
      changed considerably, so just to prevent anybody thinking they can
      combine librpmio from rpm 4.7 and librpm from this...
    - librpmbuild hasn't seen much real change but some (unfortunately)
      exposed struct members have changed so...
    - librpm is mostly compatible but as rpmal has become internal API,
      we need to bump (annoyingly, as nothing actually uses rpmal ... sigh)

commit 75c171f20891e58f1f14f37bba8e3e13b47bb783
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 7 12:38:39 2009 +0200

    ts.check() is supposed to return empty list when no problems

commit 9b94ae3dbc6d98f5c1929829735ff918aab5ef4d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 7 11:32:51 2009 +0200

    Disable the entire rpm.td type for now
    - need to figure out saner semantics & stuff...

commit f859d8aa7742eae7b5d48409a26086c6f5d80343
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 7 11:21:14 2009 +0200

    Make python header iterator return keys, not values
    - this is how python built-in dictionary works, be consistent with that
    - also avoids having to decide what to do with rpm.td type right now...

commit 1567738feddc9202308bc6e0be9c60d3adf5d56f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 7 09:12:44 2009 +0200

    Always return malloced data from i18n extensions for consistency

commit 105518cf5766bba6647e26b96e31b5a2d8ba18c4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 5 11:05:39 2009 +0200

    Bump up the python version requirement to match reality
    - Python >= 2.6 is known to work, 2.5 might be possible with __future__
      imports but until somebody tests it...

commit 46d1ba04483af09590ed070186a1be10b57c5977
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Dec 2 15:30:27 2009 +0200

    Fix the posix chown test for selinux enabled systems (RhBug:543035)
    - on selinux enabled systems, 'ls -l' mode string is longer than
      expected here, just trim the ls output to our expected length

commit 1adb46a9f864a327c93496ee0f8a149bd1def175
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Dec 2 13:22:11 2009 +0200

    Back out overly strict file type check on %ghost files
    - we currently only support creating regular files for %ghosts which
      dont exist in the buildroot, but for files that do exist in the buildroot
      there's no reason to place such limits (eg its perfectly ok to have
      symlinks marked as %ghosts)

commit 480095ab2a319d29965ced868ecc8253b1ba7420
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 27 15:57:44 2009 +0200

    Use AS_HELP_STRING() for configure option output

commit f0f94cc9df146a016f1d3c3b2d508b3e7336efea
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 27 15:12:52 2009 +0200

    Remove pointless message translation from check-buildroot
    - these scripts aren't processed for translation and unlikely to ever be...
    - fixes the remaining complaint from ticket #62

commit 45e4cca1ea3ed5073a54431dbf505e1e252fd678
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 27 15:10:57 2009 +0200

    Remove ancient broken scripts from the repository
    - also eliminates one more "bashism" complaint of ticket #62

commit 443d4f4d725a29d405b8c52e3a6ffbc1f514bd27
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 27 14:55:12 2009 +0200

    Fiddling with ulimit settings does not belong to dep extractor scripts
    - fixes one of the bashisms in ticket #62

commit 8440195af4f1d578fa3c631df8ff186c6168e368
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 27 09:28:11 2009 +0200

    Correctly pass optional tag argument from hdr.dsFromHeader() (ticket #108)

commit 8ba98b21789e6e3525cf2fd47a3a473aa17a0f9d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 27 09:15:56 2009 +0200

    Dont try to pass anything to rpm.fi() from hdr.fiFromHeader()
    - hdr.fiFromHeader() is just a deprecated backwards compatibility dummy
      and doesn't need to pass anything as there were no meaningful
      parameters in older versions

commit cfcd1f9bd98d5d0fc46a84931984efec3b9d47e2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 26 10:22:41 2009 +0200

    Dont fail build on unrecognized non-executable files (ticket #105)
    - Generally only executable files are critical for dependency extraction,
      whereas oddball application data files can cause unnecessary build
      failure due to libmagic misdetections etc, so just let non-executables
      pass with a warning and mark them as unknown data

commit ae74744e65e4e3e83f51cfae5418e8d7e12b9857
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 26 09:46:48 2009 +0200

    Just use autoreconf in autogen.sh (ticket #109)
    - instead of doing stuff by "hand" and getting it wrong, let autoreconf
      worry about getting it right (if nothing else, we can then blame
      autoreconf instead of ourselves ;)

commit 2b41860984f0c4ebba5ebce93a18c9c0ca5e1065
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 25 16:42:43 2009 +0200

    Fix signature password checking result on abnormal conditions (RhBug:496754)
    - Execve() failure wasn't returning an error code, causing rpm to
      think the password was ok when we couldn't even try verifying
    - Stricter return code checking from the password checking child:
      the password can only be ok if the child exits with WIFEXITED() *and*
      WIFEXITCODE() of 0. Also WIFEXITCODE() should only be called if
      WIFEXITED() returns true.

commit 98213fc4192c7af07037a0f3e9cce9e3b8509c02
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 25 15:07:17 2009 +0200

    We can't handle OpenPGP subkeys or secret keys, so dont even try
    - parsing subkeys ends up overwriting data in the main key, causing
      bogus signature checking failures
    - this is the final missing piece of RhBug:436812, short of adding
      proper support for subkeys (maybe someday...)

commit ded0c513836cee2f237edc2e28ee8aec261fd675
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 24 16:57:46 2009 +0200

    Dont blow up on headers without requireflags in rpmdsNew()

commit 649fc271d59eacbeeaacd3c24625ee95c884f0b0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 24 14:36:03 2009 +0200

    Nuke leftover debug junk

commit 39e15e7a678dfcfc0c03449b2ebc327569933f31
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 24 14:27:49 2009 +0200

    Support constructing header data from python
    - no proper i18n string support yet, just the basic types
    - basic data type validation only, no overflow checking for integer types
    - for lists, validate all items before trying to insert anything
    - duplicate tag creation not permitted
    - appending not permitted, only assignment (and deletion already)
    - there are probably bugs...
    - watch out for librpm bugs when passed incomplete headers, ugh

commit 897f06b9100437e6074ec5a816a83779662ddb4b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 24 14:02:52 2009 +0200

    Disable rpm.td type for now
    - its not that clear that exporting the td type is really useful
      (certainly not in its current form), avoid exporting an interface we
      might not want to support

commit 15c806b8ae53753481d09c0e980d668b44817b95
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 23 16:45:09 2009 +0200

    Fix hdrContains() return value on error

commit c225bb88cbe33d97ab9d7de20c4d6597b5394cc8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 23 15:12:04 2009 +0200

    Remove bogus rpmpopt.pot file (ticket #77)
    - this is supposed to automatically generated on update-po time from
      toplevel rpmpopt file and merged into rpm.pot but has been broken for
      a long long time and also the $ translation marking has been deprecated
      for security reasons

commit 1e2f382ff7912f743d4c18d85386506048d0d29d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 21 11:31:50 2009 +0200

    Remove hdr.has_key() method, support 'key in h' style instead
    - Python 3 removed has_key() from dictionaries, as the 'in' way is the
      way of the future support that from the start (has_key() is not in
      any released rpm version so its safe to remove)

commit 6711bcc6b948f8f60373af42e232e160c8cb89ed
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 21 11:07:29 2009 +0200

    Remove unnecessary hdr.instance() method
    - this is available through tag extension now, no point having a
      separate method for it

commit 6834d35460466992cfe43ce908443fb61f49b94c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 21 10:55:24 2009 +0200

    Typos in exception messages

commit 5e26a8be7193a90612d800eadba46b417c1d4944
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 20 16:00:06 2009 +0200

    Error out early trying to install on readonly fs (ticket #99, RhBug:464750)
    - simply make readonly filesystems appear as full, preventing install
      early instead of failing with cpio errors in middle of transaction

commit 58bece68422ba6cf20a1fe4ac4816ecd811e384d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 20 11:49:05 2009 +0200

    Convert French man page to Unix line-endings

commit 50773c56482f3cd7d48cbcad477d1a48a0395c66
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 20 11:12:39 2009 +0200

    Fix build with binutils-gold (ticket #107)
    - patch from Michal Čihař

commit d0eaa7c0485f670371fc6a99bf27021fdaa467e4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 20 10:32:47 2009 +0200

    Add RPMTAG_EPOCHNUM tag extension
    - return RPMTAG_EPOCH but convert non-existent epoch to 0, as rpm internals
      treat it

commit 71eeeb8d5e18ac6f6b349b09fd22fe2c650f8beb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 19 10:51:23 2009 +0200

    Add spec package type with access to headers to python rpmb module
    - permits python to access parsed headers from spec files
    - ticket #14, RhBug:462726

commit 4beda5ab2676d6b007034b6f9906724c813fce68
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 19 10:16:15 2009 +0200

    Permit header object generation from PyCObjects
    - needed for generating header objects from other extension modules &
      might be handy in some other cases too
    - patently type-unsafe but as long as python doesn't provide a better
      way, not really our headache

commit 2ba9153bfa8abd642c4529b454f65ed807f0afb6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 18 17:41:39 2009 +0200

    Lose the silly specFromSpec()
    - its our own private data structure we're accessing here, no need
      for extra wrappers and nothing outside spec-py.c needs this

commit 0e77aba7fb9f3e5cd875acc124ea4de2dea3c247
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 18 17:28:13 2009 +0200

    Turn current python spec methods into read-only attributes via getseters
    - incompatible, but nobody is really using the spec bindings for
      anything as there's been very little you've been able do with them
    - having the stuff accessible as getseters means we dont need 50 different
      fooSet() methods if/when we permit modifying these from python

commit ecb0f7df9a273f89706a6a8b3129dfcae29536d5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 18 17:16:45 2009 +0200

    Sanitize python spec.sources()
    - spec without sources is perfectly valid, just return an empty list
      instead of invalid NULL return without exception set and simplify

commit 8f056c7849944abdf2e5ee86e8668bc6c61834ca
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 18 17:03:32 2009 +0200

    Put a bit of sanity into python spec methods
    - prep, build etc missing from spec are not errors
    - instead of returning NULLs (with no exception set!), just return None
      for anything that doesn't exist
    - dont bother with NULL checks, if s->spec is NULL then something else
      has screwed up big time

commit 698c141fe3bfbc520a825e993a01a082cf80f69b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 18 12:21:41 2009 +0200

    Oops, tp_dealloc doesn't return anything

commit c6a2f7f3f293c0947ea015cea40618c78fc93de9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 18 11:33:57 2009 +0200

    Fix rpmProblem reference counting in python bindings

commit 67961448c192a932868efb6ee652bd228597d4dc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 18 11:22:55 2009 +0200

    Declare a few python functions only taking one argument as such

commit f6b5a7f622c0b34fc261cc0c27b596730d9597dd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 18 11:17:51 2009 +0200

    Rip python checkSignals() wtf'ery
    - since the function takes no arguments, declare it as such...

commit 841e0ad4ee360bf43acb1adb2b86977d7b74abe1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 18 10:54:58 2009 +0200

    Python archscore() only ever takes one argument, simplify

commit 66dd52d94c34b51eba9f0ef48bbe4df0e1b174ad
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 18 10:41:26 2009 +0200

    Make python ts.problems() return a python list, not rpm.ps object
    - rpm.ps object only supports iteration and subscript (with wonderfully
      wacko semantics), returning a regular list serves us better
    - rip the now useless rpm.ps object type

commit a237f1eee0b495691c2a9565ad2b0f5140c00298
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 18 09:40:30 2009 +0200

    Remove unused psFromPs()

commit a2a4599fc6947d30c5ecf42420528a91d1975093
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 18 09:35:41 2009 +0200

    Eliminate broken rpmps append() method from python bindings (RhBug:538218)
    - The code to insert new problems has been using invalid conversion code
      causing crashes since 2004 and nobody noticed, safe to say this is an
      unused interface. Additionally the method argument flags were wrong, it
      was declared as METH_VARARGS but actually expected METH_O semantics. RIP.

commit 2e76d0e6c83ca7bd08651c47fd1aee40f89333c4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 30 13:31:01 2009 +0200

    Add in-memory hash for caching rpmdb dependency lookups
    - worst case behavior for uncached dependency lookups can be disastrous,
      eg > 35s vs < 1s on my laptop for trying to remove /bin/sh provider
    - we only bother caching rpmdb lookups, the other cases plenty fast already
    - using in-memory cache avoids nasty in vs out of chroot issues with
      temporary db files, which otherwise were about as fast

commit 77c25db03a55113a77d6a56c422615f9cd4ec435
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 30 09:37:36 2009 +0200

    Document our new format extensions

commit 0dcf65119a0f7f96e460eaf7ef6aa970e9ee2f14
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 30 09:33:26 2009 +0200

    Add file verifyflags format extension

commit 491cf5e3442af05b79b33e346f60e565f343513f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 30 09:15:36 2009 +0200

    Add file state format extension

commit 18cc1d23120c5eadad251f0edde1697ed418a3e1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 30 09:13:08 2009 +0200

    Update rpmtd formats comments to match reality
    - we dont require exact int32 etc types, any numeric type works

commit 7ae58177d051f5989974b58930980371388b9d6d
Author: David Malcolm <dmalcolm@redhat.com>
Date:   Thu Oct 15 15:14:58 2009 -0400

    Convert relative imports within rpm package to absolute imports

commit 91a83f5e76790e5248780fa604ad2ac0aa3a68a1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 29 13:54:47 2009 +0200

    Permit any file-like object in rpm.setLogFile()
    - using PyObject_AsFileDescriptor() does away with another Python 3
      incompatibility and makes it more flexible.. but also more murky
      as rpmlogSetFile() only accepts FILE * streams...

commit 121b69b39529c2b5b9557eba96fd551e5295ccbd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 29 12:56:18 2009 +0200

    Update module initialization to work with both Python 2.x and 3.x
    - split common initialization tasks to separate functions, so that only the
      details of initializing a python module object differ
    - patch from David Malcolm

commit 16f9d7f4dcce8edc6458cd94ee0c7afda2c5dece
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 28 16:35:32 2009 +0200

    Remove unnecessary (cmpfunc) casts from our type objects
    - cmpfunc is no more in Python 3 and casting NULL to anything makes
      little sense anyhow

commit 70bebfdd7eab41d7e0aecf569692aac113958a3a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 28 16:27:15 2009 +0200

    Eliminate ambiguous tp_compare and tp_richcompare from rpm.ds
    - both comparison types are highly ambiguous, eg what on earth does
      "foo provides bar = 1" < "bar requires foo > 2" mean?
    - if straight ds1.evr vs ds2.evr comparison is needed, we'll be better
      off supporting that through rpm.versionCompare() which has clear semantics
    - avoids another Python 3 incompatibility too

commit 46b2df7c577b3f22a167bc46e655c9c008133a66
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 28 16:25:46 2009 +0200

    Add ds.Compare() as a thin wrapper over rpmdsCompare()

commit 1ce136261da3bd39ee10e31ed02d11a37da19126
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 28 16:24:16 2009 +0200

    Type-check rpm.ds() comparison/search arguments everywhere
    - also make them only accept one argument, simplifying arg parsing

commit 615c45ec34dc5c5c00aa50854524cd428e70de50
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 28 16:17:21 2009 +0200

    Rip out stillborn code

commit 3df9dc71118b498694437f083bb174d774cc6337
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 28 15:22:39 2009 +0200

    Eliminate the ambigious tp_compare from rpm.hdr
    - Avoids another Python 3 incompatibility as it doesn't have tp_compare
      anymore
    - We could add a compatible tp_richcompare, but comparisons like
      'if h1 < h2: do foo' are highly ambiguous, callers are better of
      using explicit rpm.versinCompare(h1, h2) instead

commit a5e123bc3e07fa6cc3fad13bc7f8a471ecb17d9b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 28 15:55:35 2009 +0200

    Use PyObject_AsFileDescriptor() instead of object type comparison
    - this permits any file-like object implementing .fileno() method
      (including rpm.fd) to be dup'ed, not just PyFile subtypes
    - this also avoids yet another incompatibility with Python 3 which doesn't
      have PyFile_Check() and PyFile_AsFile() at all

commit 590914c32c1d514d1a4e5c2244e1e38548a2e87d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 28 10:05:39 2009 +0200

    Dont bother calculating hardlink counts when the value is not used
    - speeds up non-verbose --list queries by some ~50%

commit 5463570061a7ca9ed947885b516c2551778b636f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 27 16:07:43 2009 +0200

    Support multiple python versions in brp-python-bytecompile (RhBug:53117)
    - For *.py files in python library path, use matching python interpreter
      version for compiling, this is especially needed for Python 3 which
      is very different from Python 2 code. And even if the code compiled
      with a wrong version, it would be recompiled at run-time when invoked
      with the intended python version, causing verification errors etc.
    - For *.py files outside the library path, use "system python" which
      might not be always correct but we dont really know any better.
    - Patch from David Malcolm

commit 714ebe92a1052e67967221e73ac8b788140e684c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 27 15:15:20 2009 +0200

    Argh, typo in previous commit

commit 376b1bd8b79392a3d8a019c0409c05c13b55f2b2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 27 14:52:08 2009 +0200

    Remove explicit Py_TPFLAGS_HAVE_RICHCOMPARE from ds type flags
    - HAVE_RICHCOMPARE is included in Py_TPFLAGS_DEFAULT since at least
      Python 2.3 which is as far back as we care, no need to explicitly
      set it. Eliminating the unnecessary flag removes another Python 3
      compatibility issue as the flag is gone there.

commit 2222e81e1cf74bbf78acb9969592ab06e2289929
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 27 10:57:35 2009 +0200

    Make disk space problem reporting slightly saner (RhBug:517418)
    - Only report disk space/inode issues for packages that are affected
      instead of piling up all the problems in all filesystems on all remaining
      packages
    - The messages are still ambiguous at best, need to figure out something
      better one of these days...

commit dcba14d575eb6396075e3976b898c98ddd17874f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 27 10:44:04 2009 +0200

    Let Python decide the "native" string type for tag names
    - Py_BuildValue("s", ...) returns unicode objects in Python 3, string
      objects in Python 2, this is what we want here

commit 490d79b976c2b52d838f72f553a20af0dcf81c1d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 23 22:10:44 2009 +0300

    Permit unicode as ts.dbMatch() key argument

commit d728e466d2055432b874b4777b3ca07ef3260518
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 23 21:25:08 2009 +0300

    Allow unicode objects as tag name everywhere

commit e13a476fc94509cb6f3d159e6135b48fdafcdf78
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 23 21:07:10 2009 +0300

    Add conversion function for handling python unicode -> string issues
    - In Python 3 everything is unicode, forcing each and every caller to
      convert manually is a bit much.

commit b457e2200bd41b0c7420e32d5c9ed5a10b082922
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 22 10:37:49 2009 +0300

    Make --httpproxy and friends work again (RhBug:529214)
    - macro fiddling to pass the macros set by --httpproxy etc popt aliases
      to the url helper
    - ftp proxy and http proxy control the same thing (http proxy) now to
      keep things simple
    - while at it, add %__urlhelper_localopts for local customization needs
      without having to override the entire __urlhelperopts macro

commit 4047397da1875b1e5b561e039ad6cd7d99206aa1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 21 13:21:01 2009 +0300

    On Python 3, use macros to alias all usage of PyInt_ to PyLong_ equivalents
    - Python 3 unified PyInt with PyLong, the former no longer exists
    - Based on a patch by David Malcolm

commit 1866fc41c8fdf5a82705cee7f1043d5fb634c3be
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 21 13:15:44 2009 +0300

    Replace PyString usage with PyBytes everywhere
    - In Python 2.6 PyBytes is just an alias for PyString, Python 3.0
      removed PyString entirely
    - Add compatibility defines for Python < 2.6
    - Based on David Malcolm's Python 3.x efforts

commit 33c569ce6fd827ba039e36ffe0a77b0643d0ac70
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 21 12:47:35 2009 +0300

    Ensure RPMSENSE_RPMLIB bit is always set on rpmlib() requires in rpmds flags
    - some packages in the wild have rpmlib() requires without the flag set
      due to older rpm not always setting it (for one, rpmdeps as external
      dependency  generator could cause such deps), causing bogus dependency
      failures on install
    - only fix up dependencies coming from a header, rpmdsSingle() correctness
      is up to the caller

commit 9fbaedd1ac655d978883d68a1f7d34df0d703363
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 21 12:23:50 2009 +0300

    Ensure rpmlib() dependency sanity on build
    - only requires on rpmlib(foo) are allowed
    - make sure any rpmlib() foo dependencies have RPMSENSE_RPMLIB bit set,
      rpmdeps as external dependency generator can create rpmlib() dependencies
      without the bit set, and also any manual rpmlib() deps wont have it

commit b6cfb4399ca4c5f2335c9351a7599ccfdf28add6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 21 12:22:33 2009 +0300

    Fix memory leaks on error exit from parseRCPOT()

commit cab7c23976fa83a57fcb63694748e0e40ca9fda4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 19 17:21:31 2009 +0300

    Remove leftovers from python header files
    - rpmSingleHeaderFromFD() and hdrLoad() are now implemented in python,
      declarations accidentally left here
    - make rpmMergeHeaders() static, nothing outside header-py.c needs

commit 72165c7717e26050c6785f8b103f69baaa06dd40
Author: David Malcolm <dmalcolm@redhat.com>
Date:   Thu Oct 15 15:14:59 2009 -0400

    Convert all usage of two-expression exception raising to modern syntax
    
    Replace all usage of the two-expression
        raise foo, 'bar'
    syntax with the long-equivalent
        raise foo('bar')
    as the old syntax was removed in Python 3 (PEP 3109), and the new syntax
    is supported in Python 2

commit a2999caaf4fb6a1d121b71d009d15a7a651d2bf0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 19 11:06:07 2009 +0300

    PyVarObject_HEAD_INIT() compatibility macro for Python < 2.6

commit 4b8e0ebde64321a8dbcbdce9f44557185fde1719
Author: David Malcolm <dmalcolm@redhat.com>
Date:   Thu Oct 15 15:14:57 2009 -0400

    Generalize type object initialization to work with both Python 2.* and Python 3.*
    
    The layout of PyVarObject changed between python 2 and python 3, and this leads
    to the existing code for all of the various PyTypeObject initializers failing to
    compile with python 3
    
    Change the way we initialize these structs to use PyVarObject_HEAD_INIT directly,
    rather than merely PyObject_HEAD_INIT, so that it compiles cleanly with both major
    versions of Python

commit 2f5a5c44cf5cd5406b153ba9ba37a7a5a4f4c6db
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 19 10:57:13 2009 +0300

    Py_TYPE() compatibility macro for Python < 2.6

commit 7b51c4a1eb8f14207f9a8d3e0be514a41c0e2197
Author: David Malcolm <dmalcolm@redhat.com>
Date:   Thu Oct 15 15:14:56 2009 -0400

    Generalize access to ob_type so that they work with both Python 2.* and Python 3.*
    
    Python 2's various object structs use macros to implement common fields at the top of each
    struct.
    
    Python 3's objects instead embed a PyObject struct as the first member within the more refined
    object structs.
    
    Use the Py_TYPE() macro when accessing ob_type in order to encapsulate this difference.

commit 63175d4d31469816c575ddb4de9698fb5c053536
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 19 10:30:48 2009 +0300

    Avoid using python's print for Python 3.x compatibility
    - print chanced from statement to a function in python 3, for our
      purposes sys.stdout.write() is sufficient and avoids dealing with
      the incompatibilities between Python 2.x and 3.x
    - suggested by David Malcolm / Ville Skyttä

commit 9707b5478bbad7251810247b2383f91b5bbabc16
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Oct 17 12:15:29 2009 +0300

    Unbreak lzdio/xzdio
    - revert part of b0d1038190be5cb4b45e2c2c7c84ee7022164ba2 which broke
      the split-personality lzdio/xzdio

commit 98cd707af0f284eeb4cf5864fbb14445d529bf31
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Oct 17 11:25:53 2009 +0300

    Skip posttrans scripts of failed transaction elements

commit f677a821adf7d18f5e9a9b053d805052a1bf5a39
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 16 13:36:39 2009 +0300

    Since we really want a FILE* then get one in the first place
    - this is visible in the librpmbuild exported goo, rename the member
      to force build-breakage on the poor souls trying to use librpmbuild API

commit b0d1038190be5cb4b45e2c2c7c84ee7022164ba2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 15 14:51:28 2009 +0300

    Use a helper function for finding fileno of specific io type
    - also differentiate between lzdio and xzdio here

commit 852398f8c6dcb4ad5ed0310e49e7d342a262be91
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 15 14:33:45 2009 +0300

    Lose unnecessary url type checking from fdSize()
    - we're only dealing with local files once fd is opened

commit ec68fc79551b27821bb400eb0136f8416b9ccb2a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 15 14:23:20 2009 +0300

    Lose unused leftover definitions

commit 2dc82d4e3e9c2959f4f731895993645761905073
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 16 12:49:08 2009 +0300

    Eliminate ufdio-specific read, write, seek and close
    - we dont do network IO anymore so ufdio only differs from fdio by
      downloading the file on open if necessary, after that it's just fdio

commit 7324a69d62f56faa9a621b0554723be85b6cac13
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 15 11:36:46 2009 +0300

    Use fd.tell() instead of going around through os.lseek() with fdno

commit ed40ff253a88ab399c28cf22005193e7389d5e23
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 15 11:34:26 2009 +0300

    Add tell() method to rpm.fd class

commit ed6ed5c1ce0854f1b8956a66e8a429c9d5ff3948
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 15 11:33:18 2009 +0300

    Add ftell() clone to FD_t

commit aa2ae3803cc38db786a662d70826a0131293b674
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 15 11:01:00 2009 +0300

    Avoid unnecessary dup'ing of file descriptors on header list operations
    - convert to rpm.fd on python side already add use internal helper
      which only accepts rpm.fd type to avoid having to dup around on
      every item

commit 410bf68fd41a6b8c98e48ca77e093f2f6c6bda75
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 15 10:11:37 2009 +0300

    Accept rpm.fd() types file objects everywhere in python bindings
    - turn rpmfdFromPyObject() into a python-level object converter, add
      a separate C-level getter for the fd pointer itself
    - take advantage of python refcounting to handle differences between
      native vs converted rpm.fd in callers so we can simply decref the
      rpmfdFromPyObject() result without having to worry whether it was
      converted or not (ie should we close it or not)

commit 50ad087c1030be242ab658c818ba2990dd33855e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 14 17:22:25 2009 +0300

    Make fdUpdateDigests() static now that its possible

commit efd874ae6a0ac17e269b1c5daed4f09939c94815
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 14 17:19:09 2009 +0300

    Make FD_t struct opaque, finally

commit e5be860abb675dff7f5488073f8f836cb5c908c5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 14 17:14:38 2009 +0300

    Eliminate last remaining FD_t struct privacy violation
    - add getter for fd digest bundle, uninline the setter

commit bdfb54bd238ab2d09175d2b921435b7a059ac25c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 14 17:03:14 2009 +0300

    Remove cpio stuff from FD_t
    - only the FSM cares about cpio position, move it there - there's only
      ever one fd active inside FSM, no need to track it per-fd

commit 1ff10e3d68417ebdeac329c79299187710afb697
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 14 16:15:05 2009 +0300

    Move debug goo defs to rpmio.c

commit 63432905a413d33085061674e00d6c23c60056e4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 14 16:10:08 2009 +0300

    Move pile of helpers from rpmio_internal.h to rpmio.c, remove unused stuff
    - nothing outside rpmio.c needs these, make 'em static there
    - uninlining reveals quite a few completely unused helpers...

commit f2e55c2ff8f6fba0cba61251027fe899ac2fdd1a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 14 16:02:00 2009 +0300

    Move fd statistics foo inside rpmio.c now that it can

commit fa0337c9b64d69b3a4e50d810ec63c7798cad6c6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 14 15:59:39 2009 +0300

    Move fdDigest*() to rpmio.c where they better belong

commit c716382ee899759b522b1b63e8a2acde9ca33f65
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 14 15:12:43 2009 +0300

    Remove pointless inlining of stuff in rpmio

commit 7ebe3cffc5666ffc2e04ae4722e8eaa64c0d6769
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 14 14:45:54 2009 +0300

    Use common "not implemented" implementation of Fseek()

commit aac3cb417cfd802748437b7e6e8b4b2182bc41a5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 14 14:14:51 2009 +0300

    Add first rough-cut python bindings for rpmio fd
    - attempt to mimic python file object where possible, but nowhere near
      all methods are supported, either just not yet done or due to
      underlying limitations

commit 181a3ac6a5b2d2017482da6327e0fc7a84e41c5e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 12 15:15:39 2009 +0300

    Raise exception in the converter, not caller

commit 597973befe1587decf9e1a069653b87e05be1a01
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 12 15:05:50 2009 +0300

    Permit changing rpm.td tag from python

commit c6649c55a89a9a6803c1640961de6c2412e9eeb0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 12 14:43:44 2009 +0300

    Add limited support for modifying headers to python
    - for now we only support tag deletion and assigning rpmtd objects, limiting
      this to copying data from other headers

commit bc716457e042e12f4e90efdabafe2cd69d3bff52
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 12 13:49:50 2009 +0300

    Fix format extension memleak on empty strings

commit 3ec070c5d8c16a593ab2573f0f63d58a748ba2c1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 12 11:03:14 2009 +0300

    Output dependency type on --requires etc queries if verbose (RhBug:528342)

commit df61816c7588bd00f5d6bfdb99f7539eec8bd62b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 12 11:02:56 2009 +0300

    Add RPMTAG_VERBOSE extension
    - hack to permit different query formatting based on rpm's verbosity level

commit bb47ec2695c4c2bf16b8995c53d5f92bc51cf3e3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 12 10:48:59 2009 +0300

    Permit conditional query formats on not yet cached data too

commit 856df71e8dc24219ec01619d2089a06e9ae36dfc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Oct 10 12:36:38 2009 +0300

    Oops, configuring wrong ts in rpm.ts() initialization

commit ac3d091577096a75814419e057b5c7823d0d6eab
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 9 15:44:06 2009 +0300

    Move further initialization from python ts tp_new to tp_init
    - lose the unnecessary rpmts_Wrap(), it's not needed and just gets
      in the way of things

commit dc6946e72eb69e67d1e00f741f5ba595fa740a18
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 9 11:57:46 2009 +0300

    Revert explicit PyErr_NoMemory() returns to just returning NULL
    - tp_alloc failing is likely to be OOM but we dont know that for a fact,
      and the failing method is responsible for setting the exception

commit 4e5132e443a9f062bfad83f531b43a9d5d6df1da
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 9 11:44:41 2009 +0300

    Handle NULL in ts get/setKeyring as python None

commit cbad134dd453a24d3e0e55ad81dff0e0baa2f7e9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 9 09:35:42 2009 +0300

    Kill off _rpmb.error exception type
    - import of _rpmb messes up rpm.error to point to _rpmb.error which we
      do not want
    - use ValueError for unparsable spec for now, need to figure something
      saner there still

commit 121ea41e33d19e415302c66e842a228212d80c22
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 9 09:30:37 2009 +0300

    Permit setting and getting ts keyring from python

commit 62888020f4ce618c447115e4b52015e4f7402f30
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 9 09:02:19 2009 +0300

    Add bindings for rpm keyring and pubkey types

commit 4c8afeef14708b3c25ebd8606d90991faff0c2c5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 9 08:48:13 2009 +0300

    Add an interface to parse armored OpenPGP from memory
    - refactor the decoding part out of pgpReadPkts() to separate helper,
      and add pgpParsePkts() around that

commit 0a2f3e12d25f0f71496cd2888b91b3399e23ca64
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 7 12:09:00 2009 +0300

    Support passing extra options to dep generator scripts

commit 468d5ab6cc1bd83bdef08d34e34d0cafc0c1d33b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 6 15:05:06 2009 +0300

    Lose unused python/perl/java stringbuffers from rpmfc

commit 733032dd1a682789ade39cfc0ef630d5bc83f09b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 6 13:22:22 2009 +0300

    Bah, get the new ocaml-scripts into tarballs too
    - should've been in commit 82e7dd702013d3679fda438333de30afdec17a4f

commit 35b20e13aec01252c952b3fd79665b916f5d0430
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 6 13:13:23 2009 +0300

    Python namespace import fiddling
    - import rpm instead of _rpm for exception name compatibility
    - import the lowlevel transaction class as _rpmts, makes it slightly
      easier to see which is which

commit 15fb8ccb41ce4cc02c78508b936fb74f7165683c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 6 12:34:44 2009 +0300

    Teach the internal dep generator about OCaml
    - OCaml seems to have fairly well detectable magic in its files, nice
    - Dont call the regular find-requires from the scripts now that internal
      is working

commit 82e7dd702013d3679fda438333de30afdec17a4f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 6 12:29:37 2009 +0300

    Add OCaml dependency generator scripts (Richard W.M. Jones)

commit 1c83af6e10302346a765e484d38cdac3ff619436
Author: Ville Skyttä <ville.skytta@iki.fi>
Date:   Tue Aug 4 00:48:32 2009 +0300

    Deprecate %__lzma, use %__xz in it.

commit 1892984f0216340f08e928d2bb74202cd8e87cbd
Author: Ville Skyttä <ville.skytta@iki.fi>
Date:   Mon Aug 3 21:08:55 2009 +0300

    Decompress lzma files with xz.

commit dab575b24498e2362845e7da6dc76015fa88b4a9
Author: Ville Skyttä <ville.skytta@iki.fi>
Date:   Tue Sep 22 21:11:20 2009 +0300

    Improve perl.req here-doc skipping.
    
    This version is stricter in finding the end identifier by using equality test
    instead of a regex (as a side effect, fixes rhbz#524929) as well as parsing
    quoted identifiers, and adds support for identifiers in backticks.

commit 562382ab180e2a4483485d165807f9de20fa5a1c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 5 18:23:55 2009 +0300

    Delete ancient and defunc python test programs

commit 55e6a4b905c1cf680a1235085e268b04e1b5405a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 5 17:23:34 2009 +0300

    Dont leak memory from rpm.fi() methods
    - Py_BuildValue() always copies the data mallocing here is just wrong

commit 3d1169fe698910fc62dd1893c861963dedea2945
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 5 16:42:52 2009 +0300

    Handle all rpmds variants in its constructor
    - either it takes a header + tag, or (name, flags, version) tuple + tag
      to generate a single dep
    - push rpm.dsSingle() compatibility out to python side
    - hdr.dsFromHeader() and hdr.dsOfHeader() no longer need to be in rpmds-py.c

commit 0cddef5a4a2b279d0da39c765312cc2deaa7de07
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 5 15:38:45 2009 +0300

    rpmdb open failing is certainly an rpm error, not TypeError

commit e3a1720f4e1f5ae972a8f56275113adca179c80e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 5 15:36:54 2009 +0300

    Dead code removal

commit a9958395c7000c73fe9275046f50f6b276642d48
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 5 15:25:59 2009 +0300

    Permit threads during rpmReadPackageFile()

commit af00066294bb180b609d6f612237c14ad48f0dec
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 5 15:22:12 2009 +0300

    Type mismatch is TypeError, not rpm.error

commit 564f1cb02b841f0b0b1c2b7fcafbb9a080233a46
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 5 15:19:01 2009 +0300

    Push hdrFromFdno() error code handling over to python side
    - return (rpmrc, header) tuple from C to let python do whatever it
      wishes with the information
    - let python side worry about generating backwards compatible returns

commit 60ec5008bd6bc5f9cec43f64dd8129045080fa41
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 5 14:55:37 2009 +0300

    Push hdrCheck() error code handling over to python side
    - return (rpmrc, message) tuple from C to let python do whatever it
      wishes with the information
    - let python side worry about generating backwards compatible returns

commit 0503d7f92e94a7986715756da5226fc18a1fba10
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 5 14:30:27 2009 +0300

    Teach python about the lowlevel RPMRC codes

commit fab2debfe440d677dbd072c3cd73d2c99876e7a5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 5 14:25:43 2009 +0300

    Push ts.check() return tuple-o-doom generation over to python side

commit 776f5b56ca51c7c5495f84d5823e5e646206e5d4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 5 14:08:48 2009 +0300

    Teach python about RPMSENSE_ANY

commit 4e647fd7805e85374626dd8f439f3b059ea22409
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 5 13:12:05 2009 +0300

    Enable retrieving problem key from python too

commit 66589df18f7a54f724df5e859334418d9496e5ce
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 5 12:16:26 2009 +0300

    Push transaction problem filter to python side
    - make C-level ts.run() require ignoreSet just like the real one does
      and keep the value on python side of things

commit 44ee4e164e340c0d4e4725cc57811ce637deaef3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 5 12:09:26 2009 +0300

    Push the interpretation of ts.run() result code to python side
    - at least this lets sub-typers override the hysterical return values

commit a15bfa40adc826dfa91f9bdc24b3e680cd8aecb6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 5 11:09:05 2009 +0300

    Add minimal python wrapping for rpmProblem objects
    - for now, only expose the basic attributes and return problem string
      representation on %s
    - return problem objects on rpmps iteration

commit 5b289d4491738eaa4f106eb69c70680406d63aa0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 5 09:26:43 2009 +0300

    Ugh, forgot to add the new _rpmb module to git
    - should've been in commit 8169bbde6934637ed7be58e18103330c1d5e4546

commit 9b20c706a4f93266450fae2f94007343b2e8fd9e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 2 22:48:03 2009 +0300

    Push most work of ts.addErase() over to python
    - minimize the stuff done in C: rpmtsAddEraseElement() only cares about
      header, so that's what we accept on C-level bindings
    - on python side, handle db recno's and labels like we've always done,
      additionally accept match iterators and plain headers too

commit ab04cfb08c5fbdf4e9803b2d63352fee0527937e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 2 22:15:51 2009 +0300

    Return error code from rpmtsAddEraseElement() failure
    - trying to remove non-installed header counts as an error

commit a9b77304c3ad187e8982fd67d1a6406c93ab9e40
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 2 21:41:41 2009 +0300

    Permit file objects and file names to ts.addInstall()
    - largely removes the need for the klunky ts.hdrFromFdno()

commit 92d18d1d7139b37065ea742cfe0e8cee5e9c3500
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 2 21:18:27 2009 +0300

    Push most work of ts.addInstall() over to python
    - minimize the stuff done in C: require all arguments in final format
      and report status back with a boolean, handle everything else including
      keylist refcounting hack over to python side

commit c7d62396bf3420a87ce0cd7c319d2566f9446ed3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 2 21:14:19 2009 +0300

    Oops, missing self in bunch of calls... been too long since writing Python ;)

commit 5ce5e8d75b1683fc4679037db833b55884ee1842
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 2 20:35:20 2009 +0300

    Push the rpmtsGetKeys() stuff over to python completely
    - trivially implemented in python by iterating over transaction element
      keys, we dont need no stinking extra librpm APIs for this

commit 07ad9dcf0a8eb98928178b774512bcad2bc6f169
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 2 19:19:31 2009 +0300

    Add rpmlog() wrapper to python bindings

commit bc39e17527920d47640ea35e42c961926e4e5125
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 2 18:55:16 2009 +0300

    Teach python about RPMDBI_LABEL

commit 8169bbde6934637ed7be58e18103330c1d5e4546
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 2 18:31:32 2009 +0300

    Split rpmbuild dependencies to separate _rpmb module
    - installers and the like dont need the pile of poo called librpmbuild for
      anything, avoid dragging it in needlessly
    - import _rpmb into rpm namespace if available but dont complain if
      its not there
    - arrange ts.parseSpec() compatibility on python side by dynamically
      importing _rpmb if possible

commit bc9b86ae5c14e093078506b533cede3c5f1f5ab9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 2 16:36:53 2009 +0300

    Move ts flag & color setting to python side

commit 305d634d27d8c480bc01250585d2a901e343851f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 2 16:18:59 2009 +0300

    Add a dummy python-level transaction set class, use it always
    - move the rpm.ts() vs rpm.TransactionSet() compatibility to python
    - this will enable doing only the lowest level stuff in C and rest
      in python

commit 1dffab5393fc8ed5cd81173be543db6b5c34785a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 1 15:25:01 2009 +0300

    Add python method for retrieving header "instance" (db recno) number

commit 9c88ea2c87b5d22026400fb19f064b7967aa4743
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 1 15:12:06 2009 +0300

    Deprecation tweaks
    - use PyErr_WarnEx() for better control and leave the tracking up to python
    - use PendingDeprecationWarning for now
    - document the replacing functionality in the deprecation messages
    - make hdr.sprintf() just an alias to hdr.format() without deprecating,
      at least for now it'd be only a gratuitous incompatible change on python side

commit 8b704a6bf8ae2c5a922b54ff35c65525e93f0f1f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 1 14:55:55 2009 +0300

    Oops, these are mappings, not dicts...

commit 29eb2f35047363353c851e3f8920123ce0a42217
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 1 14:27:42 2009 +0300

    Generalize python ts flags and colors set/get foo
    - make flags, vsflags, color, prefcolor (previously not available)
      and probFilter appear as internal attributes of ts objects, each
      with their own setter/getter methods
    - make the old custom set/get methods access the internal attributes instead
    - keeping them "internal only" leaves room for adding further sanity
      checking and nicer interface on python side while giving direct access
      bits for those who want it
    - not deprecating the old get/set methods at least yet...

commit f819760dc37047d464afff8a91c746114c183a2a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 1 14:06:41 2009 +0300

    Fix couple of recently introduced compiler warnings

commit b0d374528a367779e158a68a7b6a58ad929b85aa
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 1 13:21:44 2009 +0300

    Include structmembers.h centrally from rpmsystem-py.h
    - pretty much everything might need this...

commit 66c469759c497b72f24ebb6a69ca9a6333d1d600
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 1 11:37:45 2009 +0300

    Add API for setting preferred transaction file color
    - not strictly needed as it gets its default from a macro but .. shrug

commit 16e35ece9c28cfd0257102b1e1ca69746e2b4b6e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 1 11:21:58 2009 +0300

    Make python ts rootDir appear as a read-only attribute
    - there's no need to change the rootdir after object creation, better
      not permit at all

commit 24135b8443d01af4ff0c23ec11794787b0feb673
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 1 11:17:38 2009 +0300

    Add python getter for transaction id

commit eb419a86ac70aef55b8af3b964b62ada82944582
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 1 10:59:53 2009 +0300

    Eliminate python ts scriptFd kludgery
    - make the scriptFd appear as regular attribute by providing a setter
      method for it, permit any file object to be used and allow disabling too
    - readonly for now as we dont have rpmfd wrapped yet
    - remove now unnecessary custom get/setattr functions

commit 909982b1b5d2bccbf28ecfb38064ab86e09d4523
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 30 16:45:46 2009 +0300

    Deprecate custom foo.count() methods, support len() instead

commit 8cb920eb32f36297e2154b7ebcc25851d9a7b39d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 30 16:10:24 2009 +0300

    Eliminate unused python ts iterator filter stuff
    - iterators dont take arguments so there's no convenient way to filter
      while iterating, let python users filter themselves (like they've
      always been doing)

commit b8d3b273cbcc7a4130d6786ff157a87a4f383c30
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 30 15:04:39 2009 +0300

    Take advantage of headerNextTag() in python hdr.keys()
    - avoids whole lotta unnecessary copying of data, we're only interested
      in tag numbers here
    - dont filter out stuff - python consumers can look at i18n data too

commit 32b0bba8a2fb5e534bddc9c02ff170726457bdcc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 30 14:51:33 2009 +0300

    Add iterator support to python header objects

commit 3d5455d42d48740ff23899f00a903f49b73eff00
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 30 14:48:19 2009 +0300

    Add alternative header iterator method
    - we dont always want the actual contents of the tag to be copied
      on iteration, so add an interface that returns the next tag in the
      header
    - this lets callers to decide what to do with the tag and how to
      retrieve it

commit d59e715c1b8e28206bfed45b6f4cf3c00322f95c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 30 13:05:24 2009 +0300

    Add some flags to rpmtd creation
    - permit disabling extension retrieval and "raw" (untranslated i18n) tags
    - always use HEADERGET_ALLOC for data availability sanity

commit 1ddee37628b8aac48a8f1fc51af4575b5570589b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 30 12:45:07 2009 +0300

    Add beginnings of rpmtd wrappings to python
    - unlike other types, store the C-level td structure directly in the
      python object, this lets us selectively expose some members directly,
      avoids having to deal with rpmtd allocation separately and as leaves
      the reference counting to python as rpmtd's aren't refcounted on C-level

commit b53f5995969554211dde0b235c648f8d2e1cb0a6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 30 12:11:43 2009 +0300

    Turn hdrGetHeader() into argument parsing converter interface

commit 945bbf49f52c50b53f0d407e91ec46febd56d948
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 30 10:33:52 2009 +0300

    Sanitize getattr behavior of header objects
    - when generic getattr fails, only try retrieving header tag as attribute
      if it's a valid tag
    - clear any python errors if generic getattr failed
    - split generating a python object from header + tag to an internal helper

commit 600ccbe9d43e5857eb9b9925975f63d0e40286d8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 28 18:22:14 2009 +0300

    Require expected type in PyArg_ParseTupleAndKeywords() already

commit ac3f9f729b9a3b938d3344278c7e0564934f3871
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 28 17:23:28 2009 +0300

    Support numeric expansion in rpm.expandMacro()

commit f74def684253964fa03edb55f155dfbb5122b1bf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 28 16:37:14 2009 +0300

    Push rpm.signalsCaught() to python level
    - only implement the bare minimum in C by adding a thin wrapper for
      rpmsqIsCaught(), the rest can easily be done in python

commit dca55d3781401eb0ddce2759770054db7629083a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 28 16:19:22 2009 +0300

    Implement rpm.readHeaderFromFD() in python instead of C

commit 07710f3033b277ccb46decd84f994e88d5e6869a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 28 16:07:09 2009 +0300

    Implement rpm.readHeaderListFromFD() in python instead of C

commit f0e9245660100d3ce41692889899844f42feb7a4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 28 14:50:33 2009 +0300

    Implement rpmreadHeaderListFromFile() in python instead of C

commit 90042df3428414bf5f68cb6db7aed208d6455b96
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 28 13:48:19 2009 +0300

    A few cases of Py_RETURN_NONE missed on the first round

commit 65a6afbbdaf1fc81304c66ffb1b024d63c4bbb5a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 28 13:30:28 2009 +0300

    Arrange rpm.headerLoad() compatibility on python level

commit a22fda5b5add3b96ac68a39ab8a6c88b72285098
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 28 11:09:32 2009 +0300

    OSGi dependency generator fixes from Alphonse Van Assche (ticket #101)
    - This patch fix some minor bug, remove unused code and increase drastically
      the perf by threading execution of tasks and using zip command instead of
      jar to get meta-data

commit 2507ab6797926e9482ad5311697699f0a55f50e2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 24 13:57:55 2009 +0300

    Enable subtyping on the rest of our type-objects
    - not very useful atm as various places return hard-wired built-in types

commit df40d9b27b1dc8bd82f8f1c8a17d357df3fae330
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 24 13:40:44 2009 +0300

    Eliminate all custom tp_free() type methods
    - tp_free()'s purpose is only to free up the memory used by the python
      object structure, cleaning up our own allocations belongs to tp_dealloc()

commit d5f5156bd09338eceba7eeb6c5122ecc58afb0a1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 24 13:19:20 2009 +0300

    Enable subtyping on rpm.ts class

commit 374358d219b5fc35b43f929cb285c3d3d473ac89
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 24 13:05:36 2009 +0300

    Call generic python setattr if not scriptFd
    - yet more preliminaries for subtyping

commit 27bb9c9402f38dd802d61a5945f6b019c8b8ed78
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 24 13:04:34 2009 +0300

    tp_setattro() is supposed to return -1 on errors

commit 74c1040dbc4b4977045745f924ea0774c2e3ee53
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 24 12:52:32 2009 +0300

    Call (sub)type tp_free from destructors
    - more preliminaries for subtyping
    - remove pointless NULL checks

commit 36ada6c116e123f5b5b739e12f256201e5cc6fa1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 24 11:42:17 2009 +0300

    Make object allocation type agnostic
    - pass (sub)type down to wrappers
    - call subtype tp_alloc() instead of PyObject_New()
    - preliminaries for allowing subtyping

commit ab66c9ff47664fcfae6eea2146be65ccbff43195
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 23 17:44:07 2009 +0300

    Remove unnecessary header type checking
    - the type is already validated by PyArg_ParseTupleAndKeywords(),
      no need to doublecheck

commit 693838658d0a1c94bc9efa31de700e2a9898c88c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 23 14:41:30 2009 +0300

    Remove the broken tp_print method from spec

commit d55c5f6c3005858903eca1a5e98915f0c27cfb8b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 23 14:21:13 2009 +0300

    Decouple python spec objects from transaction objects
    - ts structure is only necessary for hysterically passing back the
      parse result from parseSpec(), hide this in the bindings
    - deprecate ts.parseSpec() for later ripping

commit 86527f0d45c033347880907cf7680230ecb9ecb9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 23 13:12:43 2009 +0300

    Oops, binary data can and should be presented as python strings

commit 76d8d16de0c96f2b3e66eafe9f03127e96ad5804
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 23 12:49:15 2009 +0300

    Add rpmtd to python object converter, change header code to use that
    - vastly simpler than the former goo in hdr_subscribe

commit 282efb6d4c23848b2443ffd767fd4bccda209425
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 23 12:28:47 2009 +0300

    Turn tagNumFromPyObject() into an object converter interface
    - permits direct validation and conversion from arg parsing

commit ac219c129d04952fbd205c536533723d582d1ad2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 23 12:05:02 2009 +0300

    Turn rpmFdFromPyObject() into an object converter interface
    - permits direct validation and conversion from arg parsing

commit e72a51d081b192cbc35af47f8281e939370f8b55
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 23 11:51:55 2009 +0300

    Permit any file object in rpm.readHeaderListFromFD()

commit 0913ff321b5a2034596e00c463a4888f96c19aca
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 23 11:42:59 2009 +0300

    Permit file objects in python header constructor

commit 1d2d0456890be9341f031d854a7feb69c87b2419
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 23 11:41:20 2009 +0300

    Add write() method to python headers

commit b997746271d764bebc86f207871e5e0e57e7b720
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 23 11:10:38 2009 +0300

    Accept any file object in ts.hdrFromFdno()

commit 79e87e5ea16ed7118536b12e267d93db74e15265
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 23 11:09:19 2009 +0300

    Add helper to convert python file objects to rpmio FD_t type

commit 6b9adb71f9715de3648673ae6ca4475a9e98d141
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 23 10:49:47 2009 +0300

    Add te.NEVRA() method, adjust docs

commit c29492242c97ebed80bdf561c4a15acf9d7939ca
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 23 10:44:41 2009 +0300

    Remove tp_print methods from all rpm-python objects
    - these violate the intended use of tp_print, python docs state
      "A type should never implement tp_print in a way that produces
      different output than tp_repr or tp_str would."

commit 6bbc19fb228465215b59cd2c902564d7665ce554
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 23 10:39:40 2009 +0300

    Lose the debug junk from python bindings

commit 2ede83b6598a3026c1f016a109220929f62b0829
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 23 10:31:46 2009 +0300

    Add hdr.format(), deprecate hdr.sprintf()
    - sprintf() is a C'ism and format() is closer to current librpm anyway

commit f0c3985a773706d8dc268d143120142acfbfa822
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 23 10:19:46 2009 +0300

    Make fiFromHeader() static inside header-py, deprecate

commit c21a1ea23db40bd136a859eb171a49ae51e25c64
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 23 10:11:45 2009 +0300

    Deprecate rpm.headerLoad()
    - this is not needed now, just call rpm.hdr(<blob) to load

commit d38f0eb5422252c4750eac421f0fb8c7c8d43a16
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 23 09:56:08 2009 +0300

    Deprecate header filelist munging methods
    - hdr.convert() is the generic way, the old stuff is not needed anymore
    - macro to issue warnings just once

commit 66c0d4ee6b96190426d310e6b0be4c7ce3d7fe42
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 23 09:27:47 2009 +0300

    Eliminate unnecessary function

commit e16815fc7bc40198b60f677ff7d80a84cb5356a6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 23 09:24:23 2009 +0300

    Permit headerConvert() from python

commit 42810ceae18d920f40f5193cf900f1c96bca7186
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 23 09:03:25 2009 +0300

    Add has_key() method to header object

commit cfc341170cdd5b324a3b90eb1b13620fd39298b4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 23 08:51:12 2009 +0300

    Add support for creating new header objects
    - without arguments, rpm.hdr() creates an empty header
    - with another header as argument, make a copy
    - string argument is considered an unloaded header blob and copy-loaded

commit 848de2dfec3f545801574a6219df8f081d66c5e0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 23 08:41:03 2009 +0300

    Rpm's allocators dont return on OOM, remove unnecessary check

commit 2a8da8606fc169aa9929e9a1550c8e29c466880b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 23 00:06:13 2009 +0300

    Add type checking macros for all rpm python type objects

commit 39d79f91420ba7d9527e04d00b4b0dc2d6cad9ff
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 23:02:47 2009 +0300

    Rename python system.h for disambiguation

commit 14d5aaedcb355a724ba8869a29c8e4a35efdcac1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 22:42:06 2009 +0300

    Lose the empty doxygen markers
    - nothing wrong with comments but empty comment placeholders
      are not exactly useful

commit 864220c441704e3d61fa521c682a23874b41e4ba
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 21:53:21 2009 +0300

    Put some consistency to python object creation
    - all type object creation goes through foo_Wrap() which handle OOM
      and all type specific initialization

commit ed557bbcf065905beebb42d50048cecf04c3e441
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 21:24:55 2009 +0300

    Make all python object creation wrappers return PyObject pointers
    - this way the only place where casts are needed are in the wrapper itself

commit 4d6f8e46e03ccb81a0f50848e974884b7874912d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 21:01:01 2009 +0300

    Dont leak memory on rpm.ds iteration

commit 0cb5de3fc090cae426b812ad3e38e2be8b7e63ac
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 20:50:41 2009 +0300

    Lose unnecessary next() methods
    - python adds next() methods for objects supporting iterators

commit 0793b2cf23667f515c1675e9dbe37e634c4cd730
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 20:20:07 2009 +0300

    All rpm-python iterators are self-iterators, just use PyObject_SelfIter

commit d7134660ca4300058e579bed990ab7cf7e14a40f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 19:55:34 2009 +0300

    Dont leak memory from rpm.dsSingle()

commit 07adc4e0d1aa23a4d3e3a21e8f01f615233793ee
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 19:44:59 2009 +0300

    Simplify fiFromHeader()
    - just call the fi object, no need to manually redo all this stuff

commit 708d4d8f1bd215d1dffce3f8d5680d2c967145b8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 19:42:53 2009 +0300

    Simplify dsFromHeader()
    - just call the ds object, no need to manually redo all this stuff

commit da74188fa654246ec74e0375440f326e5dc714f2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 19:19:02 2009 +0300

    Sanitize python object -> tag number exception handling
    - raise exception in tagNumFromPyObject(), not in 12 different callers
    - check against RPMTAG_NOT_FOUND consistently instead of -1 and whatnot
    - unknown tags are value, not key or type errors

commit 00d683ce88748080ce78d1e2d802a401b2d30261
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 18:55:09 2009 +0300

    Move allocations out of rpmps object init method
    - tp_init can be called several times, allocating from there leaks memory
    - tp_init gets called automatically on object creation, dont call manually
    - nothing to initialize for rpmps...

commit 4c85c7e605ee0312e51c8614d98c4740041ec4e1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 18:53:15 2009 +0300

    Move allocations out of rpmds object init method
    - tp_init can be called several times, allocating from there leaks memory
    - tp_init gets called automatically on object creation, dont call manually

commit fdc62b3decfe6a798284fc8c76372092b3ee1f00
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 18:50:32 2009 +0300

    Move allocations out of rpmfi object init method
    - tp_init can be called several times, allocating from there leaks memory
    - tp_init gets called automatically on object creation, dont call manually

commit 337cb1bfea878931bf4bc410edeaec58a857b013
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 18:22:50 2009 +0300

    Remove useless rpmts python initialization

commit e2ba570ad793dcbfdf8a80b520a54995aa56195d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 18:17:49 2009 +0300

    Realize the remaining bits of direct rpmdb interface are dead too
    - the code hasn't been enabled in a long long time despite being built...

commit 96a1b0a6816f3df9b4ff5290520b72b904182f36
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 17:49:53 2009 +0300

    Use Py_RETURN_NONE macro for returning None everywhere

commit 43f585fced06e10d4fad8208697cc323857ef535
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 17:33:07 2009 +0300

    Eliminate unnecessary custom object allocation functions
    - these are just calling python defaults, no point at all...

commit ca6591e86c201e54ae8ee876f12d5c0f0c835236
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 17:27:44 2009 +0300

    Use generic python get/set attribute functions directly where appropriate
    - no point in wrapping all this stuff in our own functions

commit 711ab9bab1596dc61825ec8c71d6bcd1f9a1917f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 17:12:22 2009 +0300

    Python initialization spring clean
    - split rpmtag enum additions to seprate function
    - PyModule_AddIntConstant() does exactly what we need for the rest

commit ae1fd3fa479bfafcc65b9ea532cafb401ed9a031
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 16:47:07 2009 +0300

    Make the python object structures opaque

commit 2b06ca079c90960242dfe173e0e8c2fbc39d7689
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 16:39:39 2009 +0300

    Lose the utterly broken rpmio FD_t wrappers

commit ae6ea9789ff395e2343d7c7a9eae551a64a74db5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 16:36:48 2009 +0300

    Lose the useless python debug stuff

commit ae1e7d83898e3c14138260ac9a8b53737da83c84
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 16:28:58 2009 +0300

    We dont support ancient python versions...

commit dd5cfeec61eeddedc842aa11d7ba2376c0e9411e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 16:24:32 2009 +0300

    Kick out mummified rpmdb python interface remnants

commit 5db17ad1ba05d03d7d543b064ee8ebf21fe9118c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 16:02:39 2009 +0300

    Missing <string.h> include

commit 6f7120694bc7f3a1c56d4639ad345aaa024ba598
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 15:36:11 2009 +0300

    Clean up + fix bad recno erasure in python ts.addErase()
    - now that we can loop over RPMDBI_PACKAGES matches, the code can
      be simplified quite a bit
    - walk through the iterator always to get a reliable installed status,
      previously ts.addErase() with bad record number wouldn't raise exception

commit e95921e29bdaa15094499d9191cd6fbfd0b7e63a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 14:53:56 2009 +0300

    Remove no longer necessary RPMDBI_PACKAGES special casing

commit 21d527f6c7111c067fc88a6ef76d201c7a653936
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 14:52:57 2009 +0300

    Fix endless loop on RPMDBI_PACKAGES iteration when offset specified

commit eae127b7d835a2cdccd0abad75141efd8cb3a2bf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 12:32:02 2009 +0300

    Unbreak --tid and --querybynumber
    - broken ages ago somewhere between 4.4.x and 4.6.x by typing changes
      presumably
    - C99 strtoul() can figure the base on its own, remove unnecessary work

commit a90f65f38b17b7ba993d9d4b7d1677ce0fd4417a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 10:21:47 2009 +0300

    DB_INCOMPLETE was removed in BDB 4.1 and we dont support older versions

commit dff73774ca68b1ab79534547fc744c510da8f0e5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 22 09:26:39 2009 +0300

    Remove ancient, unused code to filter duplicates on rpmdb rebuild
    - enabling this as-is would be murderous to multilib systems
    - getting duplicates in the db would be a bug elsewhere

commit 6bc5d87007bfe2ecfe6738f0575b3646a71361fd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 21 14:36:59 2009 +0300

    Rip out dependency caching
    - it doesn't speed up things that much, is broken in some chroot
      scenarios and is ugly ugly hardwired BDB hackery where it doesn't belong

commit 880f4cd41f8182c1c39f6678f79266932cccf082
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 21 14:04:34 2009 +0300

    Move rpmdb home creation from backends to one level higher
    - this is a common operation for any backends, avoid duplication
    - only try to create it once instead of every index open, and if it
      doesn't exist and cant be created, error out instead of trying to continue
      in vain

commit d3c75ad8eef518f413fb27184f9d81799a823e9b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 21 14:00:46 2009 +0300

    Add an internal API for getting rpmdb home wrt chroot state, use it

commit af136766a85ca003fbe8332e6ec9c63eeb4d8cd7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 21 11:59:25 2009 +0300

    Remove pointless per-index api version

commit bdea28b94511af2dee0e30635062024690e23f0e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 21 11:36:16 2009 +0300

    DB config stuff not needed outside dbconfig.c, make them static
    - additionally make the popt struct const since it can...

commit 81714312e70835c2136372d40813bd4c8423d3e3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 21 11:32:43 2009 +0300

    BDB is not optional at least currently, remove useless ifdefs

commit a232a259845adc620a187c54fb49ec91a3493e97
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 21 11:29:27 2009 +0300

    Remove useless no-op "teardown" and "usecursors" config bits

commit baa1cf0e1bdb7260be1cb41dc5e6c3bad180a34e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 21 11:00:50 2009 +0300

    Remove BDB RPC "support"
    - it's insecure, used by nobody (its been rather broken for a long
      time) and all, but BDB 4.8.x removed RPC support from upstream so we
      cant very well support in the future even if we wanted to

commit 510c6763ef745b308abd1cc6cec2a62119e87c9c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 21 10:29:38 2009 +0300

    Remove XA "support" from the backend
    - it's removed from BDB 4.8.x and this has hardly ever done anything anyway

commit a819de43ce019bc3dd305799d12c1169294c93a7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Sep 18 13:53:46 2009 +0300

    Remove unused db/dbenv methods "documentary" from the source

commit 4c12b61b34ad8da1927c25805277e8e2f90c1b80
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Sep 18 13:05:49 2009 +0300

    Use db3Free() for on sqlite backend close too

commit 37390dffe42360f5f34ef86bd7296d029c16adb7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Sep 18 13:13:59 2009 +0300

    Always populate dbi->dbi_file on db3New()
    - Only the actual db->open() wants NULL for temporary databases, everywhere
      else we want the tag name. Handle the temporary case where it counts
      instead of complicating things everywhere for the one special case.

commit b3ed4ec1b1943ffc74a5877cc6f8cb85c37da885
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Sep 18 11:56:05 2009 +0300

    Mop up leftover root/home foobar

commit 26e36775157a53ee68eff33b4863bb0e8f60f5c8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Sep 18 11:51:21 2009 +0300

    Per-dbi root and home dont make sense, always use rpmdb paths

commit f3cab403d4b3a4351037e6eab3832e92b02ae8fb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Sep 18 09:49:28 2009 +0300

    Sigh.. when not using environment, absolute paths are required

commit c01766b3f60802f9d4b1aa56151f633323377915
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 17 16:28:09 2009 +0300

    Always open database files relative to database home
    - This allows the db to move freely wherever (such as across chroots,
      RhBug:507309 and pile of related issues) .. and is even recommended
      by BDB documentation. Well duh.

commit b3dcc48824c6ddefdfe784148f3a5da0bd60ed90
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 17 12:51:28 2009 +0300

    EXIT_FAILURE being defined is already handled in system.h

commit b805731ce3d3cb9d79c224d936c41a10861baed5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 17 12:40:29 2009 +0300

    Compare pointers against NULL, not 0

commit 28919a1677f6f0e2741d9554c4bd30eb98825fb5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 17 12:25:31 2009 +0300

    Have BDB use our allocation routines now that we can
    - rfree() prototype doesn't match free() but it doesn't matter in this
      case as it's just a convencience wrapper with no extra functionality
      and setting NULL is ok according to BDB docs
    - db->set_alloc() can't be used if environment is used, fix the condition

commit 5722f2d65cc4f1a1204f71edef311ce30cadda5c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 17 11:36:10 2009 +0300

    Add runtime settable custom memory allocation failure callback routine
    - lets API users perform theirn own cleanup / exit through their own
      routines in case of allocation failure, or even try to free up
      some memory and retry allocation

commit 14925811396c7f006b9b305cdce297366c648295
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 16 17:28:22 2009 +0300

    Always use build + use our own allocator functions
    - Consistent across platforms and will allow some things macros dont and
      memory checkers these days are smart enough to get decent output anyway
    - Minimal namespacing with r-prefix, add compatibility macros to system.h
      for now so we dont have to change the entire codebase for this
    - Also make rpmutil.h where the declarations and gcc __attribute__ macros
      are available everywhere

commit 3ef73d4d9df6666e5fa3ad772cb4c177f0fd3440
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 16 16:18:06 2009 +0300

    Eliminate unnecessary variable

commit 23dd010c57a3e25938a60984186fc3b2d1fc15ea
Author: Florian Festi <ffesti@redhat.com>
Date:   Wed Sep 16 14:59:48 2009 +0200

    Remove unused local variable

commit f3d2c4a9f92ffc399cf8ed5bebdac1531081ffd9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 16 15:36:44 2009 +0300

    Remove db environment on close on chrooted operations
    - As we open the db from outside the chroot, the environment ends up
      containing paths that are not valid once we enter the chroot causing
      dumb issues like RhBug:513699 and RhBug:507309.
    - We'd be better off removing the environment always after access but
      limiting to chroots for now as the rpmdb open/close path is full of races
      when environment is not present. Chroots are somewhat special environemnts
      anyway and typically not concurrently accessed so it's less of an issue
      there.
    - While this still has all sorts of issues, it at least leaves the rpmdb
      in chroot in a functional state after initial install.

commit c339c3614484642b178716b0ccca88bb6ad3c20c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 16 15:12:28 2009 +0300

    Move the default transaction lock out of BDB environment namespace
    - We dont want the transaction lock to get mixed with the BDB environment,
      nor do we want it wiped out with it as the ts lock protects things
      *around* the environment. Unlike the environment, any fcntl locks on
      the file get released when process dies, so if there is a lock,
      there's a process alive holding it and you'd better not remove it.
    - Use the same lock path in macro configuration and built-in fallback
    - Make it a hidden file to avoid people confusing lock file existence
      with it actually being locked

commit 8b9d37302588d07e1a66f66c9b5dc84625aae061
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 16 14:08:28 2009 +0300

    There's a function to test string equivalence...

commit 6b11f58c764d2ec5edd13987bbb61d2a0c96c7ed
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 16 14:00:54 2009 +0300

    Throw out useless db statistics structs
    - besides being completely unused, these structs are defined by BDB
      public headers, no need to mirror them in rpm

commit ed8a4e20e55659593b2488df00143ff50c5829ea
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 16 13:57:01 2009 +0300

    Throw out pile of unused "stuff" from 2002

commit 4cfcdd661d398691163f27c0ad9744b33d2dfae9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 16 12:41:44 2009 +0300

    Minor cleanup to newRpmdb()
    - no need to free what hasn't been allocated yet

commit 90a0f752808d166e7d1640d60cd3a7d4cd26154e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 16 12:27:12 2009 +0300

    One complicated way to write xstrdup() .. pff
    - sqlite backend uses dbi subfile for table name, leave it alone although
      it is the same as the file name...
    - allocate + free both file and table name separately

commit 961ced7a611000b8a3d522ddc92c6230b4367e1f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 16 12:08:58 2009 +0300

    Eliminate long since unused db subfile from bdb code

commit 2bfa9ca1470d6616a4fe282a18ae0be7be097854
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 16 11:56:12 2009 +0300

    Remove unnecessary arguments to db init and fini
    - bdb environments are per db_home, not per file so dbfile etc dont make
      sense here

commit cc1a7acef9aef29218f1671760be0ac06499268b
Author: Florian Festi <ffesti@redhat.com>
Date:   Mon Sep 14 16:19:33 2009 +0200

    Also skip to be erased files in "netsharedpath"
    fixes rhbz#494640

commit 63d37be6b4a854b9cbbd47306c5c383a2ad356c2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 15 12:57:51 2009 +0300

    Initialize rpm earlier in rpm2cpio (RhBug:523260)
    - url retrieval requires macros to be loaded

commit eeceb428951fd33b9bedbd4adba9f08e46a47dcc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 15 11:05:27 2009 +0300

    Take non-gi errors into account too (SuseBug:527191)
    - patch from OpenSUSE / Michael Schroeder

commit 4018bbd08e84fd0cc227dcb8d2dc20693ce2296a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 15 10:56:58 2009 +0300

    Allow absolute paths in file lists again (SuseBug:535594, RhBug:521760)
    - patch from OpenSUSE / Michael Schroeder
    - build-time generated file lists should be placed in the build directory,
      but at least one valid use case for this is things like %files -f %{SOURCE10}

commit b69c5cf90d74738214fa46c94f223ac9800487de
Author: Roland McGrath <roland@redhat.com>
Date:   Wed Sep 9 11:07:10 2009 -0700

    find-debuginfo.sh: use permissions 444 for all .debug files
    
    Signed-off-by: Roland McGrath <roland@redhat.com>

commit d846e6b0f829791e6ffce4fd496c9378bff19c43
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 15 08:45:14 2009 +0300

    Don't add dependency type into file requires and provides (RhBug:523282)
    - unintended change in commit 3d8656f040d2c780c88d0cd63594898f0605a6fa

commit 7cd005ce7f5357e4905c920f0ddb576f055dae82
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 14 13:23:34 2009 +0300

    Clean up rpmalAllFileSatisfiesDepend() a bit
    - avoid multiple strrchr() calls
    - avoid _constfree(), the string size is known from strrchr() so we dont
      need to allocate on heap at all

commit cb2bd766ded51790353ae667aedbfeee0e9494f9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 14 09:48:49 2009 +0300

    Single point of exit for doSetupMacro() to clean up freeing allocations

commit 154462928eb6c391e2505ce7bc7354b1eb215f3c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 10 17:16:12 2009 +0300

    Oops, accessing the wrong header in processBinaryFiles()
    - dumb mistakes from commit 4177fdffd6e7414123062e8716afb22596608531

commit cdb2cd545b7ba35fb641f165f2d1269d40fad690
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 10 15:03:13 2009 +0300

    Rip the broken asynchronous psm machinery
    - it was only "used" for the ldconfig optimization and actually been
      completely broken since the ordering rewrite

commit 0b2858d13e0af704c7ace8b95751cc2dd60acd7c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 10 15:02:36 2009 +0300

    Rip the broken ldconfig optimization hack (RhBug:513224)
    - it doesn't optimize much and what's worse, it sometimes optimizes out
      calls that would've been necessary

commit 5327ac46e53205bf5f5adc83d2487e77d6e44851
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 10 14:14:49 2009 +0300

    Rip unnecessary headerDump() and the last ancient artifacts using it
    - xml format dump gives an output that can actually be further processed

commit ec8f057f64cc119ed57a1aad9dc72f56da313a07
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 10 14:05:08 2009 +0300

    Rip unused and useless headerGetLangs()
    - nothing in the world uses it and the info is available through
      regular headerGet() anyway

commit c1fb691ef7c5a9550c38ec43da69df1cc94729b5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 10 14:04:18 2009 +0300

    Rip unused and useless headerUsageCount()

commit 451bbb43a86cef57c7eb5a357ab817e9a44e6f74
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 9 14:19:20 2009 +0300

    Eliminate unused goo from findEntry()

commit fc976d13b8ba5ee3cd44c5d32804c5a1c492494b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 9 12:17:10 2009 +0300

    Handle NULL header in doHeaderUnload()
    - avoids having to separately check elsewhere and avoids segfaulting
      on headerUnload() and headerReload() on NULL headers

commit cec00141387db0435a9d60912b7bd3482596eb7a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 9 11:52:47 2009 +0300

    Eliminate pointless increment

commit 1bc60ab06ed761182c8c530c0e44bc46cf403f61
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 9 11:47:59 2009 +0300

    Eliminate yet more unused calculations from header code...

commit 43dfc9e8d01ad840a0420bcbf037244dfd300caa
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 9 11:42:18 2009 +0300

    Eliminate unused pad calculations from header code

commit 8dd2babb562aff4db947becdeeba3fd2fa1cca12
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 9 11:38:28 2009 +0300

    Eliminate yet another unused variable in header code

commit 37f739c251055c87fbafead10b05fcba33145d7c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 9 11:10:41 2009 +0300

    Eliminate last remaining leftovers from "8/98 bug" workaround
    - the ieprev calculations aren't used for anything, should've been in
      commit 4d0d54492c7d539cc806ca9e7fa59e919c28e45d really

commit 7e7c235b3438e0d147229eb14bce6f305f46bff1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 9 11:06:45 2009 +0300

    Eliminate unnecessary variable & related goo from regionSwab()

commit a2ba2b6ce0b1f924546619eb99ab60f758c07837
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 8 14:11:59 2009 +0300

    Initialize diff on declaration where it fits comfortably on one line

commit 5a13d2ed21da68b4d9f16c1266ab1250f646c9e0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 8 14:10:34 2009 +0300

    Move variable to block local scope, not needed elsewhere

commit f3389c8794f99685eb9bd113b48dd844a93c57cc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 8 14:07:52 2009 +0300

    Remove unnecessary assignment, diff is block local scope here

commit 1b617fd68b6250130bf8fb357df87a09c6691e59
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 8 14:05:16 2009 +0300

    Reduce the number of exit points in parseFormat() to error and ok cases
    - avoid a dozen redundant freeFormat() calls and dont bother NULLing
      local variable at the end

commit 9f9ab48f719e9873f45e9c08d9f13030e98d057f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 7 18:41:32 2009 +0300

    Add -Wempty-body to default options if gcc supports it
    - ...to help protect against screwups like
      2ee45ef12aaa958b78c56b6d83bb18aa6a477d15

commit f634253b323933e0383ca9097a77fc0731ad766e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 7 17:50:43 2009 +0300

    Eliminate unused variable

commit c50851aa990aab57f9906d29555fb6443571a318
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 7 17:39:39 2009 +0300

    Eliminate unused variable

commit 0232226ab1a5be31a0fdb5e0cdeb65e2430404df
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 7 17:10:12 2009 +0300

    Minor rpmfcGenerateDependsHelper() cleanups
    - initialize at declaration time, remove unnecessary variables etc

commit ec65f5e5f0a56b398878f5087685b71942db362a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 7 17:00:49 2009 +0300

    Move rl to the scope where its needed, dont bother assigning NULL

commit caa6dc91acd958928eeaf440ade7f01311bcf330
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 7 16:59:01 2009 +0300

    Single point of exit for doPatch() to clean up freeing

commit 23307f3823bc0457cabb19c4bf9bfac0cdde84e8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 7 16:39:40 2009 +0300

    Eliminate bogus variable initialization
    - i isn't used outside the loop at the end, move it there...

commit ef0adf1da10146fe5ba166feeba11961478af819
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 7 16:34:20 2009 +0300

    Add noreturn annotations to javadeps

commit 6fb96f3847bed4c2aba52497eb26b0d2d52207f9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 7 16:23:32 2009 +0300

    Help compilers/analyzers a bit with luaL_error()
    - luaL_error() never returns but as other tools dont know this it
      raises false alarms

commit 475ba2111f5c3e0eb755a954785e09d3a35dfabf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 7 15:52:36 2009 +0300

    Add noreturn annotations to argerror() helper functions

commit 59f791f591059ba44dcbbfe50fa20390e8565781
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 7 15:07:12 2009 +0300

    Add noreturn annotation to die() .. which itself should die

commit 2bdd1a5f5c4b23612a7d87992597b1cd39db86c5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 7 15:04:01 2009 +0300

    Avoid theoretical NULL pointer dereference
    - if sbp ends up not set it's an "internal error" really, dont bother
      with error message

commit d510dfa538ed04b27bd3dc2fa75596f15bd4bee7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 7 14:54:45 2009 +0300

    Avoid couple of (theoretical) NULL pointer dereferences

commit 2b6b06eac19409c5835e8498121edbe38f66a7e0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 7 14:42:04 2009 +0300

    Remove bogus non-null check
    - rpmdsInit() only returns NULL if passed NULL, which can't happen
      as ods is already verified to be non-NULL

commit e3d00e0b26c0d04f7b00be58effc36300485a88b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 7 14:15:56 2009 +0300

    Ugh, erroneously introduced ; busting the conditional
    - broken long time ago in commit 2ee45ef12aaa958b78c56b6d83bb18aa6a477d15

commit 429132522cbf52d4a062227eecc388854e253b94
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 7 14:03:47 2009 +0300

    Avoid potentially calling a NULL function pointer

commit 712ee07e6155d12736989876eb8078355065b522
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 7 13:19:23 2009 +0300

    Eliminate unused calculations

commit e16325d503cc7ba50f11fdebd77f9bae3bff0c8d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 7 13:17:47 2009 +0300

    Eliminate unused variable

commit bd1507497a2601b9a054da567d7c4c82cf39008c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 7 12:23:21 2009 +0300

    Eliminate bunch of unused variables

commit 95d9c3c21e4d84cbd98a195f5ab7579e32e58d74
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 7 12:17:07 2009 +0300

    Eliminate unused variable and pointless null-check

commit 54df179b10d4277a182f016f2c3faa8572209362
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 7 12:05:10 2009 +0300

    Fix dumb some dumb leaks in python spec code

commit 25d5968063683cb4f1059b23a83561e8dd399f9b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 7 11:49:25 2009 +0300

    Eliminate unused counter variable

commit ba3a270fb5f830b75031652018cc7b4ffe26ed38
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 7 11:44:44 2009 +0300

    Remove unused initialization

commit 90c60cc5bbf928824ab3a26b172e9432b8c93ef9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 7 11:42:28 2009 +0300

    Remove redundant reinitialization, cleanup

commit c51a167cb758aa378e27a9065b1641f758b11cac
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 7 11:16:16 2009 +0300

    Add RPM_GNUC_NONNULL macro for portable use of gcc nonnull attribute
    - gcc itself can only warn on literal NULL's passed where nonnull is
      expected but static analyzers can make better use of the info

commit d8cffcff15cb74d4b2db124711a16afdfc079500
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Sep 4 13:38:58 2009 +0300

    Fix memleak in headerGetDumber() .. err .. Number()

commit eabfd39c3b2fa3813d30bd9f1ba716c1fc88854f
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu Sep 3 07:57:34 2009 +0200

    Handle .desktop files with spaces in filename (rhbz#520920)
    - thanks to Edward Sheldrake

commit 936e2570195d076ba36aa2bb93f04f395033945d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 2 15:34:35 2009 +0300

    Use %{nvra} as default query format, much simpler...

commit a912ca599ce959dd15dc33378b1686dfc7c3aca8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 2 14:43:00 2009 +0300

    Fix dumb thinko from commit 4177fdffd6e7414123062e8716afb22596608531

commit 8c04f809e5763475ae3f778fe2a195a4417013dc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 2 14:08:57 2009 +0300

    Take advantage of headerGetNumber() to get buildtime..

commit 80d61d995f26fc17a901526ba56a465d0b69b588
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 2 13:53:27 2009 +0300

    Further simplify rpmVersionCompare()
    - headerGetNumber() side-effect of returning zero on errors fits this
      job perfectly, simplifying things somewhat

commit 9c7e4c70b9aaced9119150f7d60a1be17e4320ea
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 2 13:47:02 2009 +0300

    Deprecate headerGetColor(), implement using headerGetNumber()

commit e6b01b8f70b6bda85e83a5d50321e4745bad2b36
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 2 13:46:23 2009 +0300

    Eliminate headerGetColor() uses everywhere

commit 3661e1d525f71c78b4a866380a934759ad215237
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 2 13:57:54 2009 +0300

    Add header color tag extension, ie headerGetColor()
    - use common numeric tag generation helper for header color
      and dbinstance tags

commit 1bf0d80281f93bf84155cd9a1c741f5e3e6f8b37
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 2 13:27:18 2009 +0300

    Add convenience function for retrieving numeric header data

commit b949efaec09a40d71ef889c09e96e4ff31209174
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 2 13:24:59 2009 +0300

    Oops, dont leak memory on invalid types to headerGet[As]String()

commit 4149dfa7fedf79b28293a2b62400094de646d95d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 2 13:06:25 2009 +0300

    Take advantage of new header string getters

commit e725cbe155f09c01ad2bc1dda64911ed81db6538
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 2 13:00:59 2009 +0300

    Clean up rpmtsAddInstallElement() a bit
    - only bother retrieving os, arch and colors if actually needed
    - obsoletes are never colored, remove the related cruft

commit edc5d15db4037b981021f5fac92774b291b800a6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 2 12:43:24 2009 +0300

    Take advantage of new header string getters

commit 1d1b852ac9d9c327c6705cea7a7eead937d02e92
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 2 12:39:06 2009 +0300

    Take advantage of new header string getters

commit 5deb850f37fa4092bee51a48719a237d36842b95
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 2 12:38:20 2009 +0300

    Take advantage of new header string getters

commit 02a4806b4783ea5665a3010aca0ae05b9cf48024
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 2 12:34:21 2009 +0300

    Simplify rpmpsmStage() and srpm install a bit with the new getters

commit e5c02700ae97315da52b08f50c1ba5f51a7d82c6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 2 12:24:01 2009 +0300

    Simplify rpmVersionCompare() a bit

commit 726c6be41ae0118c25f1606fe3daecab68fe34ef
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 2 12:19:15 2009 +0300

    Simplify headerCheckPayloadFormat() a bit

commit e0a4949bd41819431022fcd364817edd02cb0bef
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 2 12:08:33 2009 +0300

    Deprecate headerNVR() and friends
    - reimplement using the new generic getters to simplify
    - mark as deprecated, no uses inside rpm remain

commit b5bc388ee38de77691b63aa7f48b4928fe60c4e5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 2 12:06:34 2009 +0300

    Eliminate uses of headerNVR() and friends in librpm (and python) code

commit 4177fdffd6e7414123062e8716afb22596608531
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 2 11:55:42 2009 +0300

    Eliminate uses of headerNVR() and friends in librpmbuild code

commit a5a9208b696860982faad4881eeec8aa0eb19648
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 2 10:24:20 2009 +0300

    Add tag extensions for common name-[epoch:]version-release[.arch] combos
    - RPMTAG_NVRA same as rpm5.org, others are new tags

commit e2239d295e6163d4d067bc3a9391997fa0e27367
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 1 17:10:18 2009 +0300

    Add two new convenience functions for header string data
    - headerGetString() for retrieving const strings directly from header memory
    - headerGetAsString() for converting any non-array tag to string

commit 0fd95da2e7105f014475843f25f7e7270f4012d6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 1 15:08:58 2009 +0300

    Add dbinstance tag extension (port from rpm5.org)

commit 7c15b12f02c0fb71f5058bdcc585a766fbf5945f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 1 13:31:20 2009 +0300

    Group tag inheritance doesn't need special treatment, just copy with others

commit a070010b7c27275dfec0e178c79a34b0ef109fa4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 1 11:38:06 2009 +0300

    Triggerprein has been implemented since January 2008, update comment

commit 75c0d816a8185177f7d69b68523a4e45b70e46f1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 1 10:52:12 2009 +0300

    Add provide-extractors for fonts and desktop files
    - extract mime handler capabilities from application .desktop files
    - extract font family names and supported languages from font files
    - both are used by PackageKit for on-demand installing of fonts and
      applications, but are useful for other things too

commit 457606983f6053eb3c053ea564937f184d2b9154
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 1 10:40:50 2009 +0300

    Further file-classifier tweaks
    - disable "known tokens" lookups in ascii files, this is way too unreliable
      to do anything but cause pain
    - consider text files as RPMFC_TEXT consistently
    - font metrics files are not fonts

commit c3770c9404ffa2fcfbe0869dd7fbdb73466a637a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 1 10:31:44 2009 +0300

    Make internal file classifier look into compressed files
    - prerequisite for various things like font provide extraction
    - at least libmagic from file-5.00 leaked fd's with MAGIC_COMPRESS,
      this has been fixed as of file-5.03

commit 1eb9ae13f66435060f7f1a6d27fea6e6965782d2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Aug 31 13:00:18 2009 +0300

    Replace equal/not equal uses of str[n]cmp() with rstreq[n] in python code

commit 555af4eadb5ecb16e101b35cb6e931d3ff502d22
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Aug 31 12:55:33 2009 +0300

    Replace equal/not equal uses of str[n]cmp() with rstreq[n] in binaries

commit 0e8dd93796a8fdafcd0328f131c8f78f1b95a3cf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Aug 31 13:03:53 2009 +0300

    Replace equal/not equal uses of str[n]cmp() with rstreq[n] in main ts code

commit 9ff1cf55f4214f582808a2ad91d68332e0c7a87c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Aug 31 12:46:59 2009 +0300

    Use rpmFileHasSuffix() instead of doing the same manually

commit 746070ebe15022b482d6b3a5d54499ba09c5aa00
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Aug 31 12:43:02 2009 +0300

    Replace equal/not equal uses of str[n]cmp() with rstreq[n] in header code

commit 4ef65c9a4218f9b354b00708ee2126fb1b7bec74
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Aug 31 12:29:56 2009 +0300

    Replace equal/not equal uses of str[n]cmp() with rstreq[n] in misc helpers

commit 1cd1eba2ab619afd17228e326dda0a1b1b4c9a2f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Aug 31 12:22:53 2009 +0300

    Replace equal/not equal uses of str[n]cmp() with rstreq[n] in fsm & psm

commit 6585c6485ab318c836d16edc069bc1be01e0e425
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Aug 31 12:15:16 2009 +0300

    sameProblem() return value is illogical, reverse it

commit feece1341e32efeafa1fbdeb893e33629147603e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Aug 31 11:30:53 2009 +0300

    Replace equal/not equal uses of str[n]cmp() with rstreq[n] in rpmrc code

commit ecfe9a04f83999c75fcbc8b6d2d9c38e266765cc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Aug 31 11:22:47 2009 +0300

    Replace equal/not equal uses of str[n]cmp() with rstreq[n] in backend code

commit 36f8bc0f8ebc6029e8edb9ab537dc75e4e0091d1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Aug 31 11:20:40 2009 +0300

    Ugh, fix brokenness from previous commit

commit f4b39519d32287b9f0ed8dc9ad93787ca6a63fcd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Aug 31 11:15:16 2009 +0300

    Replace equal/not equal uses of str[n]cmp() with rstreq[n] in io code

commit 0c4e48de7908823e19945887d5e7b6ee7fa6bedd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Aug 31 11:08:05 2009 +0300

    Replace equal/not equal uses of str[n]cmp() with rstreq[n] in build code

commit 7e841570fc4950724fb1be93512632b3fee73abd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Aug 31 10:44:35 2009 +0300

    Hum, rename rstrneq() to rstreqn() to avoid confusing neq with "not equal"

commit 8e9662bd4cedd1fc9e31ae18bc420bc2b24c7be1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Aug 31 10:19:45 2009 +0300

    Eliminate now unnecessary compressor -> ioflags conversions

commit 5ca7197e33b0e02c42abc13ad953e5fa958f01f8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Aug 31 09:44:52 2009 +0300

    Accept direct compressor names as Fopen() modes too

commit f9de5ed43058c207e70c86cdd8fa34cbfeb545cd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Aug 31 09:39:16 2009 +0300

    Switch rpmio mode checking to use rstreq()

commit a5de0df0cd48f2ad3d55657d1c79c9e88dafa965
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Aug 31 09:33:52 2009 +0300

    Add two helper functions for comparing string equality

commit 8f10f2362ec5820037591d01349e502868f89eb3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Aug 28 15:43:07 2009 +0300

    Eliminate unused result code from psm struct

commit 41eb3ac5817ca0a6672a528bdc083b632c11d0bd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Aug 28 15:36:07 2009 +0300

    Eliminate payload file handle from psm struct
    - it's only needed locally inside PSM_PROCESS on PSM_INSTALL

commit 96987afd73a68fd76b7135ea6b3a42733ecd2aae
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Aug 28 10:52:49 2009 +0300

    Don't try to incref NULL ds

commit a40e3556765b4d30642601c89319db4b03780323
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Aug 20 14:10:41 2009 +0300

    Remove extra trailing } copy-paste error from python_foo macros, duh

commit 3ae20cd53dfdffbdc98cedc7a2d6fa5ccadb9ccd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Aug 20 13:04:21 2009 +0300

    We always add EVR and Flags to dependencies, remove bogus condition

commit 10772ac7dfad60cb5a20681d22cf851468d0a8f9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Aug 20 13:02:05 2009 +0300

    Fix duplicate dependency checking on build
    - Broken by commit af8b41c64af39ce07d85fcd92fa78d566747d815 which
      simplified too much.
    - There's no guarantee that rpmdsNew() returns a sorted dependency set
      so rpmdsFind() doesn't work correctly here. Walk the ds manually instead.
    - With multiple triggers of same type, identical trigger conditions on
      different trigger script were seen as duplicates (RhBug:490378)
    - Split the duplicate checking to separate helper function for clarity

commit 634a67d055444593a47a21e9e9ea00df65e63d63
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Aug 18 11:46:12 2009 +0300

    Copy BUGURL to source rpm's too
    - should've been in commit 86032bfbb61f1bdee9ac42989a14ea40bed0deec, duh

commit 3e5097c97541fa0b8f289ef3b6011bdc3b4dc002
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Aug 18 09:20:45 2009 +0300

    Add some commonly used python macros (ticket #83)
    - macros to determine python version and library paths

commit 4ced96c5f2469786f5d8f9b14583f191b15a302f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Aug 17 15:06:39 2009 +0300

    Eliminate unused code (leftover from ordering rewrite)

commit 80a8a5b518c218287f2f2fd3d7d3b071efbb8ca1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Aug 17 14:40:39 2009 +0300

    Don't bother relocating packages no files
    - fixes a small memleak

commit d8cde273e5523364df4f4f9ae7ad3374cf02e549
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Aug 17 12:07:48 2009 +0300

    Avoid adding Lua sources and patches twice when recursing (ticket #82)
    - parseSpec() operates on newly created empty spec structures even when
      "recursing" but the lua variables are global and need recursion
      protection to avoid adding multiple times

commit 3fc58248d23d6f720942e5cbf4f92db246a802f0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Aug 17 11:22:16 2009 +0300

    Add default %clean section unless overridden in spec (ticket #81)
    - avoids silly clutter in every specfile in the regular case but permits
      overriding if needed for whatever reason
    - it's a spec behavior change but mostly backwards compatible:
      existing specs will do whatever the %clean does, and if you need to avoid
      calling %clean for whatever reason just keep an empty %clean in the spec

commit 859dd6d4d248a8277aad1bc72d1ab6b6fefa8e0d
Author: Ville Skyttä <ville.skytta@iki.fi>
Date:   Tue Jun 16 22:25:02 2009 +0300

    Add -d option to %patch.

commit 511d41abc5089884329a681935976c459b141af9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Aug 14 11:30:07 2009 +0300

    Add isSource() method to python header object (ticket #32)

commit 883253ea6af71f8063d7a045841c35bad22147e2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Aug 14 11:27:57 2009 +0300

    Add %make_install macro that does the "right thing" wrt modern autotools
    - old %makeinstall for broken/ancient autotools left untouched for
      backwards compatibility

commit 7f7188a3962b676dfa8b8996c15ff8f48ff5694c
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Fri Aug 14 09:36:26 2009 +0200

    Add bug URL macro so that it could be set distro-wide (rhbz#512774)

commit 3f29577dd441c4e7383c8fadad1d5b12a5c13729
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Aug 13 17:22:53 2009 +0300

    Update querytags test to match current output
    - should've been in commit 86032bfbb61f1bdee9ac42989a14ea40bed0deec, duh

commit 1d61f3e34b677c0b1f6d9f5de8f6996730a0261f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Aug 13 17:11:47 2009 +0300

    Build script environment tweaks to improve reproducability
    - unset CDPATH (RhBug:426955) and DISPLAY environment variables
    - always run in C locale

commit de218d7069294c615dd45cf274dd109ebbdce335
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Aug 13 16:46:18 2009 +0300

    Don't set --target in %configure (RhBug:458648)
    - --target is only ever useful for handful of compiler toolchain packages
      and cross-compiler packages are better off setting it themselves if
      necessary, rpm messing here only gets in the way

commit 91af91ef8e6b4b7314791948abd3ff4247739a44
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Aug 13 16:44:59 2009 +0300

    Nuke long since dead version of %configure macro + update commentary

commit 86032bfbb61f1bdee9ac42989a14ea40bed0deec
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Aug 13 15:45:56 2009 +0300

    Add an option bug reporting URL tag to packages (RhBug:512774)

commit 0466fd7f89d17c0a8190bd9d3d52bc4540c9f47d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Aug 13 11:46:56 2009 +0300

    Add missing <pthread.h> include in rpmsq.h (ticket #87)
    - patch from Michal Čihař

commit db5ac1e44a7c6744abde6b92d2e1c17bd238739a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Aug 13 10:35:01 2009 +0300

    Bump version to appear newer than 4.7.x branch

commit 241581a93ed0591bfae162089d9e61a606201fb4
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Aug 12 16:06:08 2009 +0200

    rpmbuild won't segfault when %including spec with zero size

commit 3b7d3b0c19e2ea4f402bba8f014920eabe0415e0
Author: Ville Skyttä <ville.skytta@iki.fi>
Date:   Tue Aug 4 00:34:31 2009 +0300

    Add *.xz and *.lzma recompress support to brp-compress.

commit f6982a403b284012abc846d9f4e5155ad84fac35
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu Aug 6 14:16:02 2009 +0200

    Correctly inherit default %attr(-,-,-) attributes from %defattr (rhbz#515685)

commit 69702c1de3e24bb44f741fbad6e961213f8db9e4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jul 23 12:24:10 2009 +0300

    Kill off lclint remnants everywhere
    - remove ancient and likely irrelevant LCL comments and bogus NULL checks

commit 52e4b9bcaca60499e8bb7f23eb590ce01c89c574
Author: Lubomir Rintel <lkundrak@v3.sk>
Date:   Tue Jun 23 00:49:17 2009 +0200

    Fix memory allocation for token array
    
    This fixes a memory corruption due to write access out of
    bounds of token array, whose size was computed incorrectly.
    It was assumed that only '%' characters separate tokens,
    which could lead to crashes on useless uses of '[' tokens,
    such as "rpm -qa --qf '[]lalala'".

commit d8071161f98ff703846281466cccf880e9dcb5c5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 14 14:25:58 2009 +0300

    Log an error on short lead (RhBug:511101)
    - if lead read returns short without it being an IO error, treat it
      as a case of "not an rpm package"

commit 7baf3235d2263d456fd11b6fa3ad605d0d1d1136
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 14 13:56:27 2009 +0300

    Honor --nofiledigest on install too (RhBug:508021)
    - query and install options clash, use popt callback to enable processing

commit 16be29746ce4ecdedb2034c8d97b6455ba5ae620
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 14 11:54:43 2009 +0300

    Document --nodeps in rpmbuild manual (RhBug:482921)

commit b8102d8f6150b0438ba91cd53ff86a5391e8bdcf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 14 11:24:02 2009 +0300

    Dont accept '-' in manifests (RhBug:461353)

commit 0114bb51dd85dc4a875d5b9f8f9775a05ecd4187
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 14 11:15:30 2009 +0300

    Dont treat filenames starting with - as stdin

commit 666e26bd7a229063e6cab068c94b56e1b35e7ddb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jul 9 15:07:38 2009 +0300

    Oops, missing include for dirname()

commit 17f9073278206a979829fa41ac496476dde1a3ef
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jul 9 14:40:17 2009 +0300

    Create the directory used for transaction lock if necessary
    - otherwise at least rpm --initdb can fail due to transaction locking failing

commit bcb71f69d2524ee51535c50b51ae6b4aa133abc8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jul 9 14:01:13 2009 +0300

    Fix package relocatability checking, broken since forever
    - at the time relocation problems are added, ts problem set is NULL
      and rpmtsCheck() wipes it out anyway so the problems from relocation
      are never seen anywhere
    - The header isn't available at transaction prepare stage where most
      problems are checked for, so we can't validate the prefixes at that
      point. Add a per-element problem set to carry the problem data and
      merge it to ts problem set when checking for other problem types.

commit 7c512db91281a53f1dc783732e310f311382ae8c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jul 9 13:43:10 2009 +0300

    Make rpmpsUnlink() behave the other fooUnlink() functions
    - pretty much everything else in rpm returns NULL from unreference functions,
      make be more consistent here...
    - also tolerate NULL in both reference and unreference

commit 6ffeec9d7f57ebb520d15581d55555244c0f3058
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jul 9 13:39:55 2009 +0300

    Add reference counting to rpmProblems

commit cb8241dda783f7e8c143b08fecf57fe89a39c3a6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jul 9 11:03:54 2009 +0300

    Unbreak --excludepath (RhBug:409637)
    - excluded paths aren't counted as actual relocations, dont bother
      checkig addPrefixes() return

commit 2db564a85a57db624cc73dd0bb7b2875cff44f3d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jul 9 09:28:45 2009 +0300

    Fix compiler warning on const free
    - we dont need popt to set the file variable as we manually process

commit 6649c23cac5823d0c7415df5cfd66e5aac5eebd2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jul 8 16:20:06 2009 +0300

    Make verify result not depend on verbosity level, uff... (RhBug:510213)

commit 33d829d362ef7d529268b99e2f943d2ecda07d59
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jul 3 15:12:07 2009 +0300

    Add an index for obsoletes (RhBug:507702)

commit 683d30858610c548ae3a3999b458b438d7275546
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jul 3 14:46:06 2009 +0300

    Unify extension + header data getter calling conventions
    - simplifies headerGet() & friends a bit and reduces redundancy
    - also just return an error instead of dying with assert on NULL td argument

commit e18aedbf97d0948b9010877270d9559d8a64c861
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jul 3 14:21:01 2009 +0300

    Pass headerGet flags to extensions too (unused for now)

commit 4d0d54492c7d539cc806ca9e7fa59e919c28e45d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jul 3 13:11:35 2009 +0300

    Eliminate unused calculations originating from the "8/98 bug"
    - the tl calculation has been unused since commit
      5994a308bc062fff53ff0184b8dea15d19fcfb63 in 2001 and we dont even
      read rpm 2.x packages anyway...

commit 5dfb57ba3c65b4b3a383361950eec31865ebf75c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jul 3 12:13:54 2009 +0300

    Re-enable rdlen vs dl header check
    - we dont even read packages from the "8/98 bug" era anymore, hardly relevant

commit c685667475d70daabac9ae224979abbd1f86987a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jul 1 13:59:53 2009 +0300

    Band-aid to dependency caching misbehaving with chrooted verify (RhBug:508074)
    - disable dependency caching on chrooted verify to avoid ugly spew
      from paths containing outside paths while we're actually inside the
      root during verification
    - correct fix would be to fix the temporary db path calculation

commit d3956140c2757a5929d4517cf54e5d9c06e11355
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 26 16:24:55 2009 +0300

    Handle hook unregistering itself

commit d1b399761ffe554f3f3c4ced187c3b8ec58233fb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 26 16:07:50 2009 +0300

    Fix rpm.unregister() Lua extension
    - hook data is userdata, not light userdata

commit 02f9f3fbdd5e0cc858ab46da9a11e04110f92462
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 26 15:39:20 2009 +0300

    Add base64 encode and decode extensions to the lua interface

commit edbc9ead961fcbeb1733b47405041b653b521bcb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 26 12:31:57 2009 +0300

    Base64-encode %policy files to ensure it can be presented as strings
    - RPMTAG_POLICIES is a string array but there's no guarantee that
      something marked %policy is a plaintext file that can be represented
      as \0-terminated string, base64-encoding them fixes that. Baby steps
      towards making %policy remotely usable, related to RhBug:505066.
    - Also remove unnecessary failure code setting, processMetadataFile()
      assumes failure already, and dont try to insert NULL strings in case
      b64encode() or pgpArmorWrap() fails

commit 8c5437b37914bd098a9f13366e53aba5a1c9cf15
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 25 17:17:16 2009 +0300

    Kick out silly constfree() use

commit 271069f90046994b005a469c80c2e150a41af9f3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 24 15:16:25 2009 +0300

    Simplify logging a bit
    - precalculate the message length, relying on C99 vsnprintf() semantics
    - generate the log message and rest of the record in rpmlog() already,
      and just pass the record to lower level to do actual logging
    - also something was wrong in the old version as valgrind complained
      about illegal reads, that is also cured here

commit 51520de1d3dffedc0c047b6da3fa68e98c92acc0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 23 17:50:08 2009 +0300

    Set HEADERFLAGS_ALLOCATED centrally from headerLoad()
    - requiring every caller to handle this separately, violating header type
      opaqueness in the process, doesn't seem that bright an idea
    - also fix a memleak on signature header read in case of failure

commit b02893092594b868296c5b3f58ab631e368bae1a
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Jun 23 11:05:06 2009 +0200

    Teach %files to accept multiple filelists (ticket #70, rhbz#475359)

commit 27263f94bd18ea0d86b1b85511ef10a7f573aa92
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Jun 23 10:56:53 2009 +0200

    Allow ':' character in EVR, it's needed for Epoch separation

commit 194b6a5c2ab37753a143cd4d710bd09ccc0502e2
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Jun 23 10:55:54 2009 +0200

    Check arg in argvJoin() before dereferencing

commit 215607caa52112bd8989ee1e6c58205c1881b6a3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 23 10:03:58 2009 +0300

    Updated Polish translation (ticket #74)

commit 0975d1e2c04718e290f868c77613640c044ce4f7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 23 09:56:04 2009 +0300

    Fix some typos in javadeps usage message (ticket #72)

commit 6b6e82f8e9c11aeecbd47038a0c637ff5c8f8a1f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 18 14:52:45 2009 +0300

    Add -h/--help to rpm2cpio (ticket #63)
    - patch originally from Debian, with slight adjustments to indentation

commit 4abfb4c186e1c3ad72c0899048d306a35a43fe8c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 18 14:19:59 2009 +0300

    Updated German translation from Micha Pietsch (RhBug:506607)

commit 31301b579faed5eb58ba648d74b0ad7c5fea66f5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 18 14:16:46 2009 +0300

    Updated German translation from Micha Pietsch (RhBug:506607)

commit 7417cd283a3641ee62b282ae36c5d2d7df349d04
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 18 13:06:42 2009 +0300

    Pile of OSGi dependency generator fixes from Alphonse Van Assche (ticket #39)
    - use Temp perl module to provide temp dir
    - re-enable deps solving in Require-Bundle, Import-Package, Export-Package
      OSGI properties
    - remove uses bundle of Export-Package OSGI property
    - use RPM '>=' as version operator to match OSGI '='
    - remove all .0 at the end of the version string
    - typo fixes

commit d5f6d1854094b659c4f02f86100c426e2279e642
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 17 15:06:06 2009 +0300

    Delay fingerprint cache init + lookup until we know there are some matches
    - no point going to fs if there are no matching files in the rpmdb
    - use allMatches count as size hint to fingerprint cache instead of
      arbitrary 20

commit 71475e314a6c7af0a3dc2cae7e846663f8f7f620
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 17 15:04:41 2009 +0300

    Permit NULL in fpCacheFree()

commit 0055fecfde5404c5106ac0fc58052e9264da8592
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 17 13:09:25 2009 +0300

    Fix invalid memory access from freeing dirName too early
    - if it doesn't crash, it can cause rpmdb provided file not seen in chroot,
      the other half of RhBug:506323
    - streamline exit points to enable freeing allocated resources sanely

commit 1b9e6d7f48a68fd677c70d04c9a6bb4274cf7b9f
Author: Florian Festi <ffesti@redhat.com>
Date:   Tue Jun 16 17:37:40 2009 +0200

    Fix Segfault for to be installed symlinks pointing to existing locations
    fixes rhbz#505777

commit 5f57e212516cc6ea5b9dd8276977d85a07eeec56
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 16 18:19:40 2009 +0300

    Eliminate now unnecessary artifacts of rpmdsNext() side-effects

commit 7fd95d099f232a54d772379fc9769eb3c2453fed
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 16 18:15:31 2009 +0300

    Generate DNEVR string lazily on rpmdsDNEVR() calls
    - avoids useless string churning on rpmdsNext() when most paths dont
      actually use the DNEVR string for anything
    - make sure DNEVR is freed whenever iterator index changes

commit 61acdba023167c73cf0db124f0ae8c09c8f387d6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 16 17:49:44 2009 +0300

    Eliminate epoch promotion debug cruft
    - might've been useful back in the day but...

commit 90b70ebd3460ca39427e37da2115f48df00879d3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 16 17:36:21 2009 +0300

    Determine rpmlib deps from flags instead of comparing strings

commit a4ac4e48ec5897e33f35c0242c4f0c79c2f9ce2c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 16 17:23:17 2009 +0300

    Validate rpmlib dependencies on src.rpm install (RhBug:490613)
    - other src.rpm dependencies are only relevant for building, but rpmlib
      features can affect src.rpm "install" too, such as unsupported file
      digests, payload compressors etc

commit 8f44f9159cb1f522f20b88821bab38e4ee0ef565
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Jun 16 13:53:42 2009 +0200

    Don't check Provides/Requires for bad characters, they could be UTF-8
    - don't check Name but EVR in parseRCPOT()
    - use rpm namespacing for charCheck()
    - remove unused variable

commit 71a7dd8b7a2df677214c18473eac661ea38fb649
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 16 13:52:43 2009 +0300

    Add debugedit support for DWARF-3 (RhBug:505774)
    - patch from Jakub Jelinek

commit 83513ded75c50be90d9b8c4c6c879f67fa1a47bb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 15 15:21:27 2009 +0300

    Rip hardcoded libgcc + glibc static helper kludges
    - shouldn't be needed anyway anymore as package of preferred color should get
      always installed first, and the issue is solvable in packaging anyway

commit ad53979f9ae024083b7f29fe7a914531e487be07
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Jun 15 08:47:55 2009 +0200

    Check for valid chars in Requires/Provides as well

commit 8d79668a137e93b5b15613b6f8769d72df831fea
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 12 08:52:38 2009 +0300

    Fix out of source directory build (ticket #65)

commit a86291ff774ff10277da4452fc47fc9870ccc991
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 12 08:33:29 2009 +0300

    Add parentheses (as kindly suggested by new gcc...)

commit 17d66c47b181d9300dfc3813e8dd86d31c7b99a6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 12 08:29:02 2009 +0300

    Make peace with newer git

commit af0ea37a918690626c9385af5d22f7d007a285dd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 11 17:22:59 2009 +0300

    Make rpmalSatisfiesDepend() smarter
    - Instead of blindly picking the first match, try to pick the best provider
      for the dependency: for colored dependencies, try to find a provider of
      matching color. For non-colored dependencies, prefer providers of
      transaction preferred color.
    - This avoids creating bogus and loopy relations between 32- and 64-bit packages
      where they dont exist, and makes behavior with things like /sbin/ldconfig
      consistent by returning the preferred colored element.

commit 31f0d94df1c50a7a2802c76f491620be22b9a8fa
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 11 17:16:03 2009 +0300

    Add preferred color to rpmal

commit 85f9df3c0d511cb1e84484c8d616b98e5e4a2123
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 10 17:22:03 2009 +0300

    Update translations for a change (due to depends.c -> order.c split)

commit 57884ba0d33482978a2f5f71a9cbad062354850d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 10 17:19:27 2009 +0300

    Split ordering code to separate source file

commit 4e331d1dbb526c0b80d9816ca4e47b21a91f17b7
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Jun 10 11:52:53 2009 +0200

    Add new function to check valid characters in NVR (ticket #59)

commit 53e48bc469eb0dbea0629efa9abadad0d9b4bdea
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Jun 10 09:16:45 2009 +0200

    Nuke useless compatibility stuff and put V/R checking where it belongs

commit 7653f770130626dd42e57395d232696d47e66ca3
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Jun 9 10:29:01 2009 +0200

    Make whitelist of characters allowed in NVR (ticket #59, rhbz#493157)

commit dee5d0e2308092b45af17bbc2a4aff4089b8a56f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 8 14:48:36 2009 +0300

    rpmdsInit() and rpmdsNext() can handle NULL ds, no need to separately test

commit 9fe316637dfef728b5f1e337d2b84955db8376f2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 8 14:43:45 2009 +0300

    Minor rpmtsOrder() cleanup
    - initialize on declaration
    - remove unused code

commit 05f44fd2647def8e08e199a7f5e492389ca66065
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 8 14:24:14 2009 +0300

    Clean up relation recording a bit
    - rpmdsInit() and rpmdsNext() can handle NULL ds correctly, dont bother
      separately checking here
    - te type doesn't change inside rpmdsNext() iteration
    - added vs installed difference is handled inside addRelation(), otherwise
      the cases here are identical except for the al

commit 7c14d1953b8d23d4cd809f896b761b607a77e483
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 8 14:13:58 2009 +0300

    Dont bother adding dependencies to self, ever

commit ace4cbf45e68d5418f4dd377c050b5b470e3b5eb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 8 14:10:44 2009 +0300

    Determine rpmlib and config deps from flags instead of comparing strings

commit 085f75f68d19913cf3ef9650ce2a73b115409489
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 5 15:00:33 2009 +0300

    Map legacy PreReq into pre and preun scriptlet dependencies in ordering
    - this gives close enough semantics for packages that rely on old
      PreReq behavior (eg RHEL 4-5 package set ordering is completely broken
      without such mapping)

commit b22fa73fea6873dfc3816433e130820a5d9f33da
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 5 14:56:49 2009 +0300

    Unbreak isLegacyPreReq()
    - long long time ago the PREREQ bit was also turned on for scriptlet
      dependencies etc, this is no longer the case so the old mask doesn't work
    - build part is not affected by the mask change as we never set
      PREREQ bit by ourselves

commit c33d80aae907726f49d9197e557cb8213bd85c13
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 5 14:50:01 2009 +0300

    Use specific types for sense flags + te type instead of generic int

commit 50fe90eeb59d7b3dc0495bc719c5a84cddd8cbb3
Author: Florian Festi <ffesti@redhat.com>
Date:   Tue Apr 14 10:54:58 2009 +0200

    Rewrite ordering
      - Detect strongly connected components (SCCs) using Tarjan's SCC algorithm
      - Use Dijkstra's algorithm to find the best relations to zap
      - Add forward relations for the Dijkstra's algorithm
      - Separate the per package ordering data and the relations

commit aefe94178709cd4c42fc87e2421106421d7f9bd4
Author: Rakesh Pandit <rakesh@fedoraproject.org>
Date:   Thu May 28 12:21:08 2009 +0530

    Removing extra useless 'break' statement lines and unhandled FTS_FOO cases.

commit 0a48a2e2425fc15f124d3333daf3f5a5a5ece232
Author: Rakesh Pandit <rakesh@fedoraproject.org>
Date:   Thu May 28 12:13:40 2009 +0530

    Removed 'xx' variable in rpminstall.c
    - in setNotifyFlag, rpmShowProgress and rpmInstall functions.

commit cff4eddfd2a619ae75383630442a6f3ac54245f3
Author: Rakesh Pandit <rakesh@fedoraproject.org>
Date:   Thu May 28 12:11:03 2009 +0530

    Removed rpmEIU memeber 'fd' -and its usage in tryReadFoo functions.

commit 2d2aaebce79d0fd48b9fe82f61356f9e8190e915
Author: Rakesh Pandit <rakesh@fedoraproject.org>
Date:   Tue Jun 2 10:43:59 2009 +0530

    Updated mailing list information.

commit 400bc0224d308b84ab0fde6157536ff67ea093da
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 2 09:42:04 2009 +0300

    Remove toplevel TODO too, nobody updates it and we have Trac now

commit 27d1f56f2945c59b39668685090977c6c2635431
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 1 15:32:26 2009 +0300

    depMsg doesn't need _notpre() anymore

commit 4cf62df1081355c5d73078c128f8dbab57d12c64
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 1 14:54:59 2009 +0300

    Remove leftover build/TODO file from 1998 / rpm 2.5 era...

commit 4913f48d08516aa9f369cec50376f2ec52afc1e2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 1 14:48:58 2009 +0300

    Generate spec part string lengths at build time and make it const

commit 317533ce3134d05fc9f30fb29238d3fb2ed4bc6b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 1 14:45:44 2009 +0300

    Generate preamble tag lengths at build time and make it const

commit 7c98cfb220e8265fc4f3530da2cd1cb95859fddc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 1 13:39:00 2009 +0300

    Issue deprecation warnings for PreReq and BuildPreReq

commit c0ed5b0a35d4819782a5f069a70893f19d3ded9f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 1 13:37:20 2009 +0300

    Turn unused spec syntax obsoletion mechanism into deprecation warning

commit 95c9e817ef6ed2cda5629f0c5edb6f7b19c209cf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 1 13:28:50 2009 +0300

    Map PreReq into Requires(pre,preun) at build
    - this gives a more reasonable mapping to former behavior of PreReq
      when cutting loops

commit 9c7bcba82ac3bbf792d6b86bcb45c60405b0db00
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 1 12:53:08 2009 +0300

    Deal with multiple scriptlet dependency bits in deptype format extension
    - rpmbuild permits Requires(post,postun) style notation which gets lumped
      into single dependency, report them all

commit 67bd81fb25a2a57adc97172cf426c475aea41612
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 1 12:42:07 2009 +0300

    Bring back former value of RPMSENSE_PREREQ
    - packages using the legacy PreReq still widely exist distros having
      patched PREREQ back in, we'll want to be able to deal with the legacy
      packages best we can, and to do that we need to be able to test for
      the prereq bit
    - this also fixes ordering brokenness introduced by commit
      502f88e0e67bb1df91e09a32a370b04bda36259e all the way from 2004:
      with RPMSENSE_PREREQ equalling RPMSENSE_ANY, all co-requisites are seen
      as pre-requisites due to isLegacyPrefix() returning bogons (RhBug:502269)
    - previously RPMSENSE_SCRIPT_PRE etc were also marked RPMSENSE_PREREQ to
      permit legacy rpm versions to deal the fine grained scriptlet dependencies
      somehow, this we dont want anymore

commit f243eb7d004394176c64ff24b5a461deb54d866b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 1 10:39:07 2009 +0300

    Avoid filtering out erase pre-requisites on install
    - Requires(pre,preun) etc are recorded as a single dependency entry,
      so filter out dependencies that are only install pre-requisites

commit 03e456de4a0ea7d574a7aeb3aa41937e1b7d18ee
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 29 11:03:33 2009 +0300

    Build prereq is exactly the same as buildrequires

commit cdfd0934841d4eccc26d7da7c35b23e6e9f76a9c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 29 09:02:44 2009 +0300

    Fix calculation of hardlinked files (RhBug:503020)
    - regression from commit 899dfb58927ec6e91014773430824462f4d0002e,
      size of hardlinked file set is the size of one file of the set
    - add isHardLink() internal helper to avoid a copy-paste code

commit 04ec8e9fca9c309367d497feef08e15e03bde120
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 22 13:54:08 2009 +0300

    Couple of documentation typos

commit 7ac37e98579ea0a870ddab0675d9bac59e107857
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 22 10:04:20 2009 +0300

    Clean platform files on uninstall

commit 03e36789b62010026f099e885546be005a62461a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 22 09:56:43 2009 +0300

    Include from src, not build dir

commit 1195233d39d327c88645eee0f7e279fd5b648ce8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 20 17:53:50 2009 +0300

    Remove unnecessary includes + other minor include corrections

commit 44cd276bc2bec0bbb3a177b845b3a02d00732506
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 20 17:22:51 2009 +0300

    Eliminate useless pgpDig nbytes counter
    - nothing at all uses this, dont bother
    - if plaintext size of digests is needed, the correct place to calculate
      it is in the digest routines, not manual beancounting all over the place

commit 9acfa19a0da514f64f17cb4e76603ed48ec8db1d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 20 17:20:43 2009 +0300

    Remove size "signature" verification
    - rpm itself hasn't used this for anything in ages, its broken too for
      quite some time and nobody noticed -> nobody's going to miss this

commit 438d943e11f54a6c8b16a742ee27f48a925cd005
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 20 17:19:46 2009 +0300

    Refactor ts out of main read package function
    - Add a lower level rpmpkgRead() which just takes keyring and verify
      flags, rpmReadPackageFile() is just a wrapper around it now
    - Avoids several levels of unnecessary keyring references through
      rpmtsGetKeyring()
    - Pretty much loses package reading statistics but shrug...

commit ddb263adfccb42cedf762a65b800f2f938506fe8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 20 17:19:06 2009 +0300

    Refactor ts out of main header read function
    - Add a lower level rpmpkgReadHeader() which just takes keyring and
      verify flags, rpmReadHeader() is just a wrapper around it now

commit 072f6fd8805cf1083eabd435689b2e6668511b9e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 20 17:00:12 2009 +0300

    Refactor ts out of main header verification routine
    - signature checking a header only needs a keyring + verify flags,
      add a lower level headerVerify() function (static only for now)
      for this and turn headerCheck() into a wrapper around it
    - some accuracy in statistics lost but shrug...

commit 9322f737819a3d81088699b1d7fa667259245411
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 19 10:26:50 2009 +0300

    Don't mess up problem altNEVR in python ts.check() (rhbz#501068)
    - the use of strrchr() silently casts away the const from the problem
      set altNEVR string, which we then happily modify..
    - similar to commit 62cc76e25cdfad78ac30bb28f626b474efdecddc

commit 2c495bd8c4cb3a28d7f359810d8e0ccacffdc153
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 12 10:30:10 2009 +0300

    Updated Spanish translation (Domingo Becker)

commit 409641c058546b9327afa0fa8417ddb88bdd4f5a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 12 10:28:00 2009 +0300

    Updated Portugese Brazilian translation (Igor Pires Soares)

commit d295d14cd0fbce33ec778ef3b07a95269a4f8bfe
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 8 15:22:25 2009 +0300

    Minor fix to Japanese translation (Masaki Furuta)

commit 5023d0a4e84e2877e3c715b57bd362f95f12ef06
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 8 15:01:15 2009 +0300

    Fix parsing of pubkey signature packets. Ick.
    - busted all the way from commit bcb226480d0b60a8667eade77d81c29de4c3d6c9

commit acd93ee570990ba3f549947dfe67a7d362f7c70d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 7 14:31:57 2009 +0300

    Dont bother generation rpmds debug messages unless in debug mode
    - rpmdsNotify() which does nothing but generate RPMLOG_DEBUG messages
      gets called often enough that avoiding the unnecessary
      rpmlog() and strcmp() calls is easily visible in wallclock times

commit 5b91b31cc98de7750a9b688c38a9e7bae0e1f58e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 6 14:54:10 2009 +0300

    rpmdsThis() is really just a special case of rpmdsSingle()

commit 2ce7b56e621b097b76a2e5059def7d0e5a64d53b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 23 15:54:09 2009 +0300

    Hide NSS peculiarities from API
    - PKCS#11 standard requires modules to be re-initialized after fork(),
      arrange this to occur automatically.
    - Most of the time child processes will exec() something else so dont
      bother shutting down NSS for every child, only lazily re-initialize
      as needed.
    - This lets us initialize NSS early to force preloading of its dlopen()'ed
      libraries to avoid issues on chroot(), without causing problems to API
      users which fork() after initializing rpm (such as func, urpmi etc).

commit 4b523ac1b85b9d4f71cd024306f01c3fb2418c8f
Author: Rakesh Pandit <rakesh@fedoraproject.org>
Date:   Sat Mar 14 18:38:53 2009 +0530

    - Removed unnecessary rpmdbFreeIterator call and moved them to enditer label.
    - Rearranged bit to keep the logic sane.

commit 5bb4b0928393a4534d9f76644f7b3d3c74ceb3f2
Author: Rakesh Pandit <rakesh@fedoraproject.org>
Date:   Sat Mar 14 17:59:03 2009 +0530

    Cleanup: Move similar patterns in rpmcliArgIter to rpmcliArgIterHelper.

commit e77498808e58a8041d3a16450fab400116aca28d
Author: Rakesh Pandit <rakesh@fedoraproject.org>
Date:   Sat Mar 14 17:32:59 2009 +0530

    - Moved code to check possibility of freshen operation to checkFreshenStatus.

commit 7cd942db0b5c09f71caeb9b732f69acb7703f4ef
Author: Rakesh Pandit <rakesh@fedoraproject.org>
Date:   Sat Mar 14 16:38:10 2009 +0530

    - Move header reading part from rpminstall to tryReadHeader function.

commit 0766138ee96fd9c0db193907a0013aba5c3d18bf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 22 17:12:37 2009 +0300

    Minor cleanups to installplatform script
    - eliminate unused target and target_platform variables
    - eliminate unnecessary temporary rpmrc (used to make a difference when
      macro path was in there, not anymore)

commit 4231321f0ccabd3f2b8b9d51f0dc8d39b34dee44
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 22 16:22:54 2009 +0300

    Eliminate unnecessary rpmts_internal.h include

commit 8f755140c31fec7106b86a3aaeb550baf2535fae
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 22 14:58:10 2009 +0300

    Refactor ts out of the main signature checking worker
    - Add lower level rpmpkgVerifySigs() function which does the real work,
      ts and full qva isn't needed there, just keyring and flags what to verify.
      Static only for now, but this is the kind of interface we'll want to
      export eventually.
    - rpmVerifySignatures() is now just a API-compatibility wrapper.
    - Use rpmpkgVerifySigs() instead of rpmVerifySignatures() in rpmcliSign()
      verify loop so we only retrieve keyring once.

commit 7f1f837877173552227a682e3ff63e9cce7b4166
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 22 14:01:40 2009 +0300

    Dont try to refcount NULL keyring

commit acfe923bf21a4402a48b6e357a53a398b107f0b3
Author: Ville Skyttä <ville.skytta@iki.fi>
Date:   Sun Apr 12 21:39:50 2009 +0300

    Fix --with-kde with KDE3 (rhbz#466009).

commit c7523b4c52a4199b4aef31c8456285586bea313d
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Apr 20 13:23:30 2009 +0200

    Require file attributes in spec if %ghost files are missing from build root

commit 8a3e9a74e363f530ed2b7fee0672477f69c3c661
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Apr 18 14:31:48 2009 +0300

    Fix a few file permissions from "make install"
    - macros.perl, logrotate examples etc aren't scripts and shouldn't be
      executable. Ditto for python __init__.py

commit e7ce53c3aea8d11d083e220bfb2e4c6b4ced9a5a
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu Apr 16 14:13:34 2009 +0200

    Introduce changelog trimming in binary packages feature (ticket #47)

commit 3b9f0b817e635b4cf03e500d83bc3e9ba60ff9f9
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu Apr 16 14:12:20 2009 +0200

    Introduce _changelog_trimtime and obsolete _instchangelog

commit 4c39a9df93c962c69e23ff48789cb950a105a44b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 16 11:03:29 2009 +0300

    Pad RSA signatures up to modulus size if necessary (rhbz#494049)
    - unfortunately can't be detected at initial allocation so extra
      copy is needed at verification time

commit fb19a98fa0ca1c152e1bd22322f5fd1350a48eb6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 14 10:42:20 2009 +0300

    Check for termination signals during signing / signature checking loop

commit 1880778844ae92f56f7a329402efd29a24b819df
Author: Florian Festi <ffesti@redhat.com>
Date:   Tue Apr 14 10:57:46 2009 +0200

    Revert erase relations before checking whether we already added it

commit 0f967d157d3a02696898f848f4cd584e646b264f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 13 13:58:39 2009 +0300

    Error out at build on unsupported payload compressors (rhbz#495429)
    - only permit bzip2/lzma/xz payload compression if rpm built with the
      necessary libraries

commit 07524b67b6cfe501bdae3df45c212b7c1c129da4
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Fri Apr 10 11:27:49 2009 +0200

    Don't require %ghost file presence in build root (rhbz#495040)

commit 2566a70b0ac8f18fbcb2c1c060465714d16982a4
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu Apr 9 14:48:02 2009 +0200

    Remove file flags that have never been enabled

commit 0f29cc7f2781cdcedab9c8a03786d45f14fc316b
Author: Ville Skyttä <ville.skytta@iki.fi>
Date:   Mon Apr 6 23:08:15 2009 +0300

    Make it clearer that PACKAGE_FILE logic applies to all -F, -U, and -i.

commit 48580a70bc7372d1227f07f3377f9c253e492385
Author: Ville Skyttä <ville.skytta@iki.fi>
Date:   Mon Apr 6 23:00:56 2009 +0300

    Remove obsolete info about --buildarch and --buildos.

commit 88e63e6460afc431133f8718acd95afd29033d8a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 9 14:06:14 2009 +0300

    Updated Spanish translation (Oscar Bacho)

commit a45e6f03194c1a404bd1924d883981bb4bcf6a7d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 9 14:04:46 2009 +0300

    Updated Serbian translations (Milos Komarcevic)

commit 01d5f668316066daaceb439eebebf8eb569c78b3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 9 13:54:09 2009 +0300

    Updated Catalan translation (Xavier Conde)

commit 7a8aecbc8651cab2bd945db8ffd749f9a165bdf1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 9 13:05:00 2009 +0300

    Fix PGP v4 fingerprint calculation (rhbz#493777)
    - patch from openSUSE / Michael Schroeder

commit 37a35d07211b612429fef6614888f5fa099dd27b
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Apr 6 10:20:47 2009 +0200

    Don't allow '~',  '/' characters and ".." sequence in NVR (rhbz#493157)

commit 26ac9b3d9417fec67b49c4434c205f3232a36c1b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 3 13:07:42 2009 +0300

    Make rpmal header & all internal
    - rpmal is hardly useful outside rpm itself, avoid exporting stuff
      unnecessarily

commit c40f6d5dcabfe0b68b830d96b01eaedac0b2d18d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 3 10:28:59 2009 +0300

    Make sure installed files have state (rhbz#492947)
    - rpmfsSetState() doesn't get called for skipped files like %ghost and
      %config(noreplace), causing incorrect file state ("no state") getting
      recorded in rpmdb, leading to inapproriate removal/rename on erase, ick
    - For TR_ADDED, always default file states to RPMFILE_STATE_NORMAL, fsm
      changes it as necessary for skipped colors and such. Lazy alloc on
      rpmfsSetState() is not correct as rpmfsSetState() might not get called
      at all.
    - originally broken by commit 8d6c4b8c95b59f5a71d90c582c2e98f5c7ed7b9d

commit 4c23ebdb81d8e8a00ccf7e92d7e3788081b2f57d
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu Apr 2 20:49:48 2009 +0200

    Use -mtune instead of -mcpu (rhbz#493696)

commit 0b8ce16f257f82d30cadce5af96350a9d54cdc48
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 1 12:17:45 2009 +0300

    Polish translation update (Piotr Drąg)

commit 7919fdb4804a6dced9cf16ccc408de8286d9987f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 31 13:14:53 2009 +0300

    Update translations for a change...

commit e38cbe5b4e7bfa9c7b10e3926093b6887d97d6a1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 31 13:14:08 2009 +0300

    Prefer more specific types over generic "text" in classification (rhbz#491349)

commit 193e1c21e1db2fc5b1c3390a8ea16abfcf357e99
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 31 13:00:59 2009 +0300

    Allocate ts->addedPackages lazily from rpmtsAddInstallElement()
    - at least verification calls rpmtsClean() which wipes out ts->addedPackages,
      make sure it's always alloced when needed

commit 312b4fcd30684793dac474abf2fa07991fa8be16
Author: Florian Festi <ffesti@redhat.com>
Date:   Mon Mar 30 21:56:09 2009 +0200

    Erase relations are reversed, use requires of target

commit 569adfa17c2839280ee43f05c0c5928e50e5ae10
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Mar 27 16:15:00 2009 +0200

    Tidy up rpmfi struct definition a bit

commit 74a4f639e2c62109ad8e44fb608382241bdb1faa
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Mar 27 16:12:25 2009 +0200

    Lose unused/useless Type and tagN fields from rpmfi struct

commit 3d63df622037796279514a4da3ddc6807a102c1d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Mar 27 14:11:43 2009 +0200

    Handle XZ in %uncompress macro
    - should've been in commit 8078d0ba24662308b10d9eb0f0da978584b7e757

commit a3922760832ec4329f1426a54012b28cf387b5f5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Mar 27 14:08:56 2009 +0200

    Rip the macro default settings in rpmrc setDefaults()
    - these get nuked and not reloaded on package builds
    - rpm(build) needs a properly set up macros file in any case

commit 2d2bda29c159f16689c05fd4dda20a3f70597e1d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Mar 27 14:01:47 2009 +0200

    Add built-in %getconfdir macro for getting rpmConfigDir() value
    - define %_rpmconfigdir via %getconfdir in the main macro config, this
      avoids it getting lost on macro reloads as happens when building
      several packages at once

commit a5eeff6d42d6452a85942f89a6b3a1196ccc47bc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Mar 27 13:45:08 2009 +0200

    Plug minor memleak from default rc and macro file paths
    - build code can call these several times, only init once

commit b2e85dae836e9695e4ef3e83c819b3523ae81634
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Mar 27 13:44:46 2009 +0200

    Plug minor memleak from rpmGetCwd() on build

commit a17b3b68f2c2320cd81e285e02c05ab881ad07fd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Mar 26 15:31:10 2009 +0200

    Eliminate pointless xstrdup()'s, the strings are never modified

commit 144830cf816c9189906b0ae826418dc079ec4348
Author: Florian Festi <ffesti@redhat.com>
Date:   Thu Mar 26 13:53:26 2009 +0100

    Fix: use the file's color to decide whether to add it to the rpmal not the color of the rpmfi

commit 1ce328aa29a7f32dde2e831856bf46efb9e50df1
Author: Florian Festi <ffesti@redhat.com>
Date:   Thu Mar 26 11:48:09 2009 +0100

    Remove rpmal from the Python API
    as it changed and never was usable anyway

commit 3a6f3133a1f7acccb787992008927704c9eb1254
Author: Florian Festi <ffesti@redhat.com>
Date:   Fri Mar 20 08:22:41 2009 +0100

    Reimplement rpmal by using hashes and cleaning up the color handling

commit 3e37044eb970fc0f30942e744fb7558f4f23f528
Author: Florian Festi <ffesti@redhat.com>
Date:   Mon Mar 16 17:02:16 2009 +0100

    Make rpmal store rpmtes and kill rpmte->pkgKey

commit 85a84fdae728138c2c526b010417ffc72c382e06
Author: Florian Festi <ffesti@redhat.com>
Date:   Wed Mar 11 14:15:56 2009 +0100

    Also order erases.
    
     - partially using code from rpm5.org

commit a1910f89e40ffe4c85c73bfb17deaac7419c4dc1
Author: Florian Festi <ffesti@redhat.com>
Date:   Wed Mar 25 16:29:09 2009 +0100

    Implement rpmteSetDependsOn and rpmteDependsOn using a rpmte

commit 64280efc1e5ba39ce05cb304e0fd1cf15843f94b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Mar 26 09:59:01 2009 +0200

    Put the PGP foobar signature generation out of its misery
    - the last freeware PGP version (6.5.8) is from year 2000 and doesn't
      come close to compiling on modern distros, commercial versions we're
      not interested in
    - "PGP" signatures in rpm mean RSA, gpg can handle that just fine since
      forever
    - the code's been unused for years, unlikely to be functional anyway...

commit 4d31b0433f3f804f81955c32e63c93f31506e8ad
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Mar 25 17:28:24 2009 +0100

    process the "./" in file paths correctly (rhbz#491388)

commit 5f445d7e8f14df6976b9fa19e9caf4a0dfbcbebb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 25 17:00:41 2009 +0200

    Make sure fd bundle is always NULLed on rpmVerifySignatures() exit
    - readFile() failure could crash on double-free otherwise

commit 678dceb2455537895e6562886d7887614d546e61
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 25 16:48:24 2009 +0200

    Unify DSA/RSA/GPG/blaa signature verification, simplify
    - verifyRSA/DSA were just the same already, differences are in key/sig
      parameters which dont matter on this level
    - these dont need sigtd for anything, the data has been parsed into pgpDig
      before we get here
    - add extra flag for header-only vs header+payload signatures,
      we know which is it in rpmVerifySignature()

commit 0ce87a7494eadb6f355b561d4e47d2226931a3de
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 25 16:35:41 2009 +0200

    Rename internal signature verification functions
    - size, md5 and sha1 aren't really signatures

commit 87b8402eaa9561fa4143c7ddb035c48a6b8d27bf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 25 16:33:29 2009 +0200

    Reorder internal signature verification arguments
    - move retval msg last everywhere

commit bdd73e4180f6ec2abccfcacda14a0cc629918168
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 25 16:30:20 2009 +0200

    Push signature identification to librpmio / pgp handling
    - retrieve names from pgp tables instead of manually duplicated strings

commit b711341a616a801137d1126067b0c0c877b0b942
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 25 15:42:30 2009 +0200

    Remove redundant RSA/DSA parameter checks
    - if this stuff doesn't match, pgpVerifySig() will fail anyway, dont
      bother checking for same things over and over

commit d65c1787857b270ca258382581644f5cc50c3bad
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 25 15:16:59 2009 +0200

    Push signature verification down to librpmio
    - new pgpVerifySig() call to perform the lowlevel verify, use for
      verifying DSA/RSA signatures
    - librpm is now free of NSS specifics (apart from what still leaks through
      including rpmio/digest.h), remove linkage

commit 19e142ee4e54ad7967f3a6dab9552450b0357e88
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 25 14:39:29 2009 +0200

    These cases are identical now...

commit 127eddaa70c5679a7a2683f4574e1198f8e3dbeb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 25 14:37:43 2009 +0200

    Rename arguments, RSA keys are certainly not limited to md5 hashes

commit eb48cfaf6bf4cc8a5ae21aed2796ff7a6e2ef758
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 25 14:17:34 2009 +0200

    Use makeGPGSignature() for both DSA and RSA
    - in reality makePGPSignature() never gets called here due to other fun...

commit e3892e2c8de0640f3e527daf215bc9e03df1e32e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 25 14:09:42 2009 +0200

    Remove unnecessary {} block inflating indentation, streamline

commit a7051403f84887791a05eaa8e20d6782eb0147e5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 25 13:51:47 2009 +0200

    Eliminate copy-paste digest calculation in rpmReadPackageFile()
    - for DSA and RSA we want to grab the hash algo from signature, other
      than that it's just the same for all header-only DSA/RSA/SHA1 sigs

commit 96e9d490a6ffc364ba9b2834bb778ec4c1411bbd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 25 13:48:37 2009 +0200

    Eliminate copy-paste digest calculation in headerCheck()
    - for DSA and RSA we want to grab the hash algo from signature, other
      than that it's just the same for all DSA/RSA/SHA1 sigs

commit 14468255716a211064f7653e9f582f94f4c25f9a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 25 13:23:19 2009 +0200

    Eliminate header/payload digests from pgpDig_s, they dont belong
    - allocate+free digests locally where needed, pass around in separate argument
    - use digest bundles to handle rpmVerifySignatures() needs
    - kill-kill-kill fdStealDigest(), dup the contexts from bundles as needed

commit 3f6b2f8c7b7b79a59664e16d6d350acb54802171
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 25 13:10:23 2009 +0200

    Implement fdUpdateDigests() and friends using digest bundle

commit c9146aa580ab87b93ba5d68cde5b606652c93cd8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 25 13:08:09 2009 +0200

    Add a type + basic api for bundle of digests (on same data)
    - in various places we need to calculate different digests of the same
      data, having them stored in a single container makes managing easier
    - can hold as many digest types as we support
    - remembers how much input data it has handled

commit 2e672f3fae9d8ff7ee64ae13cfa2fe65272c2ede
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Mar 19 16:07:18 2009 +0200

    Show query/verify selection options just once in --help

commit 46f1247a276a66561b49f97364558bf8f89d2354
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Mar 19 14:56:45 2009 +0200

    Python doesn't need libelf, nss or popt directly

commit 1df4d34fb97cd25134a89c06386b55dc98aa92f6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Mar 19 14:41:43 2009 +0200

    librpmio doesn't need libmagic for anything

commit fd2658ec8434768f75a6b6d007f7df5415f44004
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 17 16:33:12 2009 +0200

    Move leftover signature stuff from rpmlib.h to signature.h
    - these aren't really usable or useful outside rpm, and even less so
      as all the rest of the signature stuff is in a private header too

commit 298c919c00ffbc6ea988545092b02af5384bd7a4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Mar 19 10:15:49 2009 +0200

    Revert couple of checksum -> digest renames
    - physical cpio header and elf header have checksums, not digests
      (from commit 7acdfded4a82e86a673418c3953aa6bf8106d313)
    - doesn't affect functionality though

commit 54b027535f95fea9088f2cd12ed765fddb20c5c1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 18 14:53:34 2009 +0200

    Don't try to parse %patch on spec query (rhbz#487855)

commit b187977fee788888007409465c7efa644d20ddef
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 18 11:57:16 2009 +0200

    Doxygen parameter name mismatch

commit fa0898552d3c52249bc41e8104eda02c1bbd3359
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 18 11:24:52 2009 +0200

    Add support for "lzma alone" compression format (lzdio)
    - "lzma alone" is superseeded by XZ but it's used by Suse and Mandriva
      distros so it's nice to have as it doesn't need much, only open differs
      from XZ
    - rpmlib() dependency versions are an uuuuuuuugly mess here: Suse used
      "PayloadIsLzma = 4.4.2-1" so that's what we provide to be most compatible
      (hopefully). Built packages require "PayloadIsLzma <= 4.4.6-1" however
      to be compatible with Mandriva 2008.0. Did I already say it's ugly?
    - Based on similar patch in Mandriva by Per Øyvind Karlsen, but avoiding
      unnecessary renames in rpmio

commit 8078d0ba24662308b10d9eb0f0da978584b7e757
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 18 09:42:23 2009 +0200

    LZMA -> XZ renaming
    - what we support now in payloads is XZ, not the former LZMA format, rename
      user- and header-visible parts to match this
    - rpmlib(PayloadIsXz) dependency versioned 5.2-1 to avoid unnecessary
      incompatibility with what rpm5 has and what Mandriva is already using
    - only provide the rpmlib dependency if actually built with xz support
    - rpmFileIsCompressed() attempts to guess the magicless old lzma format
      by file suffix, so it can handle both right now

commit ff920a9766ea1f3ef9670c8031ff473adb0b0adf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 16 14:19:30 2009 +0200

    Handle rpmDigestDup() failure in signature checking

commit 847c991b2bde08dce508152278f515ae5ff97fd8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 16 14:13:00 2009 +0200

    Proper error handling in rpmDigestDup()
    - tolerate calling with NULL
    - avoid allocations in case of failure
    - return NULL instead of calling exit(), ick

commit ed342d7df6523ae2f7f06e5bb02f766a37614bb6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 16 12:25:03 2009 +0200

    Eliminate unnecessary pgpNewPublicKey() wrappers

commit f79353ea74c3ab255f2f9ee4b117a1fd93bb517a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 16 12:18:01 2009 +0200

    Eliminate unnecessary field from pgpDig_s
    - key/signature can't be both DSA and RSA at the same time

commit c769eb0244294c0a9ac002bd7ecd06bddc4cd3ae
Author: Florian Festi <ffesti@redhat.com>
Date:   Sat Mar 14 19:12:43 2009 +0100

    Free all keys not only those on the first buckets
     - Thanks to Alexey Tourbin for pointing that out!

commit 5f82119a306cc1e647e45df52b93989120e52b5a
Author: Rakesh Pandit <rakesh@fedoraproject.org>
Date:   Fri Mar 13 12:50:35 2009 +0530

    - Cleanup: removing unnecessary repeated code.

commit c6c8fab76ab014cf311cae6ce1529788cd0da1d0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Mar 13 09:26:04 2009 +0200

    rpm-python only accepts string objects, not unicode
    - Reverts broken commit 832fe4f01865cd17ab9393fc48b960206da223b0.
      Anything trying to pass in unicode objects is already broken as
      PyString_AsString() uses default encoding for unicode conversions, which
      by default is ascii (at least in python >= 2.5) so any non-ascii string
      will cause it to fail anyway.
    - Only accepting strings keeps encoding madness out of rpm, thank you Toshio.

commit 9317ebc83f611075874713a1099b1f5aa14c969c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Mar 13 08:55:53 2009 +0200

    Catch exceptions from python conversions in dbMatch()
    - Prior to this, string conversion (eg on unicode object) could fall
      through to rpmtsInitIterator() with key of NULL, matching everything
      and only much later spitting out an obscure traceback. PyInt_FromLong()
      can fail too, causing us to look for bogons.

commit ffe2f39adaea9f113a73aea4537a930ba99e49ff
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Mar 12 11:42:42 2009 +0200

    MD5 and SHA1 digests are handled just the same here...

commit 8eb80293d7edb3c4276730519b42ae567a7198ff
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Mar 12 11:39:29 2009 +0200

    Rip all handling of broken md5 sums from 1997, rpm 2.x era
    - we dont even recognize rpm 2.x packages as rpms...

commit 12f516842161899cdd9aa7fb5f215f19c7b7dab4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Mar 12 11:17:27 2009 +0200

    Eliminate can't happen code
    - sigtd.tag can't very well be something else than it's matching case
      in a switch, eh

commit c1180e3c4334b850d50f618f59de09a9ea06a025
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Mar 12 09:54:31 2009 +0200

    Silence compiler whining on rpmsort and rpminject

commit 23d202e928d4af945765e501ee8fc7bd0614510e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 11 21:43:44 2009 +0200

    Plug a memleak in genCpioListAndHeader()

commit 4cf26d30f8a066f855117d1661b23521e6eed647
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 11 21:41:57 2009 +0200

    NSS supports MD2, we might as well handle it too
    - also getHashType() is a "pure" function, mark it as such

commit 15042bc9095964f9fc5f5ed046f9299fad7663c1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 11 18:10:19 2009 +0200

    Eliminate now unnecessary FDDIGEST_t
    - contexts know their hash algorithm, this is not needed anymore

commit 7fc5e62eaaccd75c1ef27b1c3d21e7cefacf2772
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 11 18:07:22 2009 +0200

    Remember hash algorithm in digest context
    - alternatively we could reverse map NSS hash types (HASHContext
      knows its type) but shrug...

commit d05e4438a9efa5ad2cf0ecfef1b4537c37e132d4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 11 17:14:25 2009 +0200

    Streamline rpmDigestInit() a bit
    - single point of exit, rearrange to avoid having to free if stuff fails

commit 31288616ce442203df626ef0a49ebc3893258fc9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 11 15:28:30 2009 +0200

    Move headerIsEntry() check of sigtag into _chk() macro
    - ...and reformat the if-elses now that they fit into single lines nicely

commit c7aad74879349dcb0a7b3fbdae9ba4fdc54adea4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 11 15:16:26 2009 +0200

    Size signature failures are always RPMRC_FAIL
    - these failures cannot be "fixed" by importing a key, RPMRC_NOKEY is wrong
    - streamline a bit by assuming failure

commit b8b970a94c310346666069c781dc185d6a54aff0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 11 15:07:06 2009 +0200

    Eliminate asserts() from rpmVerifySignature() paths
    - handle rpmVerifySignature() getting called with NULL result cleanly,
      the lower level code doesn't need to double-check

commit 827101840f7327c117467a5afe063b21cd105920
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 11 14:51:46 2009 +0200

    Minor cleanup + simplification of DSA and RSA sig checks
    - eliminate unnecessary local variables
    - initialize on declaration where reasonable

commit cf2637fbb96eb84d7c5a70df8179abc439359472
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 11 14:30:16 2009 +0200

    For RSA sigs, RPMRC_NOKEY is only sane for keyring lookup failure
    - non-keyring lookups fail with RPMRC_FAIL, streamline the code a bit
      by assuming failure

commit fc023a2885bb30ae65c5b56d687bfacaf6c62b77
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 11 14:04:16 2009 +0200

    For DSA sigs, RPMRC_NOKEY is only sane for keyring lookup failure
    - assume failure and non-keyring lookups fail with RPMRC_FAIL

commit 73e0ee5df95ac22cbed73b18575d33c284f62975
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 11 14:01:11 2009 +0200

    We already assume failure, this is unnecessary

commit 32174fa9bf9fcb1f2655b46f04f6fbfc91704f06
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 11 13:58:51 2009 +0200

    SHA1 and MD5 failures are always RPMRC_FAIL
    - these failures cannot be "fixed" by importing a key, RPMRC_NOKEY is wrong
    - streamline a bit by assuming failure

commit 8b771f6906d98563b9e1811482004c97d837d6e7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 11 13:56:08 2009 +0200

    Use tag to distinguish header-only vs header+payload signatures

commit eab03f0e3b1d10e82ee20110bee8bc1d00ee991f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 10 14:27:08 2009 +0200

    Lift best signature detection out of rpmVerifySignatures()

commit 65f7f7c3ce1a9b5245bcd9deba0c6346cce9acd0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 10 13:52:41 2009 +0200

    Lift pgp parsing into separate function
    - similar to dc5fa254588be57a066ef4679c75e2bb3bda5fef

commit 33a6ed91e4a26201d2dec0bf04ab025642528b13
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 10 13:45:12 2009 +0200

    Avoid rasprintf() on potentially NULL string

commit de00e073c83644976127ae271c476fba46e66596
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 10 13:38:26 2009 +0200

    Eliminate Yet Another temporary result variable

commit a526d10e4f84178a7a6fa6b8cbb384aa88c9b16f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 10 13:33:57 2009 +0200

    Further rpmVerifySignatures() cleanup
    - initialize variables at declaration time
    - split the formatting of the result spew into separate function
    - if multiple missing/bad keys, report them all

commit 24c14d0a2ee847b75cb3fe7a666b792ea303ef14
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 10 12:10:49 2009 +0200

    Assume failure in rpmVerifySignatures()
    - simplifies the code a bit

commit 4ffdc83282e44d14115684d4349fc55506d3cbbe
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 10 11:09:35 2009 +0200

    Use rstrcat() instead of rasprintf(), less fuss involved

commit 05e16d415492f4925782e3fcf6ef155cbdf124c3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 10 11:03:08 2009 +0200

    Fix memleak on signature checking failure
    - early exit from header iteration loop could previously leak from unfreed
      buf and header iterator

commit 493913ed415acf8be26b13a785a9615ba5cb6981
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 10 11:01:32 2009 +0200

    Eliminate pointless {} block inflating indentation level
    - no functional changes

commit 65f3e55d3592da8fb482b5344a28013029199115
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 10 10:17:13 2009 +0200

    Don't return data from headerGet() on regionSwab() failure
    - callers dont expect to free data if headerGet() returns failure,
      leaking memory

commit 74c9a96c3c68de6376ee543d494a97c0f133bfaf
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Mar 9 20:19:37 2009 +0100

    Port XZ payload compression to xz-4.999.8 API
    - use SHA256 digests for archive error detection
    - use 100MiB peak memory usage for decoding
    - use better XZ detection in configure.ac

commit f70830356fd239d66e1aece81973ebc077acbe6e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 9 16:44:47 2009 +0200

    Add test for _install_lang and status

commit cdcbd324fe41cd729434576200593c0fbda44a19
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 9 14:57:46 2009 +0200

    Unbreak %_install_langs handling (rhbz#489235)
    - using rpmfiFLangs() in skipFiles() broke the %_install_langs logic,
      causing all files to be skipped if install langs, eek

commit 2b4507d852ac8469608bef2ce8e219d76b0c543e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 9 14:48:47 2009 +0200

    Fix RPMTAG_FILESTATES in rpmdb
    - sizeof(rpmfileState) != sizeof(char), and char is what goes to headers
      resulting in some pretty weird states despite being correct on disk
    - add rpm_fstate_t type for the header presentation of states and
      use where appropriate

commit 1c1faec7a30099f62c12ccef2f15e37a202e6518
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Mar 8 17:38:49 2009 +0200

    Initialize on declaration, cosmetics only...

commit 37f85b81bbbd902f5c66a2cc539b97713d6e59ff
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Mar 8 17:36:39 2009 +0200

    Dont bother retrieving colors unless actually needed

commit a526ff75f490b4f9a83909c856de2f21066a0f3d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Mar 8 17:28:17 2009 +0200

    Dont bother retrieving ts problem set unless actually needed

commit 90582c6103071cf73f79e7dcc2a260298aadf554
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Mar 8 17:25:45 2009 +0200

    Dont bother retrieving altNEVR unless actually needed

commit ac7c3412278a03da6633758bca999827d4b59038
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Mar 8 11:33:22 2009 +0200

    Load macros before creating directories (rhbz#489104)
    - %_sourcedir and friends can have things like %{name}, load macros
      before trying to create any directories when installing src.rpms

commit e9cc481202429ed06b2b0b14cb1b1065df6c8ba7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Mar 6 13:43:17 2009 +0200

    Document libcap, libacl in INSTALL

commit 095e1ba71d28300a31bf411f1e9045bbb539a130
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 4 15:12:02 2009 +0200

    Condense "shut up gcc" switch-cases to just default no action case

commit dc5fa254588be57a066ef4679c75e2bb3bda5fef
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 2 13:23:35 2009 +0200

    Add internal helper to verify signature parse results
    - eliminates bunch of copy-paste variants of the same thing
    - additionally verify pgpPrtPkts() return code

commit c7fc09d585ff3831924f72f61d990aa791f2c3f2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 4 14:40:21 2009 +0200

    Move the common part of DSA+RSA verification into helper function

commit 5cb107b843797307e2a0d21018707bd893222fe5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 4 13:41:27 2009 +0200

    Dont leak memory from hdrmd5ctx

commit d50db40ceed7083467f7b548da7b2fbe96aaec61
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 4 13:08:30 2009 +0200

    Handle V4 signature trailer for RSA too (ticket #34)

commit 682b3d8aa9a0f0a777a3858887f09c1c098db649
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 4 12:20:44 2009 +0200

    verifySHA1Signature() doesn't need pgpDig at all

commit a54ab2c1b94416e4dbf36ecd26fe7885bcd19295
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 4 12:19:02 2009 +0200

    verifyMD5Signature() doesn't need pgpDig at all

commit baac9bd4f8e5f216729a5000429a4bda74aec01a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 4 12:17:21 2009 +0200

    verifySizeSignature() only needs size, not all of pgpDig

commit 30f60665bbbe2547551531a3da299ca8818e74bc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 4 11:49:29 2009 +0200

    Eliminate unnecessary fields from pgpDig_s
    - sha1(len) and md5(len) are only needed and used locally in
      verify[DR]SASignature(), no point carrying them around in pgpDig
    - use wrapper function to hide type differences between rpm and NSS

commit d39a6c7de51c0d01ce69ee1f464b94ca70309751
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 3 08:48:12 2009 +0200

    Add ISA bits for alpha (Oliver Falk)

commit c45cfb9c047b05ca29acb8d42186e97e17204e27
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Mar 2 10:29:05 2009 +0100

    Fix documentation and help for the md5 -> filedigest move
    - add aliases for obsolete {RPM}VERIFY_MD5 flags to
      not to break API

commit 7acdfded4a82e86a673418c3953aa6bf8106d313
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Fri Feb 27 14:34:39 2009 +0100

    Add md5->filedigest aliases (rhbz#487597)
    - add %verify(nofiledigest) as %verify(nomd5) file attribute alias
    - reference digests as digests not checksums
    - make old md5 related symbols obsolete and use newer reference
    - update man page accordingly

commit 299a7af7a8acc0fa4979d71dfaebe87658ee0c1d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Feb 26 21:48:06 2009 +0200

    Missing test material to tarballs, oops

commit fca96226f99c67b70854a66bf6eb8648497d9e23
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Feb 26 17:07:39 2009 +0200

    Permit python bytecompile to fail for now
    - various things like python templates in packages, jython code etc
      blow up on the syntax checking, only warn instead of failing the build
      until parametrized with a spec override possibility

commit 0e6c5e890460224f455c0ef9e516ea675b00208d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Feb 26 16:38:10 2009 +0200

    Honor --nomanifest on install too
    - bit of a kludge but works.. and fixes test case 39

commit caf759f8ca53fc45f49f4bc07dbf1724ec9d5b93
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Feb 26 16:09:51 2009 +0200

    Switch to libtool versioning, starting at 0:0:0

commit 3448b552964a526641d2e85b4ed27ebe3465f100
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Feb 26 10:49:38 2009 +0200

    Handle empty argv passed to rpmdsDupArgv()
    - same as 94552b96256c3620b4be407c501d0d926c081963, apt-rpm expects to
      pass empty version as NULL to rpmdsSingle()

commit f6985705749403735207dee509b1a3e88a6df1ce
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 24 14:24:40 2009 +0200

    Clean up runTransScripts() a bit
    - move progtag figuring out of the loop, and return with error code
      instead of asserting on invalid tag

commit 493275977c5c3bf152d783a850eb5a45b2ade040
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 24 13:55:27 2009 +0200

    Single point of exit from rpmtsRun()

commit 8fd5ed25b4ae44b726894389927ca7cbfb691f55
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 24 13:41:18 2009 +0200

    Split initialization of flags etc out of rpmtsRun()

commit 8dff47155ab8fa90fd2d219e68d0a8371902a4d4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 24 12:35:21 2009 +0200

    Formatting cosmetics only

commit 690a297ec0974c28312ba73bf0cee2624527b8b7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 24 12:31:22 2009 +0200

    Calculate need to chroot just once

commit 750a8e656252413b3290497444f2d403890ad8a2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 24 12:22:21 2009 +0200

    Eliminate unnecessary variable
    - per-fi filecount only used once per loop for stats

commit 720a8ebe3183ec067aed1c55bf77ded8f57569f5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 24 12:13:26 2009 +0200

    Push fingerprinting chroot in+out into rpmtsPrepare() too
    - rpmtsPrepare() can now fail (for chroot), handle exit

commit a309664ba3fd87b5fa7ce8ccd6b60a7a53754ae7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 24 11:50:23 2009 +0200

    Push the rest of the prepare-calculations down to rpmtsPrepare()
    - total file count only needed during prepare, pass around as necessary
    - all fingerprinting related activity (skipping files, memory allocations
      etc) are now handled locally from rpmtsPrepare()
    - throw out some ancient no longer relevant comments

commit 924a5d82f60a964b4fab7f173a3c86d94c831921
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 24 11:23:25 2009 +0200

    Eliminate fingerprint hash from ts struct
    - only prepare stage needs, alloc+free locally and pass around to
      helpers as necessary

commit 9e9060269386c76701358a9840a743008e0e37e0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 24 11:15:53 2009 +0200

    Split fingerprinting stuff out of rpmtsRun()
    - only prepare stage needs fingerprint associated caches and hashes,
      push the allocations where needed
    - do fingerprint additions in separate function

commit 15f17c4b48ff75673acda7bb074ae5eea9a1787b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 24 10:40:43 2009 +0200

    Split prepare phase out of rpmtsRun()

commit ddf1df4abf7b6d810064de7e4a1c500ccda805c0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 24 10:28:27 2009 +0200

    File skipping only done for added packages, simplify

commit 3bf701761e2ec16317d51835ac4c3e7f3f60e8b7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 24 10:24:10 2009 +0200

    Eliminate unused variables
    - numAdded and numRemoved not used for anything at all

commit 68995d52386e7bd1512b9d8e9250ee8de833f21d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 24 10:13:43 2009 +0200

    Split sanity checking out of rpmtsRun() to helper function
    - with file counting done elsewhere, this becomes a clean and obvious split
    - ensureOlder() doesn't need ts, just needs the problem set, simplifies
      the problem set refcounting a bit

commit 1cce1b2444bd3f8c56dca01701b062f054f2832c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 24 10:10:38 2009 +0200

    Fix memory leak from unused problem set
    - checkInstalledFiles() doesn't need ps for anything, this just caused
      a memleak due to refcount

commit 62d27e0bd56b765bd32c22f4062e5b864fcf29c0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 24 09:21:20 2009 +0200

    Calculate files in transaction just once
    - move file counting from rpmtsRun() and rpmFindBaseNamesInDB() to single
      helper, store result in rpmts struct

commit 7770aaea226b9a5fbb5296654eed8572f16e8e0d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Feb 23 14:39:48 2009 +0200

    Fix busted up Korean tranlations of manuals (rhbz#466597)
    - previous conversion to utf-8 (from 2003!) was busted, wrong encoding
      or something

commit 58e92b976aebe43ebddbe2d2ec41bff0dd46b6fc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Feb 21 12:11:54 2009 +0200

    Loosen up restrictions on dependency token names (rhbz#455119)
    - Package names aren't restricted to ascii, no point restricting
      dependency names either.
    - This lets UTF-8 to go through but also all sorts of other junk but
      as we haven't got a clue about the specs encoding, no can do. So we
      only check for bad characters from plain ascii.

commit b86d6a76d71c037727b20172299d418c6bc4ae11
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Feb 20 22:02:49 2009 +0200

    Only run script dependency extraction once per file, duh...

commit 89a7b15c22df2c36d4fe85a5e08240a6044ef505
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Feb 20 21:44:13 2009 +0200

    Check for errors from python bytecompile (Tim Waugh)
    - since we're byte-compiling all python pieces, might as well use the
      opportunity to catch syntax errors while at it

commit 519f8fce17fa9fbfabc5673bee1f5b3c23147306
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Feb 18 08:11:18 2009 +0100

    Revert fix for rhbz#442105, it has unwanted side effects
    - mainly RPMTAG_BUILDARCHS missing from SRPM in case of BuildArch: noarch

commit 73d7e90f5e074f8cf03dafa8a4493ca7cbd66848
Author: Rakesh Pandit <rakesh@fedoraproject.org>
Date:   Thu Feb 12 17:57:35 2009 +0530

    -Moved maifest checking from rpmInstall to rpmCheckManifest.

commit 1fbdab96551c55e55fd6d9ebcf7b6a824e594816
Author: Rakesh Pandit <rakesh@fedoraproject.org>
Date:   Thu Feb 12 17:30:21 2009 +0530

    -Move notify flag settings from rpmInstall and rpmErase to setNotifyFlag.

commit 182e2fdfd456257ce369e51c078dda914da42a57
Author: Rakesh Pandit <rakesh@fedoraproject.org>
Date:   Thu Feb 12 17:25:09 2009 +0530

    -Move vsflags calculation from rpmInstall and rpmErase to setvsFlags.

commit 359e16d75f31a046d73e7e931743f94a43b502c9
Author: Rakesh Pandit <rakesh@fedoraproject.org>
Date:   Thu Feb 12 17:13:20 2009 +0530

    -Remove stopinstall variable and moved rpmcliPackagesTotal to correct place. -Changes in rpmInstallSource for getting reused in install mode and removed.

commit 430b8cc3770413bc5b29688ce0823d15bb7de12b
Author: Rakesh Pandit <rakesh@fedoraproject.org>
Date:   Thu Feb 12 15:54:31 2009 +0530

    -Use installInterfaceFlags in place of eraseInterfaceFlags. -Moved transaction part from rpmInstall and rpmErase to rpmcliTransaction.

commit 61a5043d7d04e2ba330a0441cbc1de5f9ea801cc
Author: Rakesh Pandit <rakesh@fedoraproject.org>
Date:   Thu Feb 12 15:49:28 2009 +0530

    -Merge rpmInstallInterfaceFlags_e and rpmEraseInterfaceFlags_e to rpmInstallFlags_e.

commit 51157419623f025149d154c8eb6a0874f1bc23f7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Feb 12 14:11:22 2009 +0200

    Only initialize file capability cache if actually needed
    - very few packages are going to have RPMTAG_FILECAPS at all

commit b9b8af5c4b9518fc36d6ab30b1e6124deb78dc8f
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Feb 11 13:19:22 2009 +0100

    Fix output of pkgconfigdeps.sh so that rpmfc parses it correctly

commit d0595eaa8ad8a9155a0fa4b0b6f4b3d5ed08b2e8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 10 14:11:28 2009 +0200

    Be more careful about errors in lua rpm vercmp()

commit 67d6e7549720551f4750ac303c50e6c7e6ca00f5
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Feb 11 10:03:16 2009 +0100

    Handle broken pkgconfig files better, rhbz#484970 (Lubomir Rintel)

commit 750802fd5f1accd5eacf54e3676f656019a24d1f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Feb 6 09:10:23 2009 +0200

    Bump FileDigests rpmlib dependency version
    - should've really been in commit 1f6ad324dbf4bebec905ddbb76d274345673a5b3,
      the tag adjustment is an incompatible change from what was in
      rpm 4.6.0-rc's, ugh

commit 3d314301ca11b10d2fed8f7a48acb735847d78b2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Feb 5 22:37:27 2009 +0200

    Use per-rpmfi string cache for file capability strings
    - most packages wont have file capabilities attached at all, and even
      where present, the files with capabilities are likely to be few and
      far in between, avoid wasting memory

commit b04c67136f6e980f7eeabd5708e37ac078b0275c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Feb 5 17:22:00 2009 +0200

    rpmfi is only needed by verifyHeader(), push it there
    - no functional changes, just slight reorganization

commit 990bdf1683c6aec9caeadeeb959029dc411935f8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Feb 5 12:34:07 2009 +0200

    Fix typo in python bindings documentation (Masatake YAMATO)

commit 1c3bd7c7ca56acfc4611fb8535acd453c159dc20
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Feb 5 12:20:20 2009 +0200

    Exclude bin and sbin directories from python bytecompile (rhbz#182498)

commit f456b6feb12e84379e9b3f4c44c71b1260947de2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Feb 5 11:36:58 2009 +0200

    Updated Brazilian Portugese translation (Igor Pires Soares)

commit 62091898aeae70f71d13923f18855174a7a65251
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Feb 5 11:32:18 2009 +0200

    --requires lists required capabilities, not packages (rhbz#483253)
    - also fix a minor typo in --whatrequires description

commit 5e71c9bd7635aff9630f74f1b19fabb81ff798de
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Feb 2 14:24:33 2009 +0200

    OSGi dep extractor update (rhbz#466111):
    - From Alphonse Van Assche: "This patch fix some parsing bugs and add some
      new functionality (disable for now)"

commit 6ce7def270994a675836e2b945a7f70eb2b03c2b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Feb 2 14:17:27 2009 +0200

    Prepend PKG_CONFIG_PATH on pkgconfig requires extraction too (rhbz#473814)
    - similar to ab02fb183a441b6a30c863aebf49be992cd431fe but for requires

commit 1f6ad324dbf4bebec905ddbb76d274345673a5b3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jan 30 13:53:18 2009 +0200

    Undo tag clash on RPMTAG_FILEDIGESTALGO[S] with rpm5.org
    - no rpm.org distro uses non-md5 digests yet for anything real, revert this
      unnecessary clash while we still can

commit 64335052e35836168f18e190da464b8c2abe0373
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jan 30 08:28:57 2009 +0200

    Try to play nice with OpenPKG (rpm5.org) packages (rhbz#478907)
    - rpm5.org writes populates lead with partially bogus data, but
      does have explicit RPMTAG_SOURCEPACKAGE for srpms, dont think of them
      as binaries
    - switches around the lead check in 50a4ed783c300a92948f7933a3c726d4bd2d643d
      to look at source type explicitly and retrofit RPMTAG_SOURCEPACKAGE
      for srpms early, this lets binaries without RPMTAG_SOURCERPM
      (which is documented as "information only") and with explicit
      RPMTAG_SOURCEPACKAGE be identified correctly

commit 81afe5e6a210171f58c989bc43bab1881821ddad
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jan 30 08:18:58 2009 +0200

    Updated Swedish translation (Göran Uddeborg)

commit fe720f6c3c0657d0ed4fc3855ac85a727dac47ce
Author: Ville Skyttä <ville.skytta@iki.fi>
Date:   Thu Jan 29 20:42:43 2009 +0200

    Improve Qt translation support, particularly --all-name.

commit dbc1f4ccf66451568a0a6b6ffd4252475a30b341
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 28 16:34:06 2009 +0200

    Don't throw away modified config files on cross-hash upgrades (rhbz#479869)
    - different digest types can't be compared, take a backup to avoid
      data loss
    - Patch from Miloslav Trmac

commit 381605a9bb32cf88209a2dbb72f58c31167779e9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 28 13:21:39 2009 +0200

    Honor rpmtd type, not tag type, in rpmtdClass() (ticket #25)

commit 216eb1c7e02b4bfb9a1d301538b41533db489b55
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 28 13:17:17 2009 +0200

    Add API for getting the class of tag type directly
    - no point requiring going through rpmTagGetType() if we already have
      the type at hand, and in some nutty cases (wrt extensions) these dont
      even always  match
    - rpmTagGetClass() is now just a convenience wrapper around
      rpmTagTypeGetClass()

commit fe846402a45dc1ceadf3a7538071b40b0fd70a60
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jan 27 14:21:40 2009 +0200

    Add RPMTAG_SOURCERPM to gpg-pubkeys in rpmdb
    - ptooey, but otherwise gpg-pubkey "packages" are considered source packages,
      thanks to Ville Skyttä for pointing this out

commit d55660aaf0dbf86a73e58479a47055c9e6e8bbab
Author: Alexandr D. Kanevskiy <kad@kad.name>
Date:   Sat Jan 24 00:18:34 2009 +0200

    Added ARMv7 architecture support

commit 0611b59a4a408579db66f0742a7df5ca6c001fec
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jan 23 14:34:08 2009 +0200

    Check for binaries in noarch packages at build-time
    - this will only catch elf files for now, but better than nothing
    - default to terminating build (as binaries in noarch package is pretty
      nasty thing to do), permit macro override just in case though
    - only works for internal dep generator

commit abe5f4abac7a282a15de5325954c3a09893769c9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jan 23 10:08:53 2009 +0200

    Update dbi tag names for Filemd5s -> Filedigests change, oops..

commit 111efde99a42c8841fa50d682fd340c73c00268b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jan 22 12:09:45 2009 +0200

    Return NULL instead of blowing up on invalid rpmpsGetProblem()
    - thanks to Pixel for spotting

commit f128fea076b510bbe934579d9821eff5f9dff802
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jan 20 14:48:05 2009 +0200

    Add a bit of documentation for headerGet() flags

commit 9f755c72c7390304e8799edae3afbf872ad4cff9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jan 20 14:16:58 2009 +0200

    Grab dependency name & version strings as argv-style arrays
    - Related to ticket #22, this fixes rpmdsMerge() on data grabbed from
      header. Prior to this, the following would blow up with memory corruption
      as headerGet() string array wasn't NULL-terminated
     rpmds d1 = rpmdsNew(h, RPMTAG_REQUIRENAME, 0);
     rpmds d2 = rpmdsSingle(RPMTAG_REQUIRENAME, "foo", "1", 0);
     rpmdsMerge(&d1, d2);

commit 05f5c353242d0dfa70ef49918c73c92dec923920
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jan 20 14:15:24 2009 +0200

    Add headerGet() flag to request argv-style NULL-terminated string arrays
    - easy to do and some places would like the data this way so why not...
    - also add corresponding rpmtd flag so caller can verify he got what
      was requested

commit a46dd884ec348b3ce23b9f53302e6626a90ae631
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jan 20 12:55:35 2009 +0200

    rpmdsMerge() expects ds->N and ds->EVR as argv-style arrays (ticket #22)
    - use rpmdsDupArgv() instead of cloning headerGet() behavior to fix this for
      rpmdsSingle() and rpmdsThis() generated dependency sets
    - this is a regression from rpm 4.4.x, but for rpmdsNew() created sets
      rpmdsMerge() has never worked as headerGet() doesn't return NULL
      terminated arrays

commit 6bd713ab319514d655d0466330a246703a7c16ae
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jan 19 14:50:18 2009 +0200

    Update / clarify documentation on building API/hacking docs

commit 479a4a9d75a1973247f1a937228ec96479f88c31
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jan 19 14:37:09 2009 +0200

    Don't segfault on "" rcfiles argument (spotted by Pixel)

commit 2084762882240f9a5b5bc27c4623291e745611a5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jan 19 14:29:48 2009 +0200

    Missing space in Russian translation (patch from Mandriva)

commit b4bb89603844699754695e9f1484779e758060f3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jan 19 14:27:39 2009 +0200

    Use temporary table for Depends DB in sqlite too (Olivier Thauvin)

commit 2a20195616ad366f060bea772342a99b96735205
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jan 19 14:25:55 2009 +0200

    Fix segfault if ts rootdir is NULL (Pixel)

commit ffcd6a60fc38dfe0d10f22c6023df3e932996310
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jan 19 14:21:45 2009 +0200

    Fix memleak from i18n strings if retrieved with HEADERGET_ALLOC

commit 0b2d7775c5e828652e45829f551352b93890bbc8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Jan 17 14:15:50 2009 +0200

    Make it possible to disable %pre- and %posttrans scriptlets (rhbz#475582)
    - rpmtransFlags_e is overcrowded, map to "regular" %pre and %post
      scriptlet behavior for now
    - split up the transaction flags at some point to permit more precise
      control

commit 947ad7acb3e237446f0e20c0009e9ae07e0f26dc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jan 16 09:09:40 2009 +0200

    Add getenv macro, use instead of $(echo $HOME) for defauls
    - inspired by http://lists.rpm.org/pipermail/rpm-maint/2009-January/002332.html

commit 91e0004f0bb772e7e6f9f84b8a476b5e5783d97e
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu Jan 15 13:25:08 2009 +0100

    Update Brazilian Portuguese translation (rhbz#480040)

commit af9801a162ef41ca85ca62a30b17f745d2ac855a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jan 15 10:28:13 2009 +0200

    Add extra parameter to rpmteClose() to control fi behavior
    - pre/posttrans scripts should leave te->fi alone, now it was getting
      nuked, oops..

commit 6d26397a616586999f93fc4de30203db267a28f4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jan 15 09:22:58 2009 +0200

    Make sure global state gets NULLed on free
    - otherwise repeated read config -> free config like rpmbuild does
      will crash and burn
    - somewhat kludgy, figure a better way to do this

commit 267e77f1c1ef007138610a8812a7ca1e3b632eca
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 14 15:18:25 2009 +0200

    Add central Lua shutdown function and use it
    - makes sure lua garbage collect runs
    - memory used by lua gets freed

commit 6be1bdf1252c42a1b7329185b0c6e589b47ca948
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 14 15:09:25 2009 +0200

    Add librpm level lua extension stub
    - inspired by rpm5.org, details differ
    - just rpm.vercmp() for now
    - push all Lua initialization to rpmLuaInit()

commit 2e0d5ef2d0b90fc43ac015396e78cd15225bb86e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 14 14:39:55 2009 +0200

    Add rpmluaGetGlobalState() to internal lua api
    - retrieve global state, init if needed
    - inspired by rpm5.org, details differ
    - use it for initialisation in rpmrc instead of abusing rpmluaGetPrintBuffer

commit 11c6c2b00454c68dba6cd17600bb7f8e1acada99
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 14 10:52:41 2009 +0200

    Actually check for dbenv->failchk() exit code
    - failchk() can't resolve all cases, we shouldn't try to continue if it fails

commit 2f239d1dfa47da51ebc002f1229eecf3c71b3271
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Jan 13 19:28:47 2009 +0100

    Remove redundant anyarch check

commit f4c997cac8a9205067a584f03ed7039cd0877b4e
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Jan 13 19:26:35 2009 +0100

    Simplify group tag inheritance for a bit

commit cc38c36a3163f17110d8005bf5069465b27cdc6c
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Jan 13 09:20:00 2009 +0100

    Inherit group tag from the main package if unspecified (rhbz#470714)

commit 744846875b198558c904fa9fca3c55bfcfb647cb
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Jan 12 16:15:27 2009 +0100

    Ignore BuildArch tags for anyarch actions (rhbz#442105)

commit 0ba387b45b71096a75eb66bb38805af99d0c70b8
Author: Florian Festi <ffesti@redhat.com>
Date:   Fri Jan 9 17:27:35 2009 +0100

    Make the data array part of the hash bucket to save one pointer per bucket

commit a7716784f26c11efe00cc737246f63e29a1aac71
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jan 12 12:57:03 2009 +0200

    Updated German translation (Fabian Affolter)

commit e58db2eda1609a8b48d8248390a2d87ff99fdd6d
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Sun Jan 11 16:54:07 2009 +0100

    Don't check package BuildRequires when doing --rmsource (rhbz#452477)
    - export doRmSource() and make it more consistent with the API

commit 1909784075c064ec8102b785415e6e7b832df3c3
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Sun Jan 11 15:23:59 2009 +0100

    Don't fail because of missing sources when only spec
    removal is requested (rhbz #472427)

commit 83e5726aadf1ca9111f3cf61130b2be9e3ea4c29
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jan 9 17:13:19 2009 +0200

    javadeps format string fix from Pixel

commit e1474f4b1d71a0021ca359dbdd60a17984eb0ff2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jan 9 17:05:40 2009 +0200

    Don't bother calculating string sizes manually
    - ...and one less strcpy() to grep for

commit 00d48deffecb109c1dcfff2ee9c9a27ca412a16f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jan 9 16:59:43 2009 +0200

    Fix buffer overflow on oversized filelist entries (susebz#397006)
    - original fix by Dirk Mueller

commit fcbe5a6e01661e0ffc75a95252789f6309373edf
Author: Florian Festi <ffesti@redhat.com>
Date:   Fri Jan 9 15:36:36 2009 +0100

    Implement dbiGet and dbiPut as normal function

commit a3d682200bd0a74f2c05f9f618700ebcbc00e0c7
Author: Florian Festi <ffesti@redhat.com>
Date:   Fri Jan 9 15:25:56 2009 +0100

    Remove unused database intex functions: dbiCdup, dbiAssociate, dbiPget, dbiJoin
    and their implementations in the backends

commit 6e816c44dc70197477845de924b9d58f35d01e6c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jan 9 15:05:40 2009 +0200

    Don't try to create "" directories

commit b7ba50c3a5ab54a142d6616f1f4029c475f23cf5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jan 8 15:04:58 2009 +0200

    Update docs wrt gpg path macro (rhbz#476201)

commit 685361038e87d321bd0f0d1451471df923ce667b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jan 8 13:57:59 2009 +0200

    Don't bother translators with messages from unused utilities (rhbz#466834)
    - these things dont even get built normally, translations hardly needed

commit 5ff05aec30ddceb0fc8f2ecd7797aa7bdc9bc96d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jan 8 13:17:22 2009 +0200

    Delay NSS initialization until actually used
    - since NSS is allergic (ie becomes non-functional) after forking, delay
      it's initialization until really needed, ie lazy init in rpmDigestInit()
    - however as NSS init can fail if attempted in completely empty chroot,
      we force crypto init to happen at transaction set create time, forking
      past that is pretty much doomed anyway
    - this is the other half of the fix for rhbz#476737, and similar case
      noticed by Pixel in Mandriva (due to urpm forking)

commit c895fb1541f5edbdc53eb6d70eb6157566e6e528
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jan 8 13:03:55 2009 +0200

    Only touch _crypto_initialized if we actually change it

commit fa299eec0cca00e87a826e2c487705cb6c80c217
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jan 8 12:24:50 2009 +0200

    Cast to intptr_t instead of int to kick out the final warning

commit bb45458b6eb11149011a88d12b38f05569571c6a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jan 8 12:07:54 2009 +0200

    Include isaname support for arm (Kedar Sovani)

commit 4516645506211e32f64b1cffc20d1a93d7e47f4c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jan 8 11:47:31 2009 +0200

    Make rpm_tid_t unsigned type to shut up remaining compiler warnings

commit 882e8df3d0b0d1f88db7550fcc83b821dc0ba93a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jan 8 10:38:29 2009 +0200

    Don't segfault in pgpPubkeyFingerprint() if rpmDigest* fails
    - this can happen if NSS is not properly initialized, such as after
      forking
    - partial fix to rhbz#476737 (as in "doesn't crash"), deeper issue is
      that NSS init needs to be done later than we do now

commit 4bf615989c2c7519abab043f4a152f594d8b7a69
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 7 15:54:06 2009 +0200

    rpmRelocateFileList() doesn't need rpmte for anything, just file states

commit 54cabb77124e633e7c60b6293869bf72665abb32
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jan 5 17:43:37 2009 +0200

    Validate + sort relocations in rpmte at create time
    - instead of storing "raw" relocations in rpmte to be passed down to
      rpmRelocateFileList() for copying over and sorting etc, do the work
      just once at rpmteNew() time
    - much of the real work to relocate header contents still needs to be done
      twice as the header gets thrown and needs to be reconstructed in
      rpmtsProcess()

commit c115ea730ff7ac078f4269328b4b558f08911d25
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 20 09:41:04 2008 +0200

    Split saving relocations to header out of relocateFileList()
    - also realizing we can just headerMod() the td's we originally got,
      they contain the modified data now... duh :)

commit 0d472e2574009a895ff2bb7f7f36bf99847e2778
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 20 00:43:51 2008 +0200

    Split sorting out of relocateFileList()

commit 0b8f01c19a4ad1ca8957e9c3c78e79f7ee442905
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 19 20:05:55 2008 +0200

    Split RPMTAG_INSTPREFIXES addition out of relocateFileList()
    - just splicing up the monster function...

commit 36541e3b99afefa86ee6cd285aa2927be26e7d0d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jan 5 15:02:34 2009 +0200

    Relocation needs to be called even if no relocs specified
    - relocatable packages expect RPM_INSTALL_PREFIX in scriptlets even
      default prefix is used, relocation adds RPMTAG_INSTPREFIXES

commit 0ab10fbf3f9e3432f180f1033b4900e2d0521657
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jan 5 11:25:03 2009 +0200

    Lose deprecated DETAILS_AT_TOP doxygen setting

commit 122e2d204e0bdcf081bf1bd05f410a6040a110bf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jan 5 11:18:46 2009 +0200

    Eliminate header magic duplication all over the place, export it
    - avoid several copies of the same thing...
    - there are valid reasons for wanting to know header magic outside librpm,
      export it as rpm_header_magic

commit b0a4ff421bec790eaa3c16684e7ffcc49ac92314
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jan 5 10:42:54 2009 +0200

    Use enumeration for header flags

commit 3de166775cac39a3fc34f225ef4859517040b328
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jan 5 10:39:15 2009 +0200

    Remove unnecessary HEADER_FOOTAG definitions
    - header.c can include rpmtag.h these days

commit abaa8d12b777d03fb9800953ad6a3c9997fdcc40
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jan 2 12:27:32 2009 +0200

    rpmte struct is now fully opaque throughout rpm

commit 54e77ab1e3f0773527ab06294792af23b3b6d780
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jan 2 12:25:02 2009 +0200

    Use rpmteSetFd() instead of direct access to rpmte structure

commit 7d80218f391cf2d6902eb6c031d117956f05e8e7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jan 2 12:21:52 2009 +0200

    Add rpmteSetFd() (internal) function
    - this is dumb as only rpmInstallSourcePackage() needs, figure out something
      better eventually

commit c9e46a798598e30e152338c95225a25072264022
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Dec 23 09:48:16 2008 +0100

    Increase XZ decompression memlimit

commit 64c16c1fa88b35f95726203d23a26de0002f1b85
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 22 15:30:23 2008 +0200

    Eliminate silly and now unnecessary rpmts(Set)RelocateElement() stuff

commit 16987711822e1443792ab5dd09666a7acff93095
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 22 15:29:52 2008 +0200

    Perform relocations from rpmte instead of rpmfi
    - rpmte has all the necessary bits at hand, rpmfi does not
    - avoids silly ping-pong through transaction set to get to the ts element
    - avoids recalculating number of relocations
    - avoids accessing rpmte struct outside rpmte.c

commit dd91ff060c2053a482fd540afa8e30c1f874f3ab
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 19 18:59:38 2008 +0200

    Perform relocations *before* filling out rpmfi structure
    - no need to fuss about recreating bits and pieces of rpmfi after the fact

commit 107d941650eb34197d52404859da9ddf4bf02bc9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 19 18:49:09 2008 +0200

    Avoid a whole lotta shuffling wrt file info header
    - We're not making copies of header or anything here, all this linking
      and unlinking is unnecessary. Just have relocateFileList() modify the
      header we pass to it, as that's what it's doing *anyway*

commit 41b1fe84cc1bb66a9ea4f8e704b3c671a71e2607
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 19 17:46:16 2008 +0200

    Ignore rpmfi flags for basenames + dirnames
    - we make a copy if we modify, doesn't make a difference

commit 31a2eac384b7527471d25f1f93be017882c98aea
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 19 17:44:33 2008 +0200

    Sanitize file modes handling in relocation
    - we dont modify fi->fmodes so fi->scareFlags is irrelevant, just use minmem
      always
    - use rpmtd accessors instead of direct access

commit 1c20dc3aa1de5c2f73a49ed0a2d40b1ef5c005af
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 19 17:34:40 2008 +0200

    Remove unused file/dir color calculations

commit 881a0034f7a93dd19de5fd90c5ad46edc2936c10
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 19 17:27:28 2008 +0200

    Save some trouble by grabbing malloced copy of dir indexes

commit 8becdbfd73d49ca353254d4e40687a9399e240ae
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 19 16:19:32 2008 +0200

    Remove couple of leftover, unused local variables
    - one more fi-> access still lurking...

commit 8db2c32a74cb667051b2b824e434b1e13da6511d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 19 15:05:54 2008 +0200

    Eliminate last remaining fi-internals use from transaction.c

commit 897bbbd7086ef90840a89d7d0bf39721dd698b24
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 19 15:04:51 2008 +0200

    Add internal rpmfi apis to deal with fingerprints, use them

commit fe5e906ab2cfe9ed82ba7fa63277d5ce41451b87
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 19 14:57:37 2008 +0200

    Eliminate file info set argument to rpmpsmNew()
    - ts element carries the info already, no need for separate argument

commit cbcff3bd99f11ad2e8ff8e37ebd32993f25d9a2f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 19 14:34:45 2008 +0200

    Optionally reload ts element file info in rpmteOpen(), free in close
    - avoids having to muck with te internals here and there...

commit 87cf7db1edb922587dd26bb19abada7c4fba51c7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 19 14:13:48 2008 +0200

    Use rpmteSetFI() to clear memory instead of direct access
    - also fixes memleak in case rpmfi has no files

commit de383d126a9e2789afbda412bb8e8d6d02462fee
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 19 14:04:18 2008 +0200

    Add rpmteSetFI() ts element method

commit 4d1826f63fc022456f083076c27313e52a57d24b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 19 13:18:31 2008 +0200

    Free rpmfi fingerprints in rpmfiFree()

commit 83f02cbb74d150cc1c05d73286a4be803f1d5316
Author: Florian Festi <ffesti@redhat.com>
Date:   Thu Dec 18 11:57:02 2008 +0100

    Reduce the hash size of the fingerPrintCache by factor 4

commit feaad9829a264e2dff3549c165278a8629b32c59
Author: Florian Festi <ffesti@redhat.com>
Date:   Thu Dec 18 11:54:45 2008 +0100

    Do only add symlinks the the first rpmFpHash as these are the only one we really need in there.
    Also reduce the hash size.

commit e6f1c592fb7db65d3fd5756d202ec9c4a75ec34b
Author: Florian Festi <ffesti@redhat.com>
Date:   Thu Dec 18 09:41:12 2008 +0100

    Reduce the size of the FpHashes by factor 4.
    It turns out that large hashes leave a lot of buckets
    empty without really reducing the maximum of collisions
    a lot. Tests show that the collision lists are only 2.3
    items on average and the maximum is only about 1.5 worse
    than with a big hash table (~12 for large hashes).

commit b2496694fac38f07d1be61abceaacc46db12ab1d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 18 11:48:09 2008 +0200

    Oops, rpmfiFInode() gone missing...
    - forgotten from commit 9c15574bae8b11a7381c01a1af39e0d59558bad5

commit 465841e8501762b36899b6f31b47968efbe74538
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 18 10:01:01 2008 +0200

    Check chroot() result in psm too (Pixel)
    - bail out early if it fails instead of blindly continuing and potentially
      messing in real root (chroot can fail for priviledged user too)

commit e0f57801c0cb035814509bdae13c3e859d471d59
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 18 09:51:37 2008 +0200

    Permit --noscripts on verify (to disable %verifyscript execution)
    - this used to work but got broken by the rpmi/rpmq/... unification

commit 766821b8a30bc962c8c3b1e5542882110c197d68
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 15 10:26:45 2008 +0200

    Check for termination signals during query loops

commit 7d371fdaa3a796fd23d18ffc80785637262c7434
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 11 10:17:48 2008 +0200

    Add public rpmfi function for retrieving digest algorithm of the set
    - the information is already available through rpmfiFDigest() but that
      requires valid iteration index to be active, whereas the digest algo
      is per info set, not per file

commit 5e23e148aab02a37e0f64c1987aaaf0c80963845
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 11 10:11:59 2008 +0200

    Avoid direct access to rpmfi directory indexes

commit 3c22ec586f58e210331988b56873a109f74ca182
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 11 10:07:41 2008 +0200

    Add (internal) rpmfi accessor for dir indexes
    - except for dnli->active setting in dnlInitIterator(), "give me
      directory matching this basename index" would be sufficient

commit f19c89778520b8c9e61db8744cb45abf489c6132
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 11 10:01:10 2008 +0200

    More rpmfi direct access avoidance in fsm
    - the remaining bits need further accessors / refactoring

commit f32e6c9d05db10122166d15694ad91a68268a9d5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 11 09:54:29 2008 +0200

    Avoid most of direct rpmfi accesses in dnlInitIterator()
    - need to add an accessor function for fi->dil info

commit 284de053fdbc7cd9de2614dc4af042d29abfd190
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 11 09:52:31 2008 +0200

    Avoid bunch of now unnecessary direct accesses to rpmfi struct in fsm

commit f7275ffa1321e6d123ff434e00e4c39b302f83f0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 11 09:23:40 2008 +0200

    Use indexed accessors instead of direct rpmfi struct access in fsmMapFCaps()
    - also rpmts is unused here, remove

commit b24b9ab7ffffda9841a031d75ce444a223d3cb07
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 27 11:19:24 2008 +0200

    Avoid bunch of unnecessary direct fi-> accesses for file and dir count

commit 42df29f4e45ed5cc1e990637ca555b954b980673
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 11 09:15:27 2008 +0200

    Use indexed rpmfi accessors instead of fiddling with iterator index
    - saving and restoring rpmfi iterator index has unwanted side-effects
    - reverts brokenness from 6fd987b7291fb1306297c193f4ab01bf6cd02618

commit 9c15574bae8b11a7381c01a1af39e0d59558bad5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 11 08:57:16 2008 +0200

    Add indexed accessors for rpmfi data, internal only for now
    - fsm in particular wants random access to rpmfi data and there's little
      reason to deny it, file info data is not going to move to linked lists
      or anything like that anytime soon, permitting indexed access allows
      eliminating direct accesses to rpmfi struct

commit ac9c8dc48a8f3efb662b7d615c6a0c94fbe42018
Author: Florian Festi <ffesti@redhat.com>
Date:   Mon Dec 8 17:46:36 2008 +0100

    use RPMFI_KEEPHEADER instead of numerical constant

commit 6e7b664e81b37be79d3dfdda4c7a72c4b7fdc760
Author: Florian Festi <ffesti@redhat.com>
Date:   Wed Dec 3 10:59:05 2008 +0100

    Kill rpmfiUpdateState and free all the rpmfis as soon as file conflict check is done

commit 395b5cffd5a64cdd2a4c2ade06a1a0b8b38d59eb
Author: Florian Festi <ffesti@redhat.com>
Date:   Wed Dec 3 10:39:48 2008 +0100

    Move ->actions from rpmfi to rpmfs
    and adjust all code using it
    also "revert" a4e345455d817ce6c56940600d652ee3ae0aa5d5
     "Move file action calculations to rpmfiNew() where possible"

commit 9c9410dc7afb3bfa9b47d2d4691fabd8f067c86b
Author: Florian Festi <ffesti@redhat.com>
Date:   Tue Dec 2 10:11:44 2008 +0100

    Move the run time part of fstates from rpmfi to rpmfs. Keep fstates in rpmfi as long as it is a real tag read from the rpmdb.

commit 8fa80b0f35574683beb19930264fa3b572c5ed24
Author: Florian Festi <ffesti@redhat.com>
Date:   Mon Dec 1 21:34:59 2008 +0100

    Create a new datastructure rpmfs aka File States that contains the information what to do with the files of an transaction element.
    This patch moves rpmfi->replaced over to rpmfs. Move struct members to follow.

commit 196cc2ad71a6af90961ab62f9f32fc14f7c61878
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Dec 9 16:24:16 2008 +0100

    Nuke the useless bits of rpmsort.c declared in global scope

commit 95b8a910893bba3ad66507975cd96f1c082004e3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 9 14:58:43 2008 +0200

    Update testsuite expectations as per array format change, add another test

commit b838d10e0cf3d747ac0ce5b756688e6677dbc513
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 9 14:52:12 2008 +0200

    Return nothing at all on empty array queryformats
    - (none) might make some sense on non-existent tags elsewhere but
      it's plain annoying on arrays

commit babf17a17dbad3593fe23a407932ba703b5d6d76
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 9 14:15:45 2008 +0200

    Throw out ancient python/ChangeLog

commit 2ba56c673bf1c168cccb827938c09f5ff2a3beda
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Dec 9 12:47:50 2008 +0100

    Move declarations of local variables for parsing functions
    from global scope. Compilers aren't that stupid these days

commit 0643dd44e6e243870868b0fb8196adad1632dfbb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 9 13:23:00 2008 +0200

    Test for fakechroot in configure
    - only needed by the testsuite
    - optimally testsuite should skip parts needing fakechroot if it's not
      there but at least this'll leave a trace into rpmtests.log...

commit 7115b8257edc2374b4833e7c4d23ae94a24f52ea
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 9 10:06:39 2008 +0200

    Add "deptype" format extension

commit bafed0d85c12a2e709dd1a50c817361659bfbd11
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 9 09:51:17 2008 +0200

    Oops, missing wrappers for C++ in headers (ticket #12)

commit 50a4ed783c300a92948f7933a3c726d4bd2d643d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 8 19:51:10 2008 +0200

    Work around broken packages which dont have RPMTAG_SOURCERPM
    - check the lead type and headerIsSource() agree on the type and if not,
      fix it up based on the information in lead

commit 826969f2c7307e7292f192fd10625d863d656723
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 8 19:50:07 2008 +0200

    Add function to retrieve type from rpmlead
    - for some messed up packages there's no other reliable way to determine
      source vs binary than looking at the lead :-/

commit 0eb0e79a80e20723d0ebe58996bc3f30a3feabe7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 8 18:46:02 2008 +0200

    Fix segfault on packages without separate scriptlet program tags (rhbz#475214)

commit ab02fb183a441b6a30c863aebf49be992cd431fe
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 5 21:44:43 2008 +0200

    Fix pkg-config provide generation on interdependent .pc's (rhbz#473814)
    - based on Arjan van de Ven's and Dan Nicholson's suggestions

commit 9799e69263d94ba1aca61795645c750bcd497ee5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 5 21:17:16 2008 +0200

    Avoid "template" which is a reserved word in C++

commit cc74d50a6dadfa30914844e6c4c0513287bf7f68
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 5 13:38:30 2008 +0200

    Tighten up setting and checking of buildroot (ticket #10)
    - expand, set and verify buildroot just once for after main package preamble
      is parsed to avoid sub-packages from overriding it
    - spec can still mess with %buildroot by defining it to something else
      after preamble but that's another issue...

commit 23c55fa997b66f39a50c34512fb07c78343f1062
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 4 22:30:35 2008 +0200

    Adjust test result to meet current expectation

commit e232344a496798e4ca143630a127d8ead086eb7c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 4 21:55:52 2008 +0200

    Pay more attention to tag extension return codes
    - behave like headerGet() does and return failure if there's no data
      to return
    - some cases where we used to return nothing previously now return (none)
      in query, but the former behavior was inconsistent between tags
    - not sure if (none) is really wanted for a return of no value always,
      but at least it's consistent now

commit 1f61d628fdd07a11aa7a84dc15a74150865e575f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 4 19:10:41 2008 +0200

    Fix triggerconds return when no data is there (rhbz#474550)

commit 144463a82d493981281dda24cdbb4b785f7bbab1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 4 19:04:14 2008 +0200

    Avoid rpmtd abuse to report "(none)" when no data is found
    - simplifies things quite a bit and avoids having to deal with
      numeric data suddenly appearing as string elsewhere

commit d9082ba95ccb911a452803078eece0140097e78c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 4 21:25:12 2008 +0200

    Comment out rpmbuild --root tests
    - no point testing what's not really ever been implemented

commit c0e6b4579fa0e392dd2de1d4aa0245983c58ae06
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 4 21:23:06 2008 +0200

    Add a pile of queryformat tests
    - a few fail, and for a reason (rhbz#474550 and breakage from
      tag class changes)

commit ba87a938b77c7ab9ffa4cca1255a877600ba2305
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 4 10:23:40 2008 +0200

    Fix python header retrieval return type for non-existent tags (rhbz#473239)
    - array / not array return needs to be honored even when tag doesn't exist,
      wtf have I been thinking, duh!

commit 5e5aa256d70987dca7af5e6cbeb9269ead0e3011
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Dec 3 14:56:52 2008 +0200

    Add tests to verify on-disk status after install+erase

commit 34159ad112684156aac6e91364c7c7bce1f1df14
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Dec 3 14:17:25 2008 +0200

    Add test to verify status of installed files

commit d54e0e899f02b3e492e7f8e73c10f9c3a8f53034
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Dec 3 11:29:19 2008 +0200

    Remove unnecessary fiddling with rpmte internals
    - rpmtsClean() will free the transaction element we added and rpmteFree()
      will in turn close the file descriptor, free headers etc

commit 0b80d38f32b438d81d47fc61074b1f163518d68f
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Dec 1 12:16:56 2008 +0100

    Include rpmfileutil.h from rpmmacro.h so that
    prototype for rpmGenPath() is included (rhbz#473420)
    - unbreaks net-snmp

commit 7ea78b4927bdf86bc09f2dee025cb43c44609099
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 1 13:03:05 2008 +0200

    Retrieve header from rpmdb on rpmteOpen() if possible, even for TR_ADDED
    - as rpmte now knows if it was already installed and the exact rpmdb
      record, we can avoid having to re-re-re-open the package in %posttrans,
      this means yum & the like can throw away package immediately after
      it was processed (rhbz#457126)
    - rearrange rpmteOpen() code a bit, split the actual header retrievals
      from fd/rpmdb to separate static helper functions

commit c99197616491693fa0fec5ff39ac4c2cd336f65b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Nov 30 17:36:46 2008 +0200

    Set/clear rpmte db instance on RPMDB_ADD / REMOVE from psm
    - transaction elements now know their real status all the time now

commit 31f0e1db5897d127cc772e775d522ff551b36cc3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Nov 30 17:28:11 2008 +0200

    Record header instance on rpmdbAdd() on success

commit 5d1791e32aa8d4b7baef2639a097af7a6efb665e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Nov 30 17:22:24 2008 +0200

    Add and use (internal) method for setting header instance
    - avoid mucking with header internals directly...

commit dc818e63544f8cdd53bb6e1f2a6fba70f6aea437
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Nov 30 14:07:03 2008 +0200

    Move install langs to per-transaction level
    - install langs is clearly a per-transaction thing, avoid checking
      and splitting etc over and over for every package
    - also if any of the %{_install_lang} is "all", disable all language
      processing as we'll be installing every language anyway

commit 5dfc982dda733dbf06128358816bf6d91cd18c2a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Nov 30 13:51:32 2008 +0200

    Move netsharedpaths to per-transaction level
    - netsharedpaths is clearly a per-transaction thing, avoid checking
      and splitting etc over and over for every package

commit 26da2e12d7b11d967656a19300ed871f21bfed63
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Nov 30 13:02:34 2008 +0200

    Eliminate unused filename field in rpmts_s

commit 0372f7d8b4ea14f8248fa287fe43eaf04e93a638
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 29 14:34:01 2008 +0200

    Read default popt aliases again (rhbz#473167)
    - makes testsuite depend on local aliases but using default popt config
      is widely documented and needs to work, testsuite needs a better method
      of disabling it if necessary

commit 604a54277ed21f1fd983c205ad7d36feb3950ce1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 29 13:52:32 2008 +0200

    Unify dbiIndexSet allocations, allocate in power-of-two sizes
    - allocating in even sized chunks helps avoiding crazy memory fragmentation
      when resizing a lot (rhbz#472507)
    - HEAD doesn't really need, rpm 4.6.x does, but doesn't hurt either...

commit 881f14d277eefc70d65bf48b1837f0219572f397
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 29 12:04:46 2008 +0200

    Use proper type (unsigned) for index set offset/count

commit 2d918b5aa0a4afec407937bb6e1d4b7ae6ba4be3
Author: Florian Festi <ffesti@redhat.com>
Date:   Fri Nov 28 12:16:00 2008 +0100

    Delete _dbiIndexItem.fpNum as it is no longer used and also does not really fit in there

commit 17f2cbe6e617dff5516498dd3b4cf15da6d0e747
Author: Florian Festi <ffesti@redhat.com>
Date:   Fri Nov 28 11:14:41 2008 +0100

    Delete rpmdbFindFpList and skipDir
    as those are replaced by rpmFindBaseNamesInDB
    and parts of checkInstalledFiles

commit b12afb1a40d0acdfa9fee058c9c0ef2c5891690d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 27 11:35:56 2008 +0200

    Oops, it wasn't unnecessary include afterall..

commit 0b5e0fb58c0577af6824d313c2a06045e94da70f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 27 10:59:14 2008 +0200

    Remove unnecessary rpmfi_internal.h include

commit b76b945b2dc3aee2db429f5bc422780c71b802f2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 27 10:48:22 2008 +0200

    Add + use internal rpmteHaveTransScript() function
    - struct rpmte_s is now fully opaque within transaction.c

commit 8b7cb846d4e466461e9faa34d354cef986ed4472
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 27 10:28:10 2008 +0200

    Eliminate the now unused ts element field of file info sets

commit 2654bc49ab3ac604a3ecd0a39c5815927bbe112d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 27 10:25:47 2008 +0200

    Eliminate rpmtsiFi(), use rpmteFI() instead
    - rpmtsiFi() is only used within iteration, so we already have the
      element, no point in recalculating it
    - also eliminates the sneaky side-effect of setting fi->te which otherwise
      wouldn't exit

commit 9a155dd3505aa4e708c0ad9e0da747626a93c6f5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 27 10:22:45 2008 +0200

    Avoid unnecessary use of file info sets transaction element
    - ts element is available, no need to go backwards

commit 48cefd2d785d3ba70eb5636789d63dd14c229e7f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 27 10:02:00 2008 +0200

    Drop the useless "tag" argument from rpmteFI()
    - only RPMTAG_BASENAMES was ever allowed as the tag, makes no sense
    - API change but AFAIK nothing outside rpm itself uses it anyway

commit 01739ef5d6a56cea818f88a693a21359e3badca3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 27 09:39:40 2008 +0200

    Eliminate silly rpmfiTypeString()
    - Transaction elements are install/erase type, not file info sets,
      rename to rpmteTypeString(). It's only "useful" for fsm internal
      debugging messages, hide it there.

commit c2bde085f546efe016419ee286b2448f63c3a0e0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 26 16:14:29 2008 +0200

    Eliminate fi->te use in fsm

commit 610347de6e07e7cd4ff13cfa9467280ed5e8c175
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 26 16:08:04 2008 +0200

    Pass rpmte to fsmSetup()
    - fsm needs the transaction element for progress notification and
      element type anyway

commit 54b683d2f395998178b760257f1145c246f9e09f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 26 16:01:40 2008 +0200

    Remove bogus consts from fsmSetup() args
    - rpmfi is messed with (at least file actions)
    - const on typedef'ed type doesn't mean a thing anyhow

commit 5bfe30320102e54a0428eb500fe19ec8f7adb5e1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 26 19:04:02 2008 +0200

    Use rpmteNEVRA() for scriptlet messages now that it's possible

commit e56768c279b7214300bade2e82ab90e4d1afe7ff
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Nov 26 16:43:49 2008 +0100

    Increase memory limit for LZMA compression
    so that everything up to level 9 gets safely
    expanded

commit d6f6d2b565f8ccb005b3230fa987a9d961c64533
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Nov 26 16:35:07 2008 +0100

    Initialize the LZMA encoder/decoder on demand, not both

commit 2494b26f01ae7447365fd3b62d028e4402f9cff5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 26 17:41:22 2008 +0200

    Fix up couple of missed pieces rpmffi_s fi->te switchover

commit 7ef6ab6bcf220ca411f764d18ee2b7efef9b81d9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 25 17:05:26 2008 +0200

    runTransScripts() don't need no steenking rpmfi's now

commit 2fb2614c38cd6b7be7ce5698aef03876074eb790
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 25 16:35:21 2008 +0200

    Lose unnecessary and silly rpmpsmGetTs() internal func

commit 26faceebed701d23e775f68d3b10b3ec93cac3ac
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 25 16:32:36 2008 +0200

    Resurrect rpmVerifyScript()
    - fake up a transction element set that we can use to convince psm to
      execute the script
    - a bit hacky but by no means worse than the former fi->h abuse with
      scaremem and all

commit 30d15deedf6b4fc7cfb7d3c252f9d838445979af
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 25 14:01:38 2008 +0200

    Rip the remains of fi->h in psm
    - erasure doesn't need the special casing anymore
    - remove useless comments

commit e939e16364fa4a42880e2b15727c8aa68b83b257
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 25 13:44:51 2008 +0200

    Use ts element, not fi header for PSM_PKGINSTALL (PSM_POST + PSM_INIT)

commit af4d4ed806347acd3cb58d9885191b6a9444233a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 25 12:53:46 2008 +0200

    Use ts element, not fi header in PSM_RPMIO_FLAGS and PSM_RPMDB_ADD

commit 25a02d47bc766386e526c38c9db03a1fbbc4f5f0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 25 12:39:37 2008 +0200

    Make psm scriptlet operations use ts element header instead of fi
    - this breaks %verifyscript operation (temporarily)

commit 440e843f87d190bc197c398510b89d8225a38bce
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 25 12:22:10 2008 +0200

    Lose now unnecessary PSM_RPMDB_LOAD stage and rpmpsmSetFI()

commit 801a36d69beffc383b5cf4fbb63534ead9dfe84a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 25 12:17:47 2008 +0200

    Rework rpmtsProcess() to use rpmteOpen+Close for both install and erase
    - both cases behave fairly symmetrically now
    - helps streamlining the process, we get the transaction element and
      file info for it before creating the package state machine so we dont'
      need to fuss around with updated file info etc

commit 8690479c3c4a1434624341db8b5a1f3914b26911
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 25 11:51:30 2008 +0200

    Teach rpmteOpen() and rpmteClose() about TR_REMOVED type
    - for open, fetch the header from rpmdb, closing is nearly a no-op
    - allows install and erase behave more symmetrically, and also paving way
      for better erase callback notifications and other things
    - watch out for them refcounts...

commit 1cbb50c7b99803613b3dea967bc51b3e2687ffe5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 25 11:46:20 2008 +0200

    Record failure of install/erase into transaction elements
    - add rpmteMarkFailed() (internal) and rpmteFailed() methods to rpmte
    - skip any elements marked as failed in rpmtsProcess()
    - this is fairly close to the previous "now mostly broken" hack, except
      we remember more than one failure so it's slightly better and it's
      possible to do much better by making rpmteMarkFailed() smarter

commit 504b2cf8a74962910502d9389d8e175cedd8fc93
Author: Florian Festi <ffesti@redhat.com>
Date:   Wed Nov 26 16:29:55 2008 +0100

    Change rpmffi_s to contain a rpmte instead of rpmfi to be able to get rid of fi->te

commit a3c316bbcb77849258474242ee862426f521a3f4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 26 08:58:35 2008 +0200

    Don't bother zeroing rpmte twice on free
    - delTE() already zeroes it out

commit 4e4a7a2d8cd9aa99d6767e90a45c9c8940e07bff
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 25 18:06:31 2008 +0200

    Add rpmhash.[CH] to EXTRA_DIST so they get included in tarballs

commit ca8c7dcaef9ed791563d38bfa24c65fcc0edcbd6
Author: Florian Festi <ffesti@redhat.com>
Date:   Tue Nov 25 16:05:00 2008 +0100

    Switch rpmfiUpdateState to the new ->replaced API
    and free the additional allocated but unused memory

commit 028030350a623e4c00c8ecb3c39e09fbb73d1ce1
Author: Florian Festi <ffesti@redhat.com>
Date:   Tue Nov 25 14:33:11 2008 +0100

    Kill handleRmvdInstalledFile by moving remaining code to checkInstalledFiles
    Switch to strict use of rpmtd accessor functions

commit 93ca811b4ff30659c67aa2c54957abe6cafe5acb
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Nov 25 14:12:50 2008 +0100

    Port LZMA (XZ) support to the new API
    - get rid of the obsolete "alone" LZMA file format
      and hacks around it

commit 5dfa17a77c4335dadc652d6c93df2a8bb30c3393
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Nov 25 14:11:08 2008 +0100

    Use the new LZMA (XZ) magic

commit c8d0ea1c7203070fb4553a243418be9608912f25
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 24 16:45:22 2008 +0200

    Move transscripts stuff from rpmfi to rpmte
    - transaction scripts have zero to do with files...

commit dc8f991671b7e0171b6052bc11021e48b16bedf3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 24 15:39:17 2008 +0200

    Unify statistics timer calls for install+erase

commit bc5c667d4f95a6d277860f22889e0be98404f8d6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 24 15:33:23 2008 +0200

    Unify install+remove logging, unnecessary code duplication..

commit 5e4d353d7924dd63d36e5bc20ec37a2303966216
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 24 15:31:31 2008 +0200

    Lift the main install+remove loop out of rpmtsRun()
    - no functional changes

commit 9c66a6daa71bf34dbdf3e8b5f4e7a4fc63d0b322
Author: Florian Festi <ffesti@redhat.com>
Date:   Mon Nov 24 14:34:59 2008 +0100

    Fix timings and no longer pass the file count more than once.

commit 024e370ae700640bfdd9a0d77cb927ed14a1cd01
Author: Florian Festi <ffesti@redhat.com>
Date:   Mon Nov 24 14:10:11 2008 +0100

    Remove check that is commented out since ages

commit c849fac52da4aec3ed133ca65885d8fd2e29416b
Author: Florian Festi <ffesti@redhat.com>
Date:   Fri Nov 21 15:00:36 2008 +0100

    Remove no longer needed sharedCmp

commit 490a0e4affd79908f54f21a5c4c4f5d94521b9d3
Author: Florian Festi <ffesti@redhat.com>
Date:   Mon Nov 24 13:36:28 2008 +0100

    fix statistics for looking up symlinks

commit 95de4c8fef3ad2a5db1d818e4149568e0b9d476b
Author: Florian Festi <ffesti@redhat.com>
Date:   Fri Nov 7 10:58:25 2008 +0100

    Reimplement the check of the files in the transaction against the files in the rpmdb.
    Use a global list of packages/files from the rpmdb that may conflict with files in the transaction to avoid loading packages from the rpmdb over and over again.

commit 35f0d2d3bc1d95993f25b8318d8fb9e33d02360f
Author: Florian Festi <ffesti@redhat.com>
Date:   Wed Nov 19 17:11:59 2008 +0100

    switch to new rpmfiReplaced API

commit 2f58ae809f18aa9fa5ce76900f298ad17ea24283
Author: Florian Festi <ffesti@redhat.com>
Date:   Wed Nov 19 17:09:57 2008 +0100

    API for accessing and creating fi->replaced
     - new API is not compatible with accessing fi->replaced directly!

commit d0651021fd0f8edc31332f9f440ce8b51d9b1b2e
Author: Florian Festi <ffesti@redhat.com>
Date:   Sun Nov 9 18:02:20 2008 +0100

    Expose rpmdbSortIterator in rpmdb_internal.h

commit 69b4090f81c06f12bfd4a9003020f540aebb6442
Author: Florian Festi <ffesti@redhat.com>
Date:   Sun Nov 9 17:54:47 2008 +0100

    Add rpmdbExtendIterator to do repeated searches

commit 88db14fc6e116c67c42c9b56122e2163399994d2
Author: Florian Festi <ffesti@redhat.com>
Date:   Wed Nov 5 15:11:55 2008 +0100

    Support leaving HTDATATYPE undefined to use hash as key only hash (set)

commit 4662e50781d6cef4761bd34bbee7312b3012dd10
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 24 13:33:41 2008 +0200

    Add couple of tests for symlink <-> directory behavior

commit f0d61e42cf69164dca0d20e48caf1643616db2f2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 22 18:00:10 2008 +0200

    Remove RPMCONFIGDIR sedding from Makefiles where no longer necessary

commit a15a475a409e4f929736d28af10bb55b5dfee17c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 22 17:51:42 2008 +0200

    Don't bother setting rpmrc specifically in tests
    - RPM_CONFIGDIR environment is used

commit a0d1153085ed74f758b5f104eef3be1e73506973
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 22 17:47:44 2008 +0200

    Remove -DRPMCONFIG from CPPFLAGS where unnecessary

commit 68230da9cea19ea989c94f4872289e9b5cb2d736
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 22 17:36:05 2008 +0200

    Use %{_rpmconfigdir} throughout configuration macros and such

commit eac0666a1f53ac1192f6962c94274c7c845fd236
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 22 17:21:33 2008 +0200

    Set _rpmconfigdir macro on initialization

commit 59d5892627638ed9f04fab74cb13592cd4b34e1c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 22 17:12:54 2008 +0200

    Use rpmConfigDir() instead of hardwired path for rpmrc + macro locations

commit a8abeef91f0669e726e0b9ab398d56cc38efa3c6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 22 16:44:33 2008 +0200

    Use rpmConfigDir() instead of hardwired value in Lua initialization
    - userconfig is actually unused

commit 43b08986a8aa0464566716f6fe0d510e96cd77bf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 22 18:08:40 2008 +0200

    Use rpmConfigDir() for finding our popt alias file
    - this allows popt aliases to be reliably available in tests
    - why do we have the popt initialization done twice?

commit da83a562e1fd81402a39b7c3edfdc413fa860d8c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 22 18:06:31 2008 +0200

    Add function to retrieve rpm config base directory
    - one-shot to determine configuration base directory path from
      RPM_CONFIGDIR environement or build-time default
    - rpmfileutil is a bit strange place, this would really belong to librpm
      but Lua initialization needs the path so...

commit 86645ffd4f45a799bc904df56ca9b2f48641d245
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 22 15:37:00 2008 +0200

    Nuke leftover reference to brp-sparc64-linux
    - missed from commit cbbb03e3a34d492a7163f5a02f77b7e806702b41

commit 53b14ecfea90116ce4fc1182cc645f77c4f49527
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 22 14:07:34 2008 +0200

    Add keywords to all tests
    - just initial groupings to have something there..

commit d3d3454a47da54991262b8a93711de56ed2fdffc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 22 13:45:17 2008 +0200

    Add basic tests for Lua's rpm and posix extensions

commit 4ba4e709efd8f88a7bb5a0db8121f57c937615d9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 22 13:26:43 2008 +0200

    Add banners for nicer test output

commit 19a41acb84dcb3c5e6381199f8385bfa4e48e803
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 22 13:20:33 2008 +0200

    A few more basic macro functionality tests

commit d6ea0224cc7a3e0c7dec7edfa8ba210ada9e6167
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 22 11:45:55 2008 +0200

    Add couple of tar build tests + another version of hello tarball

commit 1c57e1d337c00b475253c522461f689e911bc426
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 22 11:22:01 2008 +0200

    Add some (trivial) macro evaluation tests

commit a5ff12cc192f7a2a4bf0506006b50f7452be4cab
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 22 10:28:59 2008 +0200

    Add a couple of tests for installing non-existent files and such

commit 0bcdd93151c21b78ff64d05c05711b0bf79966c5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 22 01:15:20 2008 +0200

    Add some basic tests for config file behavior

commit 9844bfc9b776b00c56dd389bfe5fb4807b189d59
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 21 23:59:25 2008 +0200

    Add some basic tests for file conflict behavior

commit ab3e32ad97e1c2143560ceb1aafb7c7bb85cfdcc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 21 23:29:54 2008 +0200

    Add basic version comparison testing for install/upgrade

commit 357a7170fa04b5e28c75f4669689cd727b1cfd20
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 21 23:28:21 2008 +0200

    Simplify the run functions, fix --define behavior

commit 08c81de2ee753778763810add5221e9162d925cb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 21 17:48:27 2008 +0200

    Hackery to get rpm install tests to "work" with fakechroot
    - fakechroot throws it's own curveballs into the already mixed up thing,
      copy things to the test dir and set chroot base to that to avoid
      confusion with cwd and such
    - klunky and ugly but at least the tests pass now when running non-root

commit db6c332c49885163de3b83e200e0b40dcac05dc3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 21 17:39:39 2008 +0200

    Add + use wrapper function for calling rpm commands in tests
    - avoids having to define --rcfile etc over and over again

commit ddf59da4183f578cf1b7238892776bd64340282f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 21 16:13:03 2008 +0200

    Kill the last remaining varprefix
    - we're just mangling varprefix to localstatedir anyway, might as well
      just avoid that little extra trouble...

commit 3a03a2fc4617efd8fb9f90d5279a10f0d4df1ef3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 21 13:28:35 2008 +0200

    Define TOPDIR once in atlocal.in

commit 50e59e92dccdb246e07860bf20e38c5a586b67ee
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 21 12:02:17 2008 +0200

    These tests are expected to succeed

commit 31475056f8f03ff983ad4f640330865d7d9ef701
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 21 17:59:03 2008 +0200

    Rpm is supposed to create build tree by itself, don't help it

commit 217e5700c0cd76cfce32a50a11d7cb8b719dd446
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 21 10:04:41 2008 +0200

    Fix build tree creation with relative paths for %_topdir and friends
    - rpmtsRootDir() gives "/" even if root not actually set which makes sense
      for most operations but not when we're creating build directories

commit 5e58eeb2d7d0fa301785e5299b3f92da9aa64dc2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 21 09:50:53 2008 +0200

    Permit relative paths in rpmioMkpath() and rpmMkdirs()

commit 19c5eb1ed89a103f9b16e04bf9adbf2f747d6146
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 21 08:57:12 2008 +0200

    Update --querytags test vrt FILECAPS

commit f9ed52c0c7a29aa64bc44f94f7b102775ce4be44
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 20 11:47:52 2008 +0200

    fpCacheCreate() already doubles the size-hint, dont do it twice

commit 2a1e88d27e8c1cecfe4fadff15f04c87c31c689a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 20 10:26:22 2008 +0200

    Use per-rpmfi string cache for symlink targets
    - few files are actually symlinks so this is very sparse data
    - per-rpmfi instead of global as link target names are not that likely
      to be shared across packages
    - another few megs saved...

commit 5487213d18c356512655e93b7ee246d2895198c0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 20 10:15:08 2008 +0200

    Add functions to create and free string cache(s)

commit 325198377df0be04a801d4f4c10dc698ce9d6c38
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 20 09:53:27 2008 +0200

    Use global cache for file languages too
    - there's only a very limited number of language names ever present,
      and most files don't have languages attached at all
    - another few megs saved on memory

commit e5d97430b62ed01bc3311de83eccdf48bb20a8cf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 20 09:40:46 2008 +0200

    Use rpmfiFLangs() instead of direct access to structure in skipFiles()

commit 0c4aa20a0a2ed4e8316a7f468da4e249a19aed41
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 20 09:36:47 2008 +0200

    Add rpmfiFLang() method for retrieving file language(s)
    - hardly anything needs outside install internals but it's harmless too...

commit 3bdac8f451062b5bd7a68205feae0dd0c71a336d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 20 09:26:12 2008 +0200

    Generalize the user+group string cache + helpers

commit f90f2f10155ebb010b3ff3119be31fd8600c1cb8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 19 18:17:56 2008 +0200

    Be more intelligent about storing file user and group names
    - User and group names contain an enormous amount of redundancy: typically
      most files and directories are owned by root:root. Store the each unique
      name just once into global cache and store indexes to the cache into
      rpmfi user + group arrays. This saves several megabytes of memory on
      a transaction of non-trivial size...
    - The cache implementation is dumb, but it's not noticeably slower than
      uncached version, copying megs of strings around is not cheap either.
      Easy enough to replace with a hash-table or such if performace becomes
      and issue...

commit 6fd987b7291fb1306297c193f4ab01bf6cd02618
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 19 16:07:15 2008 +0200

    Make fsmMapAttrs() use rpmfi methods instead of directly accessing
    - Save rpmfi iteration index on entry, restore on exit to avoid clashing
      with other users of the given file info set. For now...
    - Remove unnecessary re-initialization of uid and gid in case not found

commit 3b8ebf377bcab9a796304056a24cd3155fa23938
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 19 15:52:03 2008 +0200

    Oops, file modes are needed on erase too to avoid suffixes on directories

commit 36d4908efe80c72a48bdce31132c4b6331c9c094
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 19 15:46:02 2008 +0200

    Use matching type for fsm->digest to avoid unnecessary cast

commit 3eed102f115bac5dbf37b4866b3d2deaf5760ec8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 19 14:06:37 2008 +0200

    Erm erm, really use the rpmfi creation flags in rpmteNew()
    - this part was missing from 763cbf05b57da940d2972e0b448548481be400c6

commit 2a50bb83889aab7a2c65402fb7eabb548594cd0b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 19 13:04:07 2008 +0200

    Erm, really do lazy allocation of fi->replacedSizes
    - this part was missing from 9bde09b59969088c06eb6f5b8a2597b2e04bd306

commit 9bde09b59969088c06eb6f5b8a2597b2e04bd306
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 18 12:00:10 2008 +0200

    Add + use internal helper function for getting+setting rpmfi replaced sizes
    - lazy allocation on set, otherwise there's no replaced size - the getter
      deals with this transparently
    - saves a fair bit of memory, by no means everything has replaced files...

commit d0b088328549c6693b127decbc1b0b8c0513a539
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 19 11:59:53 2008 +0200

    Eliminate unnecessary fi->record field
    - transaction elements carry the db instance already and is available
      where needed

commit e3370d3b470cb19773d69c1fac5cf2a5766a0857
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 19 10:02:16 2008 +0200

    Avoid unnecessary mucking with rpmfi internals

commit 30fb610125d57bfdeab38a1811bf563bf264180e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 19 09:45:05 2008 +0200

    Refer to psm->te instead of fi->te in rpmpsmStage()
    - transaction element is passed to psm, might as well use it

commit 140f0c8c8339524968e9ca595fabe9a28b38f49f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 19 08:16:03 2008 +0200

    Restore SIGPIPE to default action for scriptlets (rhbz#471591)

commit 2ca4430dd7d9c295b97cd2362a91dfe3f8819ab3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 19 08:10:21 2008 +0200

    Updated Brazilian Portugese translation from Igor Pires Soares

commit 763cbf05b57da940d2972e0b448548481be400c6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 18 18:54:26 2008 +0200

    Use finer grained rpmfi creation flags for in rpmteNew()
    - also call the erase flags erase, not remove...

commit 78e8e7c02cefc2643159f333c3067fa77108ccf9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 18 18:36:45 2008 +0200

    Limit the data rpmfi loads in query
    - makes queries a little bit faster, hardly makes a difference on
      modern hardware but loading unused data seems silly anyhow...

commit 0e65e6e92cd3c0e6e796b6595bbf1ff709d54375
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 18 18:32:58 2008 +0200

    Add preliminary rpmfi flag groupings for base operations
    - just a start and probably not 100% accurate..

commit 219ba347365e226f87d2d489e105a52a798c0284
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 18 17:38:53 2008 +0200

    Add control bits for all (relevant) parts of rpmfi data
    - permit fine grained control over what's loaded into fi for what operations,
      non-trivial memory savings are possible
    - future possibilities - these reflect closely to rpm verify control
      bits, all we need is RPMFI_DISK or similar operation to pull the
      information from disk instead of header and then verify can simply
      compare two rpmfi's

commit 9444f2d7f5bf0ffcd81fac0e04f0e831ac276306
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 18 17:12:55 2008 +0200

    Remove useless default flags and action fields from rpmfi
    - fi->actions is always allocated so fi->action is never used for anything
    - fi->flags is just as unused and unnecessary

commit f41bd90c2ac156e459e7af368481471b701ed887
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 18 16:33:07 2008 +0200

    Nuke useless rpmfiFContext()
    - it only ever returns NULLs because there are no security contexts in
      headers (and having them in headers make no sense as the context varies
      based on whatever selinux policy happens to be active on the target
      system)

commit 937d814328caba4383b8c3c71265268bb9797840
Author: Florian Festi <ffesti@redhat.com>
Date:   Mon Nov 17 13:44:06 2008 +0100

    Fix indentiation

commit 73c17671de0bf017c9600504f11a1cdf79d0ce26
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Nov 18 11:23:26 2008 +0100

    Add AC_CONFIG_MACRO_DIR for the new libtool

commit dbad382930c65934dcd204212c373e7ac20ebd09
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 18 11:30:43 2008 +0200

    Don't bother loading FILELANGS for installed packages
    - only useful for anything when installing packages

commit 03d0f43f0bf4d2883a822277c7a9ae2936b3a699
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 18 06:58:13 2008 +0200

    Lazy allocation of FSM for file info sets
    - new more or less internal rpmfiFSM() to create FSM on demand, move
      mapflags calculation there

commit 61794cf2c0575de8d557b06df7e94865d7afe592
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 18 06:28:23 2008 +0200

    Remove now unnecessary fi->keep_header

commit 341aaa65150049ba88308ff692eafcd9834997f0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 18 06:26:02 2008 +0200

    Remember rpmfi creation flags

commit ff9733b6fd8cb7dc8afccf38833d26d0363aeef7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 18 11:22:26 2008 +0200

    Generate and insert SOURCERPM tag earlier
    - needed to make checking for source/binary package in rpmfiNew() reliable
      at build time

commit 8d6c4b8c95b59f5a71d90c582c2e98f5c7ed7b9d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 17 17:29:31 2008 +0200

    Add + use internal helper function for setting rpmfi file states
    - instead of directly accessing the array from fsm, use a helper function
      which lazily allocates the state array as needed and sets state
    - also fixes the silly case of non-installed packages showing their files
      as "normal", ie installed (now it shows "no state" as it's not relevant)

commit 4927190cf1d0a2ac902bdf5d7fe5742539c4bb37
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 17 16:12:04 2008 +0200

    Simplify rpmInstallLoadMacros()
    - fi argument unused, remove
    - no relevant return value, make it void
    - handle any proper data type for per-header macros

commit 936fc19176c7baec9871bf8fd5aa6a1d56e02d00
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 17 16:04:48 2008 +0200

    Avoid reconstructing half of rpmfi manually in rpmInstallSourcePackage()
    - similarly to commit 53ee682a491ceee4392815075cabbe58d0cc0c22, arrange
      the header to contain necessary paths before calling rpmfiNew()
    - avoid unnecessary accesses to rpmfi internals otherwise too

commit a919e1d631f50b29c5d917d9bfb65f7413a59782
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 17 14:21:38 2008 +0200

    Unnecessary local variable, result only needed once

commit 4181d7adadecf64102e620a23a697c756e3602c9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 15 18:44:00 2008 +0200

    Assume failure in rpmInstallSourcePackage()
    - avoids bunch of redundant assignments

commit 899dfb58927ec6e91014773430824462f4d0002e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 17 13:39:22 2008 +0200

    Calculate total file size correctly in build (mdvbz#45820, rhbz#247374)
    - delay total size calculation until the real file list is known, ie
      in genCpioListAndHeaders() where duplicates and excludes have been
      weeded out

commit 74ed5372f681a31987ba931d72ed5d8ae96f52d0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 17 12:44:57 2008 +0200

    Make depflagsFormat() accept any numeric type

commit 05a3d00162fe343002e378e42f6458139093ba52
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 17 12:42:53 2008 +0200

    Simplify xmlFormat() by taking advantage of rpmtdClass()

commit 8ef09ba46c4698ac5a00b496c9ab4a9596bfa31e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 17 12:41:24 2008 +0200

    Make fflagsFormat() accept any numeric type

commit 53dcad2f9d5e13e246775c3b052c025406100a0d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 17 12:39:13 2008 +0200

    Make permsFormat() work with any numeric type

commit 5580fe23d50b82971c3f0fb12bc613ac64fcdfb0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 17 12:37:42 2008 +0200

    Make triggertypeFormat() work for any numeric type

commit f31b790c31d017b9fbe75d1af822c8b438c0cd84
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 17 12:34:22 2008 +0200

    Make shescapeFormat() to accept any numeric type

commit 2b3380aa3455cd6aa3a86491c33c15648360f6fd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 17 12:29:01 2008 +0200

    Take advantage of rpmtdGetNumber() for date format specifiers
    - dateFormat() and dayFormat() now work for any numeric type

commit ee6c147a239feb6a5587758b19192f274bd0a529
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 17 12:23:19 2008 +0200

    Take advantage of rpmtdGetNumber() for numeric format specifiers
    - octalFormat() and hexFormat() now work for any number, not just int32,
      fixing rhbz#471820
    - use a common helper-function for both formats, they only differ by o/x

commit d97e357560ad9111b95da9bf2126a7bde3b4bac1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 17 12:16:51 2008 +0200

    Simplify stringFormat() by using rpmtdGetNumber()

commit 2db2c376b322bd2940b15600357e9f22d8a55e0c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 17 12:04:08 2008 +0200

    Add rpmtdGetNumber() for getting numeric values from tag container
    - returns the value (as opposed to pointer to, like the rpmtdGetUint32()
      and the like do) of any numeric type as uint64_t (largest supported
      integer type so everything can be converted to it)
    - handy when you don't really care what the internal presentation is
    - there's no rpmtdGetNextNumber() as there's no meaningful way to return
      end-of-iteration here

commit ed5306b0be4b8b0df205b066d2646b4e1dd93545
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 17 11:31:00 2008 +0200

    Introduce rpm tag "classes"
    - rpm tag data can be either numeric, strings or binary data, each with
      their own "subclasses" (different sized integers etc), add new
      enumeration for these
    - add rpmTagGetClass(), rpmtdClass() public functions for retrieving the
      base class of tag and container
    - useful for getting a basic idea how to handle tag/container data

commit 4fa662abd1d1f5b5f155a734d0e87aa7e7894ae3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 15 17:55:47 2008 +0200

    Use rpmpsmNew() + rpmpsmFree() instead of manually doing the same

commit 170f5bf7d6ce45142cb750d732a625c168c3f524
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 15 17:20:13 2008 +0200

    Eliminate bogus const from fi->apath

commit 32b347aa1f909618d739d611b6bb52cde12c3b30
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 15 16:00:36 2008 +0200

    Eliminate match iterator from psm struct
    - it's only needed in local scope anywhere, no point having it psm global

commit 53ee682a491ceee4392815075cabbe58d0cc0c22
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 15 14:22:47 2008 +0200

    Avoid having to reconstruct half of rpmfi manually in genCpioListAndHeader()
    - We need the on-disk paths for the cpio list in here, so instead of
      creating an rpmfi with wrong paths (for build purposes) and then
      painfully reconstructing half the fileinfo structures by hand, pass
      the on-disk path into rpmfiNew() to begin with and then switch the
      intended cpio path back into header.
    - Duplicates and excludes are already filtered out from the header filelist
      so walk that instead of spec filelist to avoid having to filter out
      duplicates and excludes, again.

commit 2d5b4e3bace86d9eea9024ca55de941fec2acc6f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 15 13:22:19 2008 +0200

    Lose useless fsm/rpmfi astriplen field
    - only "used" for debugging output, duh...

commit 21c34a30eb1d8de4a66a9b4d0ec5e1b38cac4568
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 15 12:07:44 2008 +0200

    Don't bother creating a dummy transaction element for rpmfi in build
    - nothing in the build path requires fi->te to be present, so don't
      muck with rpmte internals needlessly

commit 3632ad454b4415a0f73abdad920cfe2ed79d023c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Nov 15 11:50:09 2008 +0200

    Minor genCpioListAndHeader() cleanups
    - rpmfiNew() only cares about ts for install-time relocations, nothing
      to "fix" wrt that
    - we know fip is never NULL (this is a static function with whopping two
      callers), don't try to cover up for potential programmer errors

commit 2eb2569d4662a3f0f391da9f0b50641495d9f6db
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 14 22:45:42 2008 +0200

    Don't have to free what's not allocated to begin with...

commit 79225b74872717613c39ca23fbe48f521e4323a3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 14 22:44:07 2008 +0200

    Add more flags to control rpmfi creation
    - RPMFI_NOFILELANGS to avoid loading lang data (only useful on install)
    - RPMFI_NOFILEOWNER to avoid loading of user+group data

commit 2835e091ae72ba53b3fac478a1e7d080affa4cfe
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 14 17:16:42 2008 +0200

    Yank rpmfi state update out of rpmtsRun()
    - stuff it into internal rpmfiUpdateState() function to get it out of
      sight for now, the mechanism needs some...

commit ee59836299c5e1c293b64a826552cbfba5dcec12
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 14 13:05:41 2008 +0200

    Move open + close of files during install to separate functions
    - internal rpmteOpen() and rpmteClose() functions replacing copy-paste
      slop between rpmtsRun() and runTransScripts()
    - eliminates bunch of rpmte privacy violations
    - rpmtsRun() doesn't need the file descriptor for anything, might as well
      keep it hidden

commit 0b0fe33d3c086e9bc3d7caa351a8304a64f724f1
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Fri Nov 14 15:16:08 2008 +0100

    Allow setting custom parameters to patch (rhbz#471006)
    - added new _default_patch_flags macro with the rpm default
      "-s" parameter

commit e0e0a5e0352fcc8aebe5a4752d672399151aee3f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 14 09:47:43 2008 +0200

    Fix capability verification
    - the previous implementation was pretty busted up, cap_size() is the
      size of external representation, not sizeof(*cap_t) as pointed out by
      Andrew Morgan
    - check for cap_compare() in configure and use it if available (it's
      a Linux-specific extension to the posix draft and only very recent
      libcap has it)
    - if cap_compare() isn't available, use internal variant that does it
      the hard way by converting the capabilities to external presentation
      and comparing those

commit 9646a117c916022067fd07e4f6a7e25c5c9814de
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 13 13:22:03 2008 +0200

    Filter out mtime difference of shared files on verify (rhbz#426672, #444661)
    - Since we permit shared files to differ by mtime on installation, it makes
      no sense to consider them as verification failure either. This has been
      especially bad on multilib where timestamp differences make verify warning
      lights look like Las Vegas at night. If file mtime on disk and header
      differs, check if it's owned by more than one package and if so, ignore
      the timestamp difference.

commit ed8b3fb2c54aab97b616e83a24a3f65a36f9c0f2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 13 09:56:52 2008 +0200

    Move ts and preferred color fetching out of the loop
    - neither ts or preferred color changes per-file

commit ed9836dfba18ac021d8c083679db8d2a9e7c38a3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 12 21:58:56 2008 +0200

    Remove unused fsm subbuf field

commit 526c052dd28e8c8ff68c15c752f7349e721b9f01
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 12 19:36:58 2008 +0200

    Don't bother retrieving db offset for erase elements
    - not needed anymore, use -1 to mean "unused" and document dboffset
      to rpmtsAddEraseElement() as unused

commit c11c97b67cda698486cee26f0621231731c6ebcf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 12 19:33:12 2008 +0200

    Take advantage of headerGetInstance() in removePackage()
    - avoids an extra argument, making dboffset argument to
      rpmtsAddEraseElement() unused

commit 3cfac308be04a96a192afa8be8f6a7c08bfbdfba
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 12 19:24:58 2008 +0200

    Take advantage of headerGetInstance() in rpmte
    - rpmte db_instance and dboffset are exactly the same thing, unify
    - pull rpmte->db_instance from header directly, ignore dboffset argument
      to rpmteNew() (removing the arg would just break api needlessly, otoh
      it's unlikely that anything outside rpm actually uses rpmteNew())
    - eliminate the now unnecessary added/removed union from te struct
    - rpmteDBOffset() is now just an alias for rpmteDBInstance(), deprecate?

commit 0c707fe3707aae08ce6b00393ff625cf80547074
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 12 18:29:51 2008 +0200

    Grab rpmfi record from header on rpmfiNew()
    - avoids having to muck with it from rpmtsRun()

commit b9010a3510033bfd83f07b47eda1cfa269b38429
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 12 17:58:37 2008 +0200

    Add rpmdb record number to headers retrieved from rpmdb
    - set "instance" number on retrieval from rpmdb
    - add public headerGetInstance() function for retrieving the value
    - ported from rpm5.org, useful for number of things

commit 90cb50ff7abab62d99cfcda1f64f451aa2fe41cb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 11 22:37:12 2008 +0200

    Convert obvious bits of rpmtsRun() to use rpmfiFAction()
    - remaining access to fi->actions is the nasty part of copying things
      around to transfer state

commit 2f6e1bbb809c7e15b30b909db1a810087c7f94a0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 11 22:35:45 2008 +0200

    Convert skipFiles() to rpmfi(Set)FAction()

commit 2ee45ef12aaa958b78c56b6d83bb18aa6a477d15
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 11 22:32:23 2008 +0200

    Convert handleOverlappedFiles() to rpmfi(Set)Faction()

commit 715917ba3754833031b605ec5b8d1c220774dc76
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 11 22:19:29 2008 +0200

    Convert handleRmvdInstalledFiles() to rpmfiSetFAction()

commit ac2eb4f66d954fd9ae7f5b49e3dbc28b348d3657
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 11 22:17:10 2008 +0200

    Convert handleInstInstalledFiles() to rpmfi(Set)Faction()

commit 78df12924875d332a0f0f647ebc075929287c940
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 11 22:09:48 2008 +0200

    Add rpmfi get + set methods for file actions

commit 23ce4966bbfab43974e874d20adb260bfb634ec9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 11 21:06:32 2008 +0200

    Remove unnecessary initialization
    - rpmfiNew() zeroes out everything, including striplen + astriplen

commit a4e345455d817ce6c56940600d652ee3ae0aa5d5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 11 21:01:02 2008 +0200

    Move file action calculations to rpmfiNew() where possible
    - for build operations and src.rpm install this is known

commit fbc06059cbf7e3f7c6330e46b38bf68850ff1e26
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 11 20:48:35 2008 +0200

    Cpio flags are only relevant for fsm, get them out of rpmfi

commit 0d168d173a4c361cc1f548304743cb63bdc5946d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 11 14:50:53 2008 +0200

    Figure out cpio mapflags in rpmfiNew()
    - all the necessary info is there, no need to litter the logic all over
      the place

commit 8e3b5bfc894c5e5b0e3dd41d49fc62db3eaef7aa
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 11 13:53:31 2008 +0200

    Push archivePos from rpmfi to fsm
    - archive position is only meaningful in fsm context anyway

commit f553d2f0cd99678528ac73d1258cacb9c8dc3bf5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 11 17:35:42 2008 +0200

    Adjust rpmfi directory count while messing with it in build
    - otherwise we can miss some entries in rpmfiFN() max calculation and
      blow up in entertaining ways

commit d535818f425ae7ef0fd53203a18751cb19367e1c
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Nov 10 15:52:29 2008 +0100

    Add the post-build package checker feature in rpmbuild (upstream ticket #2)
    - allow to run helper programs to check:
      - generated SRPM
      - each generated RPM after it is successfully written
      - whole successfully written binary package set
    - it is possible to use it for sanity checks

commit 34cda292e3523d3ca2d5fee7c71cce813f886357
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Nov 10 15:46:43 2008 +0100

    Add macros for the post-build package checkers feature (upstream ticket #2)
    - _build_pkgcheck: helper that is ran for every binary RPM successfully
      written
    - _build_pkgcheck_set: helper ran for the whole binary package set
    - _build_pkgcheck_srpm: helper ran for SRPM
    - _nonzero_exit_pkgcheck_terminate_build: indicates whether the build
      should fail if helper returns non-zero exit status

commit 6404a00063bb3865f2332990fc88b90107d41306
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 7 12:03:54 2008 +0200

    Reset return values to zero + NULL's on not found in hash GetEntry()
    - avoids having to separately check for return value in some cases and
      accidents from not resetting the values in caller
    - in line with headerGet() behavior

commit 844cec311f3690b51033dd25d10d9ade22e2e7f2
Author: Florian Festi <ffesti@redhat.com>
Date:   Wed Nov 5 15:23:19 2008 +0100

    Rename doLookup to fpLookup to get rid of one function just calling the other

commit 5f01f7e02bb99be176add41488a6943dc17d6df4
Author: Florian Festi <ffesti@redhat.com>
Date:   Fri Nov 7 10:53:55 2008 +0100

    cleanup fpLookupSubdir
     * rename end -> endsubdir
     * fix assignment mess
     * fix memory leak (no longer free current_fp.subDir which may be NULL)
     * set endsubdir to NULL while not valid

commit e542ec8682f2929f31efb85d04ac4667d6a85ff0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 7 11:52:24 2008 +0200

    Reverse hash GetEntry() return code vs found/not found
    - be in line with hash HasEntry() and other similar functions
    - internal-only function with just one place actually checking the return

commit e30ad6d1af8d425fd8390d67965a136e50f3ff8b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 6 10:03:10 2008 +0200

    rpmsqAction() proto needs SA_SIGINFO conditionals too (ticket #4)

commit f2a7667482475bbb9bf3acbcf57b755fee987a83
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 5 20:37:48 2008 +0200

    Nuke the ugly hardwired "known arch" list from label parsing
    - we have list of known archs in rpmrc, might as well use it
    - avoids having to patch & recompile rpm just to have rpmtsInitIterator()
      recognize new archs when it's runtime configurable otherwise
    - checking against compatible arch would be saner except then people could
      get into even more trouble using --ignorearch...

commit a9169c46abddfe620c6ae83f3064b50c8bb35852
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 5 15:44:35 2008 +0200

    Calculate filename buffer len lazily in rpmfiFN() as needed
    - avoids having to separately calculate in genCpioListAndHeader()
    - avoids unnecessary fnlen field in rpmfi struct
    - avoids having to calculate it at all if rpmfiFN() is never called

commit 6bd6677f86ba130cd4bc952c37fa848450f1920a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 5 15:24:08 2008 +0200

    Remove yet more unused fields from rpmfi struct

commit b43998c374abae3b689a3698a2006fdafac1146a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 5 15:18:43 2008 +0200

    Remove fperms and dperms from rpmfi struct
    - only fsm internals need and just use straightforward defaults, move
      the defaults there...

commit 52d53b8bf91c8a9ed28257ec9ed1b65fd5b98fbc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 5 14:30:01 2008 +0200

    Use cpioMapFlags type consistently for map flags

commit c0a6f7b14169adb0f3961f6ccda8ca49eb87fead
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 5 14:10:16 2008 +0200

    Stop messing with fi->fstates allocations in psm
    - it's alloced and freed by rpmfiNew()/Free() always anyway

commit ed412a0e2f8538ab952908b06f24d769f1fd1fb6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 5 13:17:50 2008 +0200

    Kill off per-file cpio map flags
    - nothing uses/needs them, one less murky fi->foo allocation less
    - apparently a leftover from initial multilib attempts long long ago

commit cbbae4cdb102ac9de4023e6e35ddc668eae09ca9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 5 12:45:03 2008 +0200

    Remove unused variable

commit 455a4cd65269272a8e11a50d1242b36b2163e2de
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 5 12:38:16 2008 +0200

    Remove useless rpmfi->gid and uid "defaults"
    - they were only ever set on src.rpm installation yet overridden by
      package contents unless src.rpm didn't contain RPMTAG_FILEUSERNAME
      (and GROUP). Which would mean rpm 1.0 packages or thereabouts, or a
      malformed package. In that case we just fall back to root uid,
      it hardly matters

commit 1f5d9dff73cd8ea0ee0c6ff1e843c99c78f9c38d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 5 11:49:33 2008 +0200

    fi->actions is already initialized earlier in rpmfiNew()

commit 4d786000b3d92548eb1822221271d7e02ca93a1b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 5 11:47:29 2008 +0200

    fi->actions cannot be non-NULL in rpmfiNew()

commit 2102efa0b1136c5c19da75f670c9c0436a187754
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 5 09:36:20 2008 +0200

    Dead code removal

commit 4f7f014537d0b4e20a63785de68dd74bdbcd2446
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 4 14:53:11 2008 +0200

    Stop messing around with rpmfi user+group names from psm
    - these are loaded and freed by rpmfiNew/Free() as necessary

commit e07d37aebc1ab6213507375b8916f5259cf4dc0f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 4 13:33:27 2008 +0200

    Fix segfault on relocation (rhbz#468319)
    - a bogus free() / thinko

commit 9a509e9f1acc27fdce547a47513e9d083a17b6d4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 4 11:36:01 2008 +0200

    Avoid leaking C99 options into stock %{__cc} macro

commit faaa0305f5593bd3165e444d4ec4516490b4610e
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Nov 4 10:09:31 2008 +0100

    Return proper exit code when exec*() functions fail
    - the correct exit code should be 127 accoring to bash(1),
      we want to keep the same exit code behaviour as bash for rpm

commit 73f61ecf05ff1679219e49bfd029a69da66bf35e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 4 10:28:20 2008 +0200

    Fix building on systems where SA_SIGINFO isn't available (ticket #4)
    - patch from Adam Tkac

commit dc06dca3b99f809f826f25565ace384b54f3b488
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 4 09:52:19 2008 +0200

    Use more portable file access mode tests (ticket #5)
    - don't make assumptions about how O_RDWR and friends are defined, use the
      portable way of masking flags with O_ACCMODE and explicit comparison
      against the mode we're interested in

commit 759b73a4293e56209e922bf58d7028234e854338
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 3 14:13:45 2008 +0200

    Ensure PATH_MAX is defined (ticket #3) Adam Tkac
    - POSIX doesn't require PATH_MAX to be defined as constant, and Hurd
      doesn't define it...
    - deal with it just once in system.h for now, the proper fix is to get
      rid of needing it at all (ie add a mallocing realpath() clone to librpmio)
    - patch from Adam Tkac

commit b7678c0a40898debdf7396c84967411efbfaf345
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 31 17:45:43 2008 +0200

    Drop unnecessary trailing slash in pkgconfig includedir

commit 81660b26bc8e2c03f55cf928955f7ed84e00785a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 31 10:17:07 2008 +0200

    Adjust for file output change (rhbz#468129)
    - file util used reported file details along with mime type, newer ones don't

commit 05c547a30d966d03b9bda37343c7dcc3638b6be0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 30 14:12:18 2008 +0200

    Kill of scareMem / keep header flag from rpmds
    - unnecessary complication which doesn't benefit anything (anymore)

commit ce33decefa68cdde3ed332c629dc661defc963d1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 30 14:07:29 2008 +0200

    Change all in-tree rpmdsNew() uses to non-scaremem
    - shouldn't affect anything at all

commit 6c91bf043776319bc0b78852ecf65a56501c60e1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 30 14:03:34 2008 +0200

    Remove unnecessary scaremem flags, hardwired to no scaremem anyway
    - no functional changes

commit a19cc0d1fe7a5113d550d07ba70501658034f161
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 30 13:46:26 2008 +0200

    Save bunch of memory by not loading file classes in ts elements
    - install/remove doesn't need or use file class information at all,
      don't bother loading it, saves a few megs of memory on large transactions

commit 0cf0a7bb68d6f455a2e1bc4918a7a2509eb0055a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 30 13:43:44 2008 +0200

    Add couple of new rpmfi flags for file classes and file depends

commit aabdfb8866a1e3a86fa91efffe2c170d6e808570
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 29 15:23:49 2008 +0200

    Move rpmfi color union calculation out of rpmfiNew()
    - nothing at all uses the calculated value or rpmfiColor() call, take
      the unnecessary calculation out of common path into rpmfiColor()

commit eef41b6dff424992785ff67a22244d62454d796f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 29 15:06:37 2008 +0200

    Remove unused package color calculations from 2003

commit cff6230a8c55b6cf925b7e33fb04f26a7d284d82
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 29 14:34:36 2008 +0200

    Don't bother fetching and storing pre- and posttrans scripts in rpmfi
    - the psm script machinery requires the full header to do anything at
      all, so the script + scriptprog were only used to check if the package
      *has* such scripts, a single integer will do just fine there thank you

commit f01ed5459deb3b3cf5c1bd4077f89b02533071f0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 29 14:09:27 2008 +0200

    Convert in-tree users of rpmfiNew() to use flag names
    - easier to grep for than scareMem, preparing for further flags

commit 8f41e940abe6950c807f530468adeb24049271ca
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 29 14:08:12 2008 +0200

    Turn rpmfiNew() scareMem argument into bitfield of flags
    - preserve behavior with 0/1
    - more room for controlling what gets loaded into file info set, not
      everything is needed at all times

commit b24dc850a50585efa5c6c4cef245f95cb568dd8d
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Oct 29 13:38:16 2008 +0100

    Set timezone to UTC every time to avoid local timezone confusion

commit e7c604daddacf5f5ba3a25f7639eca9d7b0f5eb6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 29 12:09:27 2008 +0200

    Add --filecaps popt alias for looking at file capabilities
    - doesn't need much as the capabilities are stored in human readable format

commit db1f9af5e2a4443e64ce10112a9553204bab7f4e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 29 11:53:31 2008 +0200

    Add proper file capability verification support
    - at least for now, no capabilities is treated like empty capability set
      which are not exactly the same thing but probably sufficient for our
      purposes

commit 97ab15cc9eadc1aab563b87a0c92d559cd9e9a41
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 29 11:49:38 2008 +0200

    Add fsm and rpmfi machinery for file capabilities
    - new rpmfiFCaps() API to retrieve the info from rpmfi set
    - fsm internals quite similar to selinux handling
    - plenty of #ifdef's, another possibility would be adding cap_foo dummies to
      system.h like for selinux

commit a792c55ffae23f483121401fe4ca6724abc57bb5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 29 10:49:25 2008 +0200

    Start adding POSIX 1.e draft file capability support for real
    - Parse %caps() from spec filelists, making best-guess verification of
      capability string sanity by passing to cap_from_text()
    - The posix draft specifies capability export presentation through
      cap_copy_ext() which would be fine, except that we don't have support for
      arrays of binary data. So we simply store the textual representation
      of the capabilities in a string array which we do have.
    - Only add capability tag on packages which actually have capabilities to
      avoid unnecessary header bloat.
    - Add a new rpmlib() dependency for file capabilities, packages relying
      on file capabilities wont work correctly unless the capabilities are set.
      To be fully correct, support for on-filesystem file capabilities should
      be checked at runtime, as this depends on kernel versions and such...

commit daf28d3bf452a1d8daa8f5a70b860ce91b76fffd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 29 10:39:00 2008 +0200

    Don't try to verify capabilities of directories and non-regular files

commit 5d7cd0ab2d814e5d2a0803bede9c2d0bb09f2709
Author: Florian Festi <ffesti@redhat.com>
Date:   Mon Oct 27 13:08:39 2008 +0100

    fix compiler warning
    by removing unnesseccary return address for the hash key

commit f4b5bd55db623f83d75d0d60432f48cf0a7d2ef5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Oct 26 14:18:17 2008 +0200

    Beginnings of acl (POSIX.1e draft 17) verification support
    - if built with --with-acl, check that files and directories don't have
      any extra acl's set
    - for now, any acl beyond the regular unix permission set is reported as
      file mode difference as the acl's cannot have been set by rpm itself
    - patch from Andreas Gruenbacher, modified to use libacl instead of raw
      xattrs for portability, BUT atm this uses non-portable acl_equiv_mode()
      Linux libacl extension, the posix draft doesn't seem to have much in
      the way of comparing entries :-/

commit 770a0f34c90b59460372da181895b3a2802113f1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Oct 26 14:17:16 2008 +0200

    Beginnings of file capability (POSIX.1e draft 15) verification support
    - add minimal bits and pieces to check for capabilities in files on verify
    - for now, any capability set is a verification failure as the capability
      cannot have been set by rpm itself
    - patch from Andreas Gruenbacher, modified to use libcap instead of raw
      xattrs for portability

commit ca40fafedb504ff0195e347886e96a0268b3f5db
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Fri Oct 24 12:32:31 2008 +0200

    Remove timezone autoconf checks and system.h reference
    - timezone is no more used so it's no more needed

commit aeb672ca42cf0cce6d9f3d302f4d04513fc673c3
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Fri Oct 24 12:31:02 2008 +0200

    Fix "timewarp" bug while listing changelog via "rpm -q --changelog"
    - date is no more one day in advance than written in changelog

commit ccbb618bfc2bb2a74674b9e49c5637abb34515d9
Author: Florian Festi <ffesti@redhat.com>
Date:   Thu Oct 16 12:13:13 2008 +0200

    Use better cache sizes

commit 48554042f434fd75452e752f1156a652ffe71c8b
Author: Florian Festi <ffesti@redhat.com>
Date:   Thu Oct 16 12:07:44 2008 +0200

    Improve finger print hash function to avoid degeneration of hash tables

commit ecc8ff93740f1cc733350016b9b4304c4f14e9d6
Author: Florian Festi <ffesti@redhat.com>
Date:   Thu Oct 16 09:59:54 2008 +0200

    Switch to a better hash functions for strings: Jenkins One-at-a-time hash

commit af80c3de47278c0631f1add3da56758343fafe44
Author: Florian Festi <ffesti@redhat.com>
Date:   Thu Aug 28 16:06:36 2008 +0200

    Use only one, global fingerPrintCache to save a lot of stat calls

commit af3464a053ecb0b56cc5af494ea22955fb350757
Author: Florian Festi <ffesti@redhat.com>
Date:   Wed Oct 22 15:17:49 2008 +0200

    symlinks

commit c6ccc90d7fef0f1b65e4bf5b77d5b800d4b53ffd
Author: Florian Festi <ffesti@redhat.com>
Date:   Wed Oct 22 14:08:49 2008 +0200

    Detect to be installed symlinks for file fingerprinting (file conflicts)

commit 9766e4492cb7d90f0c60804f3bbbbc7f8ef01a7e
Author: Florian Festi <ffesti@redhat.com>
Date:   Wed Oct 22 14:03:52 2008 +0200

    Fix path problem in doLookup

commit eecfbcbeb0b102c4518c5576e88d48c039ac635f
Author: Florian Festi <ffesti@redhat.com>
Date:   Tue Aug 12 21:19:11 2008 +0200

    Delete findFps foobar
    including fpsCompare and the fpsdebug opt

commit c65019affca0988705ea4b8b9b38adbeca6ac98f
Author: Florian Festi <ffesti@redhat.com>
Date:   Tue Oct 21 15:45:09 2008 +0200

    Change both hashes to a typed version

commit 832909b4a7f093f6ab223850dad892223a71ff80
Author: Florian Festi <ffesti@redhat.com>
Date:   Fri Oct 17 16:18:45 2008 +0200

    Make rpmhash a generic datatype using macros and includes

commit 6a8c221ce6a890d6b23daed92669eb90e80ec2af
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 24 08:05:00 2008 +0300

    Updated Serbian translations from Miloš Komarčević

commit a4825e488d44e01803c377ebc4fc054adcf42b3a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 23 07:47:25 2008 +0300

    Make noarch subpackages actually work
    - noarch subpackages get their arch tag from parsePreamble() already,
      don't stomp over it later in parseSpec()

commit be651d98769ec95c0fcc2dee015a40c68b83271a
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Oct 22 09:28:38 2008 +0200

    Increase IEEE-compliance of floating point code for SH platform
    - thanks to CHIKAMA Masaki

commit 02274eecfc98a83b53b091d2fd889cf7f9393b29
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Oct 22 09:22:05 2008 +0200

    Apply patch for sh arch from CHIKAMA Masaki

commit bb74348bfb7ddfd333d94f855eabcc8658f0e9e6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 22 09:58:10 2008 +0300

    Skip adding identical NEVR to transaction on install too (rhbz#467822)
    - Installing several different versions of the same package in single
      transactions is permitted for "pure" installs, whereas on upgrade only
      one version is allowed. Identical NEVR got correctly checked and filtered
      out on upgrade but on install identical NEVR was not catched. Check
      for equal and newer separately...

commit cf89739282feadb36ef50d0b8d8c4e293cd0c24c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 22 09:25:30 2008 +0300

    Fix memleak in rpmtsAddInstallElement()
    - previously any of the "skipping/replacing already added" checks matching
      would leak memory due to early jump to exit

commit a8ba86f05bfdff3eda650d49a4aaab7bf0b4abf5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 20 11:46:42 2008 +0300

    setenv() + unsetenv() are always available in rpm

commit f7707ad3ff279fe09e8885b3a2ea2b01f629950b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 20 11:32:43 2008 +0300

    Add posix.mkstemp() to Lua posix lib
    - lifted from apt-rpm

commit d34646ab10a992092c9d2662cbbe573945c22332
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Oct 19 20:08:35 2008 +0300

    Trash more unused junk from installplatform

commit b4d6e18a9837144864fb608356acab38930a675a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Oct 19 19:59:57 2008 +0300

    Trash unused MULTILIBNO hackery in installplatform

commit 49c515179ac0d34c2bb43bf242b8eb27ff6c11b2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Oct 19 01:53:03 2008 +0300

    Permit tab as whitespace between macro arguments (rhbz#467567)
    - another dumb regression from macro handling cleanups

commit df080a1f6f313f83e41a94ff4deba1216f196a01
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Oct 18 19:23:56 2008 +0300

    Clean up hg leftovers

commit cbbb03e3a34d492a7163f5a02f77b7e806702b41
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 17 15:00:01 2008 +0300

    Trash the horrible brp-sparc64-linux hack from the ancient past
    - kludgery like this might've been necessary back in 2000 to get started
      with multilib setup but it certainly isn't needed now

commit 927f384a9bc058eb6f954e93cc515cc1293fd2a3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 17 08:11:10 2008 +0300

    Kick out the dumb defaultdocdir logic from installplatform
    - defaultdocdir is simply %{_datadir}/doc

commit df8b431be808b73b92cab5fd62637be6dcd72991
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 16 17:09:28 2008 +0300

    Bump version

commit c3c8dc476e68e385c9ac599f8651959a4fdb41d0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 16 16:11:15 2008 +0300

    Updated German translation from Fabian Affolter

commit c00bd2d29c223c832bfca68411b8fdbafee15371
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 16 16:08:18 2008 +0300

    Updated Polish translation from Piotr Drąg

commit 0b31ec61cc9b53eaf56d3f51058805be8b0b214a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 14 16:32:58 2008 +0300

    Add configure detection for SuSE too
    - distro detection in configure is pretty futile but while we still have it...

commit 085dcab72fc44cab5b1a7d810071bb8c0ea0afc2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 14 15:49:57 2008 +0300

    Move luaext include fiddling out of toplevel configure

commit bedeef1edaa4f2af99fbaecc6adfab75de1cd03a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 14 15:44:22 2008 +0300

    Add lua to pkgconfig private libs

commit fc9723a64c10e2212b1b296c2c94a0ff83e3cf05
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 14 15:41:56 2008 +0300

    Reuse toplevel WITH_LUA_LIB setting instead of manual setting

commit a1b05720b0e8219b26a81b8e2429226f094075cc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 14 15:36:23 2008 +0300

    Sanitize libluaext linkage

commit b6a0fca64d1d3b5331dffd78b7e75b4d5d86dfe4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 14 15:31:24 2008 +0300

    Fix build without Lua

commit b03953e551e4e3597b22cca71bfb583162187e9d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 14 14:46:33 2008 +0300

    Tweak autogen a bit more
    - lose hardcoded paths based on platform
    - add --rpmconfigure that can be used if rpm is already present

commit 9fe51e6e2844a5b1d7d4d4116f15449d0f327b35
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 14 14:24:30 2008 +0300

    Rip silly "usage" message from autogen

commit 9397bb61baeebcafda97c61c81dfbc1c4af96c17
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 14 14:17:14 2008 +0300

    Kick out ancient and unused LDFLAGS_STATIC configure.ac-hackery

commit 3cc7a1ef99d5832b35239afca0c3a978d1984ae5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 14 13:34:14 2008 +0300

    Yank out ancient convertdb1

commit c8ce024aa88a150ded321de8844049341d6ed04c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 14 12:46:55 2008 +0300

    Fix up some fuzz in finnish translation

commit 195f50d3cdf611f24d40c6a63b5f928d791570dc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 14 12:39:25 2008 +0300

    Adjust translations wrt the rpmps format changes

commit e2749795c3ae35f99092634fd859cba93f23156d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 14 12:25:01 2008 +0300

    Use PRIu64 for 64bit value formatting instead of casting

commit 19027cb53f3e559cca3d0ddc4e193fab37840d96
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 14 12:17:01 2008 +0300

    Avoid overflow on memcpy()

commit ef9744ff76f1ceed7645731bb6c102845eb48b79
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 14 11:56:15 2008 +0300

    Need to mask the return type from rpmTagGetType() return value

commit 055893bb4c638be34c3f6d98f6a6086718465a61
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 14 11:55:05 2008 +0300

    Error out on unknown data types in sqlite backend too

commit dcfe216da88af522ba5af55c4855dc8473c640cd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 14 09:44:58 2008 +0300

    Update translations for a change...

commit b7492daf836d1d61dd65ea586e3b1615da4c9c8d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 14 09:44:32 2008 +0300

    Add a few missing files to POTFILES.in (rhbz#466834)

commit 43a8cc7e102c97afd61a55e252ecce5a59f643da
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 13 14:45:45 2008 +0300

    Make sure dbiTags are loaded on db rebuild & remove
    - rpmdbMoveDatabase() and rpmdbRemoveDatabase() don't operate on an open
      database so there's no guarantee that dbiTags are loaded when they are
      called unless explicitly requested

commit d28c8fdfb82e45e63a82df7a0e6bc74af4a3cd87
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 13 13:25:17 2008 +0300

    Honor default nopromote setting everywhere in rpmds (rhbz#466503)
    - Of four different ways to create a dependency set, only rpmdsNew()
      explicitly set ds->nopromote to the expected value from global
      _rpmds_nopromote defaulting to on. All others fell back to legacy
      "epoch promotion" behavior inherited from xcalloc()...

commit 504e8231cbb379a0b3f41d7bcfb0bd4e3ff16f50
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 10 15:59:38 2008 +0300

    Yank out legacy.h, make the conversion functions static

commit 54f8c5891f92573ebd0ec5f478c6be25b6c4349d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 10 15:54:53 2008 +0300

    Convert in-tree users to headerConvert()

commit 56f3545c68796b1d1fb95ae4609395bc269e242a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 10 15:53:12 2008 +0300

    Add headerConvert() "proxy" for performing various conversions on header data

commit 83c4098e4f8affdb7f52aa5ed14ac15d7785f5b9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 10 14:43:39 2008 +0300

    And out goes rpmMkdirPath()

commit aa509e3c5e68ed905a80743bd3e8f3e26aa67590
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 10 14:42:34 2008 +0300

    Use rpmMkdirs() to create the entire build tree in buildForTarget()
    - also move the build tree generation earlier to have it automatically
      created for tarbuilds too

commit c17dffb6fbf8c123516ed1ca06a35beca1c7d8c7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 10 14:37:57 2008 +0300

    Use rpmMkdirs() instead of rpmMkdirPath() in rpmInstallSourcePackage()

commit a7ece3bc4ac90b7bc68522aa748e138b78ce4179
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 10 14:28:41 2008 +0300

    Add rpmMkdirs() to librpmio
    - allows creating bunch of directories including parents at once

commit 1566caf0674b372a7b9ea481a90df0f2e8e7aec9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 9 19:53:05 2008 +0300

    Get db vectors out of exported ABI

commit bd70d0e36c620a086af56d78cefdf7577a3457ba
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 9 16:17:53 2008 +0300

    rstrcasecmp() and rstrncasecmp() are "pure" functions

commit eb9649dfa9676fcf8117f799db8578b72c805673
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 9 16:14:23 2008 +0300

    Mark risalpha() and friends as "const functions"
    - they only process integers that are fed to them and no outside data

commit 505aea7fb178a174217181e02d765f41d3dc377a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 9 15:59:26 2008 +0300

    Remove silly double definition of headerGetLangs()

commit 5a76671724e989f9b841211c038272288825ce2f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 9 15:57:30 2008 +0300

    Mark htonll() as internal

commit 65c85b38e882e25aaca542f353e75334f8580280
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 9 15:52:34 2008 +0300

    Mark cpio functions internal
    - cpioStrerror() is used by build code so that remains visible, ugh

commit 0d48ac0aaa7cfda5b9e6cb1915a8588e410f2175
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 9 15:48:46 2008 +0300

    Mark fingerprinting stuff internal

commit 5589c3a81f56bbe27e0bff5eb27206e4a3aa1eee
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 9 15:47:09 2008 +0300

    Mark hash table functions as internal
    - we're not a hash-library afterall..

commit 2ec41237f8c809c9780474f28e59fbacb70d27e2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 9 15:43:00 2008 +0300

    Mark rpmdb internals as internals, including the "mergesort" gem

commit 550e156c62ff3dc397bfc167908fcbd1e523d922
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 9 15:37:59 2008 +0300

    Mark psm functions as internal

commit 7681f80e2512d60cd0b057a136fe38a3fb6575e6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 9 15:35:36 2008 +0300

    Mark various fsm* bits internal
    - build code uses fsmSetup() and fsmTearDown(), those need to remain
      visible for now

commit fb992745595893789a02a465833299493313c0a7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 9 15:32:04 2008 +0300

    Mark uidToUname() and friends as internal

commit be50cb5bf796719f23a446c807d2a79dbc8588a0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 9 15:21:15 2008 +0300

    Add RPM_GNUC_INTERNAL macro
    - allows portable use of __attribute__((visibility("hidden")))

commit c7bee0c3e6576fc78d7ab3ff64eeef3dda91bf02
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 8 18:26:58 2008 +0300

    Unknown tag is an error

commit 9e89ea52f5b3ac84bd7e63a20dba004a7ddd8acc
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Oct 7 22:06:28 2008 +0200

    make %find_lang --with-kde work (rhbz#466009)
    - patch from Rex Dieter

commit 30a975b9839b3cb15a8cd9c1ead454464523028a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 6 18:50:30 2008 +0300

    Malformed query query items (pkid's and such) are errors

commit bcda17901e38adacad42226164057f4b357ce9a6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 6 18:49:33 2008 +0300

    Make "not an rpm or manifest" message an error, not just notice

commit 9c51c44577ddee0c4d822d12b79b29f5a9bcdb21
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 6 18:41:09 2008 +0300

    Consistent log level for "unverifiable signature" message
    - all the other ones in package are RPMLOG_ERR

commit 377239b7dfa3e228748ef33c0bcec4213ac16f7b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Oct 4 15:16:38 2008 +0300

    Failure to rebuild database is worth at least a warning

commit bc1ac82d8d59f29561fe25e50292d8f0f8950817
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Oct 4 11:19:26 2008 +0300

    Restore permissions of a signed rpm to that of the original (rhbz#465586)

commit ffb886048c55b3a668d873bde7054571763c1058
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 3 13:00:33 2008 +0300

    Similar semantics for <lua> scriptlets
    - missing <lua> support is always an error
    - otherwise permit non-pre scriptlets to fail

commit 3e687d09717ee8408a781d221f0ce4506b3cef1b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 3 12:49:29 2008 +0300

    Always perform callback notify of lua scriptlet errors too

commit 17df5ad61397fb12bb2cc9a94798de97266afd9e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 3 12:13:21 2008 +0300

    Remove unnecessary {} block + indentation, no functional changes

commit 9cfb380cc4529ea6b9314cc67819069c8e9c1a23
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 3 12:04:19 2008 +0300

    Filter out error exit codes from non-pre scriptlets (rhbz#216221, #rhbz465409)
    - typos and thinkos in scriptlets often cause duplicates to be left around
      for no good reason
    - treat non-zero exit codes from scriptlets as just warnings instead of
      errors
    - ... except that %pre and %preun are sometimes used for preventing
      install/uninstall of a package, preserve the traditional behavior for them
    - abuse rpmtsNotify() "total" value for signaling warning/error to callback

commit 6f3f5a36a3938fd13b44246f8484c35ba1cc5ac3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 3 11:41:41 2008 +0300

    Emit RPMCALLBACK_SCRIPT_ERROR notification on all scriptlet failures
    - previously the fork() failures and such would not be notified at all
    - discard the not-so-useful differing exit codes (signal, exit status and
      whatnot) from commit 179ebc1745b51eb9821dd783c9df475ac539532d

commit e241f3d3ed5a12ca870ad69f7ae3240acfaa13a5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 3 11:38:08 2008 +0300

    Assume failure in psm runScript()

commit e2b89abff7f6592bb2c2f2bda996d89df1960282
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 3 11:34:41 2008 +0300

    Log message on all scriptlet failures.

commit b520b61387757e0ceb907d41cb0d0a5f6478bdbd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 2 12:43:02 2008 +0300

    Updated German translation (Fabian Affolter)

commit fca9b72432f8151427d83a2de01066e138fb2e14
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 1 17:05:35 2008 +0300

    Kick out yet another ancient artifact
    - --usecatalog hasn't done anything in 10 years if ever afaict...

commit e358af91668bb24ae93dbab99cc3acbc429e8aa1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 1 16:28:47 2008 +0300

    We have easier ways to catenate strings...

commit 65e06b20130da895fda8e3698a7ebe18e14d13ca
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 1 10:06:27 2008 +0300

    Tweaks to how ISA-macros are generated (rhbz#464754)
    - for ppc and sparc, move the check for 64bit version first and use a
      wildcards to cover more field automatically

commit aadc1a314640e5eaf31c7f76f9062b1e4e7d1772
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu Sep 25 19:44:29 2008 +0200

    Use safer xrealloc() instead of realloc()

commit 1dd1d68afe6dc55c37f1292250f1c26a5196ea0a
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu Sep 25 19:32:27 2008 +0200

    Avoid segfault in rpmLeadNew() if calloc() fails

commit 50ef054902618edec7a29f5b3b5cb4ae2e12a20f
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu Sep 25 19:30:47 2008 +0200

    Use safer xstrdup() instead of strdup() wherever possible

commit 008280006d5f6bf3696adffb431e968faabc594e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 25 19:16:14 2008 +0300

    Fix up fuzz in Finnish translation

commit 6c4e6f4fa883628c9a31a6536d8f76dbacf5639c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 25 19:10:02 2008 +0300

    Extra ')' in error message

commit 09d43dcbdff4a450a72178d695db553c2f33ef27
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 25 14:58:05 2008 +0300

    Don't go onto wild zombie rampage if python callback tracebacks (rhbz#463447)
    - If a callback tracebacks, the python program that called us in the first
      place is as good as dead already. Clean up what we can, throw an error
      message and just die, no good is going to come out of blindly continuing.

commit fe51be67e4955fd9dbc71d9fdc16231cc32b045d
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Sep 24 19:49:09 2008 +0200

    Update Brasilian Portuguese translation (rhbz#463748)

commit c4fa73c573f63d2e2b53e75d25feafbf65ebf671
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Sep 24 18:39:43 2008 +0200

    Don't treat %patch numberless if -P parameter is present

commit ea9ef92189517164cd68d80c2dd46d8c945ef6fc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 22 14:30:37 2008 +0300

    Kick out useless and ancient junk genCpioListAndHeader()
    - scareMem is not used for anything here
    - xmalloc() + xcalloc() cannot return NULL, don't bother checking

commit ef539ed4dcf5afaf55a45d339ad29c0a930a832a
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Sep 22 12:10:36 2008 +0200

    Fix error messages in processScriptFiles()

commit 9d436590568c657a475e7871d4e13021984633cc
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu Sep 18 15:18:53 2008 +0200

    Drop rpmfcFileDep() to simplify things

commit 5225a86631eaf463eb3b822dc9632820960a7bd3
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu Sep 18 14:57:52 2008 +0200

    Move type regression test back to asserts
    - no need to bother with message translations in this case

commit 74677caa5a6ad36c1836412cd5af378d71dd863f
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu Sep 18 14:56:15 2008 +0200

    Use proper codes for logging via rpmlog().

commit 10908577285571966f79c61df1acbd3913f24760
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 18 14:31:47 2008 +0300

    Force error return if headerLoad() fails

commit acb135c28bd9c637e4db722495a5c890d5593a46
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 18 13:24:25 2008 +0300

    Document manifest behavior for install options too (rhbz#461354)

commit 06fb56e64f9198f50c3f91a9ce1f2c82ea9ea790
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 18 13:05:19 2008 +0300

    Kick out references to "Red Hat Linux" out of the manpages, duh :)
    - RHL hasn't existed in years and rpm isn't specific to RH anyway...

commit 0ba35c7e5c7430bbc72200eda80395c8ede7c5b8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 18 12:38:04 2008 +0300

    Update man-page wrt supported query formatter types

commit 9a9f3c3abbab8981cd1397ddee5a16b4ce65e353
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 18 12:36:31 2008 +0300

    Don't segfault on unknown format specifiers

commit 9da6dce0e68cfe1290e4001def4e5adebb8bf37b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 18 12:15:43 2008 +0300

    Add error message on query of unsupported files (rhbz#461362)

commit a486bcc79329dd48db0c0c50391a940abbb9e75f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 18 11:39:42 2008 +0300

    Remove pointless non-NULL check
    - rpmfiNew() cannot return NULL, and even if it did, we'd be out of here
      by the way of rpmfiFC(fi) returning zero before this...

commit e69e4ebb7ad146eaf716cf0a127da89840fc9bc9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 18 11:29:58 2008 +0300

    Save some screen estate by initializing on declaration..

commit 04386244b9dfc4d9d16217ce4795236284f56d37
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 18 10:26:39 2008 +0300

    More informative output on ambiguous erase (rhbz#164743)

commit bb4d56a96d33b849dcfaf4851f6f79c5cd65f060
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 18 10:04:51 2008 +0300

    Fix rpmErase() exit code when arch is specified (rhbz#462631)
    - essentially the same bug as rhbz#124016 but for erase code
    - rpmdbIteratorCount() isn't realiable on RPMDBI_LABEL, we need to walk
      the iterator to know if there were actual matches

commit 137acc7355ae94959b241674e26006df07a2d84c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 16 12:08:18 2008 +0300

    Kick out double definitions of --(no)dirtokens options
    - mistakenly added in commit 2f7c913723baa8135201fa930678d9e3d4138f2c

commit 0eb396d2ffa5bf89fcf76a50a2080be1c85e7ded
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 16 12:03:26 2008 +0300

    Resurrect --nodirtokens build option
    - this gets somewhat hysterical: we create the filelist as uncompressed,
      then compress it for rpmfi which only understands compressed filelist,
      and if nodirtokens is used, expand the list again after initializing
      rpmfi for cpio/fsm...

commit 55fa1442f85628e75bcb21226949d8940a38b708
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Sep 15 16:42:01 2008 +0200

    Avoid asserts and output error instead (rhbz#448939, rhbz#456100)

commit 320c11f6537d6c898226f7e072497c4af253c5b1
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Sep 15 16:37:32 2008 +0200

    Use proper return codes in rpmfc.c

commit b08d25d841024a45e16c9e7a6797f995da2501f0
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Sep 15 16:35:30 2008 +0200

    Use proper return codes in files.c
    - don't go on if processPackageFiles() or rpmfcGenerateDepends() fails

commit 5e7ca9eac40f7fce940acd8b1c23e2af847478d1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Sep 12 15:22:02 2008 +0300

    headerAddEntry() is deprecated, use headerPutString() instead

commit c094de4d1017af3ffbe89fdbb1db5bb821467c73
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Sep 12 14:56:41 2008 +0300

    Make sure RPMTAG_GROUP is always present in packages
    - Many things (including LSB and various 3rd party applications) expect
      group tag to be present even if they don't actually use it for anything.
      rpmbuild doesn't consider it mandatory in specs anymore though, so stick
      "Unspecified" in there to avoid breaking things in case spec didn't set it.

commit 1327278138c7f24b11b8ea231331991eb20385ee
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Sep 12 11:38:07 2008 +0300

    Fix indentation

commit 9f857078253981e02575bd8f097e50d925476d3b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 11 16:56:50 2008 +0300

    Macro argument handling regression regression (rhbz#461180)
    - commit f366011c42e4cf13d3acdd1af45127f0a1baec3a is too simplistic and
      broke other stuff in turn...

commit 5c71ced2da7d99b870a07a83dd8f7e8d1f1b7e4b
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Sep 10 15:04:57 2008 +0200

    Don't generate broken cpio in case of hardlink pointing on softlink
    - thanks to Pixel

commit ce1db5af7036df54a21abab9a5d70b06a821fff4
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Sat Sep 6 15:30:40 2008 +0200

    Fail hard if patch isn't found (rhbz#461347)

commit f15b7e099682605994b4e7d23cb03018e72abb5e
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu Sep 4 06:34:47 2008 +0200

    Add Dutch translation to LINGUAS

commit 18aab133cdb0f1c67c624d51ce8e72855eb42046
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu Sep 4 06:34:11 2008 +0200

    Add initial Dutch translation (rhbz#461060)

commit a424702c226d6b5cf9bc7e08d97819902286d9ac
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 2 15:25:17 2008 +0300

    Updated German translation from Fabian Affolter

commit 4ff3b1166463d1173e4fff89d910b01d01357230
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 2 14:33:55 2008 +0300

    --repackage is gone, remove references from French manual too

commit 991600462bb7ccbda4d6b85031e44aff01a217d3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 2 14:28:35 2008 +0300

    Updated French translation from man-pages-fr

commit 073b1ac2a88451cc248c4b01825edd01162ffe36
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 2 10:18:41 2008 +0300

    Fix --dump usage description in Japanese manpage translation.
    Patch from Masahiro Matsuya.

commit b53ecfb11de5ab10a40c787028d793d536d64da7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Aug 20 12:01:38 2008 +0300

    Avoid dumb casts
    - switch-case on rpmTag enum makes gcc go crazy without a default case

commit ff7bfd9a32d672f8582ea334ee351af8c32229ad
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Sep 1 14:35:34 2008 +0200

    Create directory structure for rpmbuild prior to build if it doesn't exist
    - creates all needed directories that rpmbuild uses for input and output

commit b16df5b3089e36cf0bd697177983dcae11b849ec
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Sep 1 14:31:33 2008 +0200

    Create _topdir if it doesn't exist when installing SRPM
    - otherwise creating _sourcedir and _specdir fails

commit 436aab02e7f9870835d7c992734fcb2922e273c3
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Aug 26 19:03:25 2008 +0200

    Fix type in format string to avoid segfault (rhbz#460146)

commit 1d6ab4edccf75d1d663ab24938d34e9cc9af3d71
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Aug 26 13:25:21 2008 +0200

    Allow noarch subpackages
    - inspired by patch from Altlinux

commit 743415bb78a773e99e76205b3b435b77bd9ff3e4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Aug 19 13:45:20 2008 +0300

    Improve INSTALL documentation wrt NSRP/NSS requirements

commit 79fe0e0e3bd375c02a6f182508a6feea85379e3c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Aug 19 13:33:21 2008 +0300

    Use standard auto*foo checks for NSS
    - upstream NSPR/NSS doesn't ship pkg-config files (unfortunately)
    - hardcoded guesses are always wrong for somebody, expect builder to
      pass in appropriate CPPFLAGS etc

commit 77ecf321bfb1a2024cb5f6bc231a6b97c56373c6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Aug 14 15:02:52 2008 +0300

    Fix segfault on zero-length %policy file

commit 44263ec047b82f0989b6d9cb89393ee0ebe07f99
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Aug 14 14:28:12 2008 +0300

    Fix segfault on zero-length rpmrc

commit 3b9636cd69b81b6e23331bb797c0fdde79e97316
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu Aug 14 13:08:09 2008 +0200

    Accept numerical characters in macro expansion (related to rhbz#456103)

commit be02cdfa87f593bee96554d0fc5ae4ab1eb04f05
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Aug 14 09:57:03 2008 +0300

    Don't fail build on too large but excluded files

commit 9068da47a912a6983a12d1024a118c16bcb2a057
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Aug 13 09:43:41 2008 +0300

    Unbreak payload size tag generation on ppc (rhbz#458817)

commit fefae4d68e25b69301e9cee12bc54b52fbd11403
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Aug 11 15:39:47 2008 +0300

    Macroize configure command/path in %configure (rhbz#458644)
    - add + use %{_configure} macro defaulting to ./configure, can be overridden
      for VPATH builds etc

commit be56a4a888f60e461154f88843d3a57eeb991d9c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Aug 8 13:53:57 2008 +0300

    Add parenthesis to make gcc happy

commit 17b4187204af7147869d8bcf05f9f9c3c0e0a3d6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Aug 8 13:53:39 2008 +0300

    Missing include

commit 645184ac397a9e62ed3076648505b76b4551e794
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Aug 8 13:06:28 2008 +0300

    Remember to free up resources from matchpathcon_init()

commit dccb80ea38b99e94de84139cb854488e160be38c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Aug 8 13:04:10 2008 +0300

    Fix check for nocontext flag

commit 4537bc35fdc268b31d208e31a7db5729f09345fd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Aug 8 12:49:41 2008 +0300

    Plug memleak from %pretrans
    - te->fi needs to be freed if we replace it...

commit 99eb5361591f9e1a29cc73ccbd696382a4b16d2f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Aug 8 09:57:41 2008 +0300

    Add key to transaction keyring on rpmtsImportPubkey() (rhbz#457360)
    - still not 100% identical to pre-keyring behavior but enough for the
      common case
    - only add keys we don't already have (rhbz#68290)

commit ff1a549fedfe9d564791545b433887283f595ce3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Aug 8 09:32:05 2008 +0300

    Use rpmPubkey + methods for pubkey header generation
    - moves some of the lowlevel pubkey fiddling into rpmio where it belongs

commit 3897bdc37d7667ffb5414dcea91802de3342b9e9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Aug 8 09:26:30 2008 +0300

    Add couple of new rpmPubkey() methods
    - rpmPubkeyDig() to retrieve parsed OpenPGP parameters of key
    - rpmPubkeyBase64() to return base64 encoding of the key

commit 60e5ff80f7d5e19a5ca32cb95e2d2ea535733d6c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Aug 8 09:14:53 2008 +0300

    Split gpg-pubkey header generation out of rpmtsImportPubkey()

commit 4b61c45dbb9965c90ddc1da6c3b8d77959bf682a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Aug 7 18:44:03 2008 +0300

    Permit i18nstring in headerPutString()

commit a366612ec3fde6b2506072fa2288454118fda29d
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu Aug 7 14:50:49 2008 +0200

    Don't allow multiple definition of PatchX or SourceX (rhbz#458261, rhbz#458260)

commit 95925ab512c441d568e76219f006f8de3dc741b3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Aug 6 12:27:25 2008 +0300

    Clarify %buildroot in macros
    - it's no longer really optional..

commit 6c0651926aedc302816a0b61e1f1a8ef568be539
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Aug 6 12:24:34 2008 +0300

    Simplify spec buildroot hysteria
    - BuildRoot: in spec is now really ignored
    - parseSpec() uses default %{buildroot} unless overridden
    - BuildRoot is forced for all builds

commit edf4e046ec5a1c7e07ad390f806fdca2a758f0d5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Aug 6 08:38:57 2008 +0300

    Add %sparc macro for supported Sparc architectures (rhbz#455972)

commit 5a89833b93b3ceb5063c984578f355416eb409e3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Aug 5 15:32:49 2008 +0300

    Increment exit code on file not found errors too (#446202)

commit 58986581b1a4fd256bfdcbfd60bf98cb1786491f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Aug 4 15:46:18 2008 +0300

    Bunch of keyring / pubkey refcounting fixes

commit b5b9329bfa4cd357ce3cac53568aaed7d70383c5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Jul 26 11:14:45 2008 +0300

    Updated Italian translation from Francesco Tombolini (rhbz#453364)

commit 35a08d1df2cb607db8d638d6d43c6125d5688860
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Jul 26 10:56:02 2008 +0300

    Revert "plug memory leak in spec parsing code"
    
    This reverts commit ac3bd07ef2a11cbe9e7f7031786eebc8851c863b.
    - It's not a leak, valueMakeString() doesn't make a copy of the argument
      string and the free() causes invalid memory access

commit 7c4798116adb2b050016b8751b4b236efc4098a1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Jul 26 09:57:09 2008 +0300

    Fix typo in buildroot dir definition (rhbz#456708)
    - it's _buildrootdir, not buildrootdir... this caused user overridden _topdir
      not getting honored by buildroot

commit ac3bd07ef2a11cbe9e7f7031786eebc8851c863b
Author: Florian Festi <ffesti@redhat.com>
Date:   Fri Jul 25 12:28:46 2008 +0200

    plug memory leak in spec parsing code

commit 1d666839ba37da9280df8b463eed86b630a55a76
Author: Florian Festi <ffesti@redhat.com>
Date:   Thu Jul 24 17:38:11 2008 +0200

    fix parsing of empty doublequoted strings in spec file (rhbz#456103)
    off-by-one error

commit 2a44df64cf4c4d04394d1b0894b21e35736a3f49
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jul 24 11:17:50 2008 +0300

    Dumb thinko breaking tar builds from relative paths (rhbz#456321)

commit 6df78d16a68ff508a965ba2166f3b376e850b138
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Jul 19 15:27:42 2008 +0300

    Fix regression in %patch handling (rhbz#455872)
    - popt doesn't think of "008" as an integer, let popt think it's a string
      as we do our own conversion to number anyway

commit ba644ed5aa949f6583ad5e011369527241b62a3c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jul 17 15:50:35 2008 +0300

    Add (and use) refcounting for rpmKeyring and rpmPubkey

commit d422e661c1790db2d6a1e6fe36eff5eb4506ad5b
Author: Florian Festi <ffesti@redhat.com>
Date:   Thu Jul 17 10:37:03 2008 +0200

    Added .gitignore file to get sane git status reports

commit 78ccd971f2cb0dde427ba697282711e3245d0d7d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jul 16 12:07:03 2008 +0300

    Make sure there's room for \0 (packages with no files)

commit 1d9b4d00d7f1a0ca396624411b53d4fb714ccdf8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jul 16 10:52:33 2008 +0300

    Make rpmExpand() smarter wrt memory allocations
    - precalculate unexpanded size and allocate enough for that plus MACROBUFSIZ
      for expansion
    - typical allocation is way smaller than what gets allocated "just in case",
      calculate expanded size and realloc to actual size to avoid wasting
      memory

commit 889f9f5b41d3819c52aad568e75bd0ec5568c0f6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jul 16 10:33:57 2008 +0300

    Eliminate another unnecessary temp variable
    - do all the copying inside a loop

commit 1b19676027fd19369f49629945debcefeeae48df
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jul 16 10:24:05 2008 +0300

    Eliminate unnecessary variable
    - leftover from static buffer times..

commit f366011c42e4cf13d3acdd1af45127f0a1baec3a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jul 16 09:40:10 2008 +0300

    Macro argument handling regression (rhbz#455333)
    - expandMacro() wants the next \0 character to be returned, which might
      or might not be the same as lastc passed to grabArgs()
    - use memcpy() instead of memmove() for the copy, the areas can't overlap

commit e9cae5eecfab5694be7dddc3fde75f1669e6c240
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jul 14 16:23:02 2008 +0300

    Adjust querytags expected results for autotests

commit 649ed675167835796bdf97c48880d56d76bf8629
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jul 14 16:21:15 2008 +0300

    Don't put unimplemented tags into tagtable
    - no point in exposing unimplemnted (ours or somebody elses) tags to users

commit 90c8e0ea0858ec12ddabf3f9732ab9532cd5d334
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jul 14 16:16:41 2008 +0300

    Mark unimplemented tags as such in tag table

commit b3bd21e9f07272a37c2259070f95b167e6301c79
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jul 14 11:53:10 2008 +0300

    Fix mono dependency extraction
    - use "Mono/.Net assembly" instead of "PE executable" for detection,
      later file reports bitness (eg "PE32 executable"), we just care if its
      mono or not. This is compatible with older libmagic too.

commit b256dc9e25dc7e9f9886edab0f3c8ceef8a4b27c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Jul 12 17:57:51 2008 +0300

    NSS uses unsigned int, not size_t for hash lengths
    - broke ppc64 in entertaining ways...

commit 6cf51345a40d6c0eb84d5d3c590d72fa48090ced
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jul 11 13:37:38 2008 +0300

    Updated Polish translation (Piotr Drąg)

commit 6b3a81c5ddcd5a3f55af984cd33b7280f31e3c1f
Author: Ville Skyttä <ville.skytta@iki.fi>
Date:   Tue Jul 1 20:22:02 2008 +0300

    Decompressor clarifications.
    
    Signed-off-by: Ville Skyttä <ville.skytta@iki.fi>

commit 78d829a3011c07f03e1e4db4b3ba128f88591c9b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jul 9 11:56:22 2008 +0300

    Include rpmlib.h from rpmcli.h, legacy stuff expects it there..

commit 5e7ea3212a93eb1dcc1d4dbebcddd9daa4c0522b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jul 9 11:55:57 2008 +0300

    Check for _RPM_4_4_COMPAT in rpmlib.h too
    - limits the chances of compat includes not working due to include order
      a bit

commit c56ca35a5315054c044f7aee03618f93af58978f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 8 18:35:56 2008 +0300

    Include legacy compat from rpmlib.h to cover maximum ground
    - rpmlib.h is the first thing any old rpm api user includes so this should
      cover most cases (only if enabled)

commit f1b25fc2576e30fd7604e2b4a6f30aa0738b224c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 8 18:35:07 2008 +0300

    Add compat definitions for renamed string functions too

commit 7a6ef7ceeccb4994f57a87333806d31934a926d9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 8 18:13:42 2008 +0300

    Missing cplusplus-wrappers in rpmtd

commit 1d0cbf396215a0167b112a80796d51a50718eab2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 8 18:11:28 2008 +0300

    Bunch of dumb macros to make up for renamed functions
    - should arrange deprecation warnings on them too...

commit d8bdf1ef85dbde77fd5f7d5bd5db6c09650580ab
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 8 16:49:54 2008 +0300

    Add deprecation markers for hge etc vector prototypes

commit 425d43a8a056685432c9f80b12b1ec8920950408
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 8 16:45:44 2008 +0300

    Add deprecation warnings to all legacy types

commit 1cab53124a70691519c751def7c75740a24484bc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 8 16:43:46 2008 +0300

    Compat typedef for Spec -> rpmSpec rename

commit 65077f4d5b5ea218082c757c79ae28d07717b0b3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 8 16:42:32 2008 +0300

    Add mappings from hTAG_t and friends to legacy interfaces

commit 9e586c62b18905691817587353e434eb03b4e7d2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 8 16:41:24 2008 +0300

    Limit the exposure of rpmlegacy.h stuff to bare minimum

commit 6ea17f9ca4f1434510be658fc048de21ec3b2109
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 8 16:37:23 2008 +0300

    Shove header legacy interfaces into rpmlegacy.h
    - get them out of sight from main header.h
    - turn headerSprintf() into macro around headerFormat(), that way
      rpmTagTable and rpmHeaderFormats lossage can be hidden away as far
      as headerSprintf() use is concerned

commit 27a6edd36d39e296e6b13e24974989ad5954abb9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 8 16:23:22 2008 +0300

    Add a new legacy compatibility header, kill rpmints.h
    - build-time generated rpmints.h causes issues with distcheck and is
      not really worth the trouble
    - we have all sorts of legacy junk to deal with, sticking them all into
      rpmints.h would be just wrong...

commit 6778dc94ce0a9234ed74cf094d98e219a295070f
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Jul 8 17:46:54 2008 +0200

    Initialize optCon
    - one could easily segfault the doPatchMacro() code without
      the optCon initialization (particularly with jump to "exit" label)

commit 096d1af159718bc8d6cea7355fa9f59bc4bd9f10
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Jul 8 17:45:12 2008 +0200

    Make the default fuzz for %patch is now configurable via macro
    - the macro is _default_patch_fuzz

commit 831c5c5460b3ae2ee25d3033db6f18f33e1fc873
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 8 13:00:20 2008 +0300

    Update translations

commit 97fe9f138ee4602cb642137b70e9b5a47da32d23
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 8 12:59:51 2008 +0300

    Add ISA-stuff for ia64

commit e5d4fc50c2286f34d0154d915c1b73927e5f7b89
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jul 7 18:11:59 2008 +0300

    Missing const in parseUnsignedNum()

commit b215f1145e59356c48f1dcefdfcd7bd4fd843ab5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jul 7 18:09:49 2008 +0300

    Teach Fflush() about lzdio

commit 1a173c486dd1d83f22f8123317303fc1fb880f54
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Jul 7 14:49:33 2008 +0200

    Make Source0: equal to Source: again

commit ff23d0a2cf9d9b9fac3e2818fd1f067e79ab70a1
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Jul 7 12:59:03 2008 +0200

    Use proper type in format

commit c69f826ade69a753da062872fe1110574277413f
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Jul 7 10:21:30 2008 +0200

    Remove useless argument to format, d'oh

commit e8cef9b4f0242b593b8541ba5a22ed6d249a0260
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jul 7 13:28:04 2008 +0300

    Eliminate couple of timedRead() uses where we control how fd is opened

commit 32d69b7adc78f1707f6f11268b68841c0b235dc9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jul 7 10:52:33 2008 +0300

    Make verifyAttrs and virtualFileAttributes static
    - nothing outside files.c needs

commit 811436ec92fa874585552204fc9e4136e0b406c0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jul 7 10:48:06 2008 +0300

    Kick out ftpStrerror() and related error code defines
    - our url-retrieval error codes depend on the helper used which is
      configurable via macros so hardcoding eg curl codes wont work either
    - need to figure out something better, pass helper error messages through
      FD_t or something...

commit 7b95061d3b77d8e6b3c80e856b66f49b393511ba
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jul 7 10:42:55 2008 +0300

    Kick out now unused printNewSpecfile()

commit 0f991a14209054da7c45c46e96750738fc62874d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jul 7 10:17:21 2008 +0300

    Kick out ancient, useless and hidden --specedit option

commit 5129c291bde8b7785fa70fb42b2844e962369ace
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jul 4 18:18:55 2008 +0300

    Remember to remove ugid.c from POTFILES

commit 9e7cfa4d5ba64baf63c604426805a71671c58544
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jul 4 16:09:42 2008 +0300

    Stuff rpmio/ugid.[ch] contents to lib/misc.[ch]
    - no uses outside librpm itself, and there they can be hidden away from
      exported ABI whereas in librpmio they cant

commit 104bc6a883cb59ce5f7de966fc6d8d11dc4ed247
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jul 4 15:02:30 2008 +0300

    Make ufdClose() static, nothing outside rpmio.c needs directly

commit ebdfd2887fcb6ebccd153f78b5b775495903e6b9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jul 4 15:00:48 2008 +0300

    Remove unused ufdGetFile() from librpmio

commit bd078a5811fb6405f9aa94177f8f0cc47eafee39
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jul 4 14:53:02 2008 +0300

    Remove unused _url_debug

commit 7985d6afb5c1f290c4c60b4b1c74e55c0353cacd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jul 4 14:52:47 2008 +0300

    Bump version to resemble library versioning

commit 3b57a00f063640a46f291f88acbbad966000e249
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Fri Jul 4 15:00:09 2008 +0200

    Really make fuzz=0 default for patches
    - and finish the Patch and Source fixup

commit 724b07bba5e802998b1b79b408c2401d2a238a3b
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Fri Jul 4 14:58:38 2008 +0200

    Don't treat Patch and Source as Patch0 and Source0

commit 6c1103fc80bd0f063338c7147e2e9fac10b3a32c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jul 4 14:33:23 2008 +0300

    Remove rpmcache from repository
    - next to nobody even knew about it, much less used it...
    - hardly useful now that solvedb is gone

commit d902040e1484f1352dafc96c77678da0ebf44035
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jul 4 14:22:06 2008 +0300

    Remove solvedb API and other related stuff
    - nobody has used the solvedb for anything in years
    - the mechanism + rpmcache are horribly broken on multilib
    - solvedb is way too heavyweight to be really usable as a general mechanism
    - solver callback hook left in place as that could in theory be usable still
    - we need something better to replace this all, getting rid of it now
      so we're not stuck with the API forever

commit d0fbf7b246d4232a3a833753d488646e0636de2a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jul 4 14:11:47 2008 +0300

    Kill of rpmtsAvailable()
    - long since deprecated and nothing uses it, might as well go

commit 47c85270631de173d873b98bc79727e2db203007
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jul 4 13:01:39 2008 +0300

    Enforce conflicts on colorless files always
    - silently swallowing conflicting files when they happen to be in the
      same (suitable colored) transaction but conflicting when installed
      separately is nothing but inconsistent
    - partial fix for rhbz#190209

commit 46592cde0dc41cf6ab26af0579fec8044502a98b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jul 4 09:17:33 2008 +0300

    rpmfiFNlink() returns unsigned int

commit ee2b26644174e1ae93ac979050e7acccc369ae88
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jul 4 09:11:20 2008 +0300

    Couple of type corrections
    - use rpmTag, rpmsenseFlags instead of int32_t

commit 6c8b729ac4df24f407f6d36fade22a7c44ab71d6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jul 4 09:07:19 2008 +0300

    Use unsigned integer for epoch, remove unnecessary static

commit 5a4807c81c0a06646ae2a9b52c310ac605590db9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jul 3 17:44:43 2008 +0300

    Parameter name + doc mismatch on rpmVerifySignature()

commit 78040394a29b3960cd6adaf8bd626936093008a3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jul 3 17:44:13 2008 +0300

    Change headerPutString() arg name to match the others
    - doxygen whines a bit less..

commit 2093f7ccebc1fd66c9ecec7d1f61eb5642070654
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jul 3 17:14:45 2008 +0300

    Plug leaks in selinux context handling in fsm

commit 03b2c395b2abaf726b6c63e06f2c6c795a2d83d7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jul 3 17:01:52 2008 +0300

    Plug a memleak on erasure
    - on erasure the order of things in ts->order is reverse from that of
      rpmtsi-iteration, causing rpmtsiNext() to abort earlier than it should
      when things are removed underneath it
    - just walk the ts->order table without iteration, it doesn't make any
      difference here

commit 1de6c1eb12e0c2d05a342649f73befc00d6de0cb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jul 3 15:30:04 2008 +0300

    Make all transaction callback notifications go through rpmtsNotify()
    - single point to control how callbacks are made

commit 9d200565744d3023053d64f627c82cf2451fa701
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jul 3 09:11:55 2008 +0300

    Only try to pubkeys from rpmdb if we non-db keys didn't exist
    - try to avoid opening the db unnecessarily, if the keys have been migrated
      out of the db, the db probably only has duplicates (right? :)

commit 0e1e131e1adf15dbce5a1fcdc197b029f3e283fc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jul 2 19:07:06 2008 +0300

    Don't leak memory on duplicate pubkeys

commit ef17f6bcfc564c6cb132e5e4013b971819ebe66c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jul 2 19:00:56 2008 +0300

    Don't leak memory on python expandMacro()

commit 20ee2a2df04f4667609b1d98bd798692bd14ecf5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jul 2 18:49:05 2008 +0300

    Plug a dumb memleak that has crept into rpmReadPackageFile()
    - jumps to exit from very early on, make sure sigtd is reset early so
      calling rpmtdFreeData() is safe

commit 83e894db7841dd4e751d785a88e49e4ae5894624
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jul 2 15:08:51 2008 +0300

    Change rpmts keyring get/set interface to more transparent handling
    - add extra parameter to rpmtsGetKeyring() to automatically load up default
      keyring if not already set
    - rpmtsSetKeyring() with NULL keyring just frees up current keyring
    - adjust callers wrt the change
    - allows more transparent action for various "legacy" code-paths

commit 43a05ea36331a3463efd33e3a6cbbc3d723f36e5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jul 2 14:24:20 2008 +0300

    Add documentation for keyring

commit 710750ebff4ded200a71107b6dba8f5daf573101
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jul 2 14:23:30 2008 +0300

    Make rpmkeyring.h public header

commit 152a6e84aa0fa801ad2ac063570119df698d8b93
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jul 2 13:55:50 2008 +0300

    Add docs for the new rpmts methods, deprecate rpmtsFindPubkey()

commit fe09759eefbe34833a39f696517e558fa429a075
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jul 2 13:48:00 2008 +0300

    Load pubkeys from both files and rpmdb for now
    - easy enough and doesn't cost really anything
    - at some point, start warning about legacy gpg-keys in rpmdb, suggest
      conversion method / do it automatically & such...

commit cea83865c944d378d3aa8632c81a68c3d669b4db
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jul 2 13:31:12 2008 +0300

    Some keyring improvements
    - keep keyring sorted by keyid, use bsearch for key lookups
    - check for key in keyring before adding it to avoid dupes
    - actually check that pubkey parameters match the signature on keyring
      lookup like was done in rpmtsFindPubkey()

commit 396ca4694710ca5a1e4ac799f1c68bb0c788b7d8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 1 17:35:21 2008 +0300

    Eliminate remaining signature checking bits from rpmts
    - rpmdb-specific rpmtsFindPubkey() no longer needed for anything, and
      neither are the cached packets and stuff

commit e0e4d1c17c873e02e0482207250369d1b27b2782
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 1 17:32:11 2008 +0300

    Use keyring even when using pubkeys from rpmdb
    - build-time switchable, could easily be run-time too

commit b10bc97b30e459b865a81fcbe4dab8eef84a0485
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 1 17:00:04 2008 +0300

    Pass keyring, not the full ts to lowlevel signature routines
    - lazy loading of default keyring on ts iteration works nicely for
      everything but package signature verification which doesn't access the
      db now that the keys are not in the db anymore
    - make rpmtsSetKeyring(NULL) load the default keyring, a bit ugly but
      it'll have to do for now

commit ca52a32bbc30ea5dbcf206e6c8023f71e02f6540
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 1 15:50:05 2008 +0300

    Eliminate digest calculation statistics from signature code
    - if we want them back, they just need to be detached from the ts...

commit 3c3fe4388fd42e177171582d03f59aea6babb324
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 1 15:46:18 2008 +0300

    Add methods for getting and setting the keyring used by a transaction
    - semantics of exactly when and how this is permitted needs more thought...

commit caf362cff8d1967bd3d9360c117beacc93470343
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 1 15:38:32 2008 +0300

    Move keyring related typedefs to rpmtypes.h

commit 0b385002cc04254ce9f994efe1c2400edd9d1516
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 1 15:25:06 2008 +0300

    Proof of concept rpmtsFindPubkey() implementation with the new keyring
    - ifdeffed out for now, needs a bit more flesh and a working
      import routine

commit 685877f2a0ec949186bbcad1f18ee7895e0da15b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 1 15:24:04 2008 +0300

    Beginnings of an abstract keyring
    - in librpmio to avoid dependencies on higher level stuff
    - for now it's just an in-memory structure that needs to be populated
      by someone simply by feeding in GPG public keys

commit e435dde712e390f1bdc729b8c792da9bd0633e16
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 1 14:35:44 2008 +0300

    Create ascii armor that rpm itself accepts

commit 0e9f7e2d62c1dec40742c25593aed7eaa436e092
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Jul 1 12:52:16 2008 +0200

    Update parseNoSource() and addSource() to accept only unsigned SourceXX numbers

commit ad60332e0494e4da632fb3ed5345bb94d607d3da
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Jul 1 12:51:08 2008 +0200

    findSource() eats only unsigned SourceXX numbers

commit 301ef0b5e66a129f952fdb410fe889940106ca69
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Jul 1 12:49:19 2008 +0200

    doPatchMacro() now uses parseUnsignedNum()

commit 7f9befa4c34dc6edace37d9961086c67d2c256ec
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Jul 1 12:48:50 2008 +0200

    doSetupMacro() now uses parseUnsignedNum()

commit e9e77e64636f4a8e7f066776e3556a8e2de2e51c
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Jul 1 12:47:36 2008 +0200

    doUntar() accepts only unsigned SourceXX numbers

commit afc392537ffb84396dedfe2cb4270e7dfa6ccd3f
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Jul 1 12:46:16 2008 +0200

    Convert doPatch() to use parseUnsignedNum()

commit e831de1040243bd33163a930137f1318ece4d61c
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Jul 1 12:43:47 2008 +0200

    Use the same number types for SourceXX tags as returned by parseUnsignedNum()

commit 206fd12122b7c2e0cf07768f639d78b5f1e4bd5e
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Jul 1 12:42:12 2008 +0200

    Convert Epoch parsing to use parseUnsignedNum()

commit 0f8102c1edc8ba7bb1b42fed1a9b59a797648ca9
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Jul 1 12:41:35 2008 +0200

    Avoid reading negative constants from SourceXX, PatchXX tags
    - convert parseNum() to parseUnsignedNum()
    - don't misleadingly declare it to return int, when the result is
      always unsigned due to strtoul() call

commit bcb226480d0b60a8667eade77d81c29de4c3d6c9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jul 1 11:22:40 2008 +0300

    Eliminate the nasty global _dig and _digp variables from rpmpgp
    - pass the digest + parameters around as arguments instead

commit 9594f656d49e1a43e2ffea94c21ffc8f3c4373f8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 16 17:49:02 2008 +0300

    Unexport all the umphteen pgp related tables
    - nothing outside rpmpgp.c needs

commit c2efff14861624a7c8d4d32ef6b681f642804e13
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 16 17:01:29 2008 +0300

    Make every pgp function not needed by other parts of rpm static

commit 1f425eb2908a293f19cdad92f81ce5b74ab520c9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 16 11:12:15 2008 +0300

    We don't need ts to stash the keyid locally now

commit 8e5b817e3cc4f1a272d3ad20807212d27174a208
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 16 10:59:23 2008 +0300

    headerCheck() does not recurse
    - it can be re-entered by the way of rpmtsFindPubkey() but that doesn't
      matter now as dig etc are local variables

commit 38b401a98986b8f627a9b0b9c3ca94b403e962b0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 25 11:17:17 2008 +0300

    Detach pgp digest from transaction set
    - pass the pgp container around as argument as needed
    - eliminate the related API from rpmts

commit 884714d326af946a06c1e690afe55e95fb351fff
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Jun 14 14:48:53 2008 +0300

    Pass pgpDig around from rpmVerifySignature()
    - get the digest once in rpmVerifySignature() and pass to verification
      functions, signature is part of the digest so no need for separate call..

commit 5c08f55432f9fdc5df7b7bc6ca52b02c68e69dc7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Jun 14 17:09:50 2008 +0300

    Detach signature tag data from transaction set
    - pass signature tag data around as argument to rpmVerifySignature() as
      needed
    - eliminate the related API from rpmts

commit d51023569e2595cfcc76204b4e2a2cb6ba56b4f6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Jun 14 14:03:11 2008 +0300

    Actually use the sigtd passed to verifier functions

commit fcc1dc3369d5821b9780ece8cfeb1fe2bca6a15b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Jun 14 13:49:37 2008 +0300

    Start getting signature tag out of transaction set
    - collect various signature items into rpmtd and pass around to
      verifyFooSignature() checkers

commit a89c861a3de113ebce481eb2e337e71dfb9e2c21
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 30 15:13:54 2008 +0300

    (char)NULL doesn't make much sense, use \0 instead
    - shuts up another gcc warning

commit 19d3b1c8203208954e8d1511bdd5d5cf064548dd
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Jun 30 12:23:42 2008 +0200

    Fix missspelling in macro name (_initddir -> _initrddir), thanks to
    - Ville Skyttä
    - old macro with typo is kept for compatiblity

commit 669a169adea04a2b2019b498fc045fce8349a3dd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 30 11:28:32 2008 +0300

    size_t for "left" is wrong, use rpm_loff_t instead

commit 4cdc43c6c2b68883d603c3363472014c7163a34d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 30 11:24:34 2008 +0300

    strntoul() is supposed to return unsigned long, not int
    - the type mismatch busts up handling on largish files...

commit c1a5eb371270c2dee3e335ab3589e90fdc0195c7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 27 15:15:46 2008 +0300

    debugedit: Include empty CU current directories (rhbz#444310)
    - patch from Jan Kratochvil

commit b55b066e9c30ec4962434fa73d701083fb4d5364
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 27 14:39:39 2008 +0300

    Use correct formatters for debug-printing in rpmtsInitDSI()

commit 5f42506680ce69837d4708d38b27a7cfe346b024
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 27 13:49:48 2008 +0300

    Use correct formatters for debug-printing sizes in printSize()
    - on x86_64 size_t happens to be 64bit, not so on x86...

commit 8b6f82feecf0b5206eff998b07e8958962190513
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 27 13:02:23 2008 +0300

    Adjust python callback for amount+total type change

commit 636311912b9426b5bc1da2d256ec4b97b92f1686
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 26 17:52:31 2008 +0300

    Argh, the tag is LONGSIZE, not LARGESIZE

commit a9d58acedc555d38228486863f72510293f3c626
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu Jun 26 15:58:40 2008 +0200

    Take new directory into account in rpmbuild test

commit 65110dfa7dba070cf0ec2467d1f4ddae458640c9
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu Jun 26 15:56:55 2008 +0200

    Add new BUILDROOT directory to default rpm build dirs.

commit d562af843e60c4c7dc27e3086130c6bcd560e74c
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu Jun 26 15:56:11 2008 +0200

    Add new macros for enabling default build root

commit 5c46ebb7a95f3076ef4ba38a30d4998057c03258
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 26 16:37:06 2008 +0300

    Make --info work on large packages too

commit 219917989ab737b145324cea1ee760bc0203426f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 26 16:34:32 2008 +0300

    Teach build about 64bit payload etc size tags
    - only use 64bit variants for packages that actually need it to avoid
      breaking compatibility when not needed
    - old rpm will bail out "cleanly" (with an obscure error message) if
      the 32bit size tags are missing but at least it doesn't crash...

commit 5831f37b215be053096ee38b090bddba2e7cdad6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 26 16:28:17 2008 +0300

    Teach rpmAddSignature() and rpmReadSignature() about 64bit tags
    - use 64bit size if it exists, otherwise quietly fall back to 32bit one
    - can't use tag extensions to convert there as legacy sigtags conflict
      with things like RPMTAG_NAME

commit 09ca0902dfacd6de00237235717af3d9162a76e2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 26 16:23:56 2008 +0300

    RPMTAG_LONGSIGSIZE and RPMTAG_LONGARCHIVESIZE are signature tags
    - move them from regular tag space to sigtag space, fortunately these
      don't require special conversion

commit e84165833f26c2cbfd32abb50439a3ab34be8767
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 26 16:10:49 2008 +0300

    Only insert 64bit filesizes on build if required
    - older rpm's get pretty upset on encounter with 64bit integers, avoid
      unnecessary breakage as very few packages actually need large file sizes
    - Add per-filelist flag for large files, enabled from addFile() as we
      check sizes. It's basically a no-op for now as we error out if any
      single file exceeds cpio per file limit.

commit a930e73ef733210df9be7e142ec81d206fbd9f70
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 26 13:59:06 2008 +0300

    Unbreak compressFilelist()
    - well and truly broken by yours truly by a Monday morning commit
      efe1665024e374a8f8704d229b1d306af232c396, blush...
    - can't use headerIsSource() here, as build calls compressFilelist()
      before RPMTAG_SOURCERPM is added to the header, revert to earlier hack
      and check filename beginning with / or not
    - dncmp() expects address of a string, not the string itself, this was
      masked by the above bug

commit ef34894567a3ecd82035fc3b82ca06a3773d9ac3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 26 11:38:28 2008 +0300

    Add a bit of protection against oversized files
    - max size of individual files in cpio new ascii format is well known,
      check we don't try to stuff in larger than what fits and abort cleanly
    - this should do as long as we only support new ascii format cpio payloads

commit 6985cd23724b8cde7f7a73507d8130be18c270af
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 26 10:50:01 2008 +0300

    Type corrections
    - Fread() and Fwrite() return ssize_t, not int32_t
    - sigtag is rpmSigTag, not int32_t

commit ae177457c8d23584f45b5f36866ce4ed20509e1a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 25 10:59:20 2008 +0300

    Fix handling of "%posttrans -p /xxx" with no body (Pascal Rigaux)

commit 1ba0513f7f68690339147a56de29883f24591952
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 25 09:56:12 2008 +0300

    Adjust autotest querytag expectations to match current tags

commit c26161d0ab973e99f065fd3ee578461175ee1211
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 25 09:53:21 2008 +0300

    Adjust autotests for rpmi etc removal

commit 997f5a24c5420676e8a3e29845902dca7f98a44e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 25 09:33:50 2008 +0300

    Force "rpm" popt context for everything but build
    - we want the same set of aliases to be seen no matter what name we're
      invoked with, except for build which has it's own set of aliases
    - this has a nice little side-effect of permitting popt aliases to work
      from source tree on non-static builds despite libtool calling executables
      lt-something instead of their real names

commit 7a02b41365399f665f0ff097f05583ae537ac6ee
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 25 09:20:27 2008 +0300

    --buildpolicy is rpmbuild, not rpm, alias

commit 52bea969c7404f646050844f8a5576d60032aaeb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 25 09:19:13 2008 +0300

    Kill the million duplicate aliases in rpmpopt

commit cb17fb722c9576502378cd79e21517819db2cdc4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 25 09:18:51 2008 +0300

    Eliminate rpmd, rpmi, rpmk, rpmq "helper binaries"
    - it's all rolled into "rpm" itself, these are just clutter and complicate
      things needlessly
    - leave symlinks to rpmquery, rpmverify, rpmsign and rpmdb -> rpm to
      preserve some compatibility (at least rpmquery is known to be used
      in scripts)

commit 600c98615426cf2e5a606e51b447250cd3be9c71
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 24 16:07:05 2008 +0300

    Update rpm manpage wrt build modes
    - the build modes have been removed ages ago, just point to rpmbuild manual

commit 30be238164bebe44b00202a7a2ee37bbd5b6bf75
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 24 15:23:30 2008 +0300

    Eliminate build-related aliases on "rpm"
    - "rpm" for building packages hasn't worked in ~six years, time to clean
      up the junk from rpmpopt

commit 94df5d41dee6dc5fc73d2e9c02e7bf9153eb6a87
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 24 11:11:38 2008 +0300

    Eliminate already commented out build mode aliases for rpm
    - these haven't been enabled in six years and not coming back...

commit 70282af4b3b5d5911ffe1dabd04ee79c64045150
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 24 11:09:00 2008 +0300

    Eliminate rpmb aliases from rpmpopt

commit 6f06fff34884e334cd5eab3b371481b4e1039876
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 24 11:08:05 2008 +0300

    Eliminate the pointless rpmb + rpmt symlinkery
    - just build "rpmbuild" executable and install to bindir instead of
      mucking with several different names

commit aa9a791d808f504781d0b75255df3387383a1809
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 24 10:08:37 2008 +0300

    Force FD_CLOEXEC on all potentially open descriptors
    - instead of just "100 should be large enough", use sysconf() to grab
      number of max open files and do them all. If sysconf() fails,
      use 1024 as "should be enough for everybody"

commit f39d2432f74bdc328ceafa8abc6cac517e02c73b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 23 16:57:51 2008 +0300

    Remove artificial limit in dependency loop elimination attempts
    - continue processing as long as progress can be made instead of artificial
      hardcoded magic "try ten times"

commit 9fb0258e04d70608982f3f65a881b053a4e19645
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 23 15:02:10 2008 +0300

    Avoid silly deprecation warning on headerFreeData()

commit 6c7f2262f3b8a1399c6ed130ff0e388236507d2c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 23 14:37:54 2008 +0300

    Simplify requires and provides addition in rpmfcGenerateDepends()
    - just feed (name, evr and flags) one by one into header instead of
      constructing arrays of them

commit f23f0a8d62aeeab6f41667c53f7bb12f1817cc83
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 23 14:31:54 2008 +0300

    Oops, provideflags were being inserted as requireflags

commit 2ed85a4ff9cc25800f607cd0450b02fde97cf3b6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 23 08:46:51 2008 +0300

    Add Malay translation from Sharuzzaman Ahmat Raslan

commit 880a28f2ff45da7d7003db9c270d1ed173cb90f0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 19 16:08:00 2008 +0300

    Dumb, dumb thinko/leftover from earlier prototype...

commit 1c445fcd974eedd0e533487a40214b328b1eaadf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 19 15:52:54 2008 +0300

    Use type-specific headerPuts() for rpmcache
    - ...to finish of undoing yesterdays damage, sigh ;)

commit 0199599a0d2bd049b54b7977b27f182bd8a23663
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 19 15:52:19 2008 +0300

    Use type-specific headerPut() for the case where it makes sense in relocs

commit 3782d652156dd3067c09f0441bc087d2b4b790ff
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 19 15:51:07 2008 +0300

    Use type-specific headerPut() in rpmpsmStage()

commit c7d0e4ada36a40e6dd77a80f7db570a22f603862
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 19 15:50:32 2008 +0300

    Use type-specific headerPut() for installtid stuff in rpmdbAdd()

commit 17cbd0f9725c3e08524b7a29c89b4574c9862710
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 19 15:49:09 2008 +0300

    Use type-specific headerPut()'s for rpmtsImportPubkey()

commit 6d7b5c0a7aa80ba2adf9892c941536acaaf78b85
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 19 15:48:31 2008 +0300

    Use type-specific headerPut()'s for legacy retrofitting stuff

commit f754ad8aa640b7a72936a19dabca64bacdbb5e63
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 19 15:47:04 2008 +0300

    Use type-specific headerPut() where makes sense in build/parseScript

commit e940b206b95f78063902518e7dd1151bd278b46f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 19 15:46:34 2008 +0300

    Use type-specific headerPut() where makes sense in build/rpmfc.c

commit 6844c194ccd7ccebd48719600b5e32f14a620f0f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 19 15:45:57 2008 +0300

    use type-specific headerPut() on all of build/reqprov.c

commit cd4c3c0cd26a2977fa4b62363b72fa3930457625
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 19 15:45:37 2008 +0300

    Use type-specific headerPut() on all of build/parseSpec.c

commit e79b9743b12438de490879fe6d5cc674b3ba6118
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 19 15:45:06 2008 +0300

    Use type-specific headerPut() for all of build/parsePreamble.c

commit e663f0bef2c6af9f773e8dccb7e3e1af43afc5c6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 19 15:44:29 2008 +0300

    Use type-specific headerPut() for all of build/parseChangelog.c

commit 683b2e5b72d36e38b9a09f9153427fe5b2d0616b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 19 15:43:45 2008 +0300

    Use type specific headerPut() for all of build/pack.c

commit a317eaadc3f7b9b4ec373719e7d60fb8b98e2c21
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 19 15:43:14 2008 +0300

    Use type specific headerPut() for all of build/files.c

commit 62d7b0c98d30c70613319e6af742827a67d9c732
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 19 15:41:58 2008 +0300

    Add type-safe headerPutFoo() methods for supported tag data types
    - requiring all access through rpmtdFromFoo() just adds unnecessary
      indirection and pain for little gain, the header is fairly intelligent
      when it comes to inserting data
    - this gives us various benefits over the old headerAddEntry() interface too:
      + basic type checking done by compiler
      + extra sanity checking (ie you can't add string data to integer tags,
        don't permit adding more than one entry to non-array data etc)
      + "do the right thing" approach - add / append as needed and supported
        by various types
      + headerPutString() can now be used on both single strings and string
        arrays making the interface much nicer for the rather common case of
        appending strings one by one to string array tags

commit a10f770ccaed6e722a0065a59887d6fc133cc86a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 18 16:14:36 2008 +0300

    Dumb thinko breaking scriptlet interpreter

commit 0d1cdd4a965cee9fac416341d3566c3d06a8b305
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 18 15:39:59 2008 +0300

    Group legacy interfaces at the end of header.h
    - out of sight, out of mind ;)
    - add deprecation warnings where missing
    - move to header_legacy doxygen group to avoid littering main documentation

commit 96470400712b6f8d68c906751f2d8df934c34bba
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 18 15:22:46 2008 +0300

    Convert headerAddI18NString() to use headerPut()

commit efe4e4c9461ef3a1b253a5bcb628882bc0b21ee3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 18 15:16:05 2008 +0300

    Switch around headerAdd*Entry() family vs headerPut()
    - make internal header add + append take rpmtd as argument, legacy
      interfaces are just wrappers around them
    - add deprecation markers

commit b6577f0e6de7f7888823eeaa396c6d2bae369f20
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 18 14:59:13 2008 +0300

    Switch around headerMod() and headerModifyEntry()
    - headerModifyEntry() is the legacy interface going away, make it
      use headerMod() and deprecate it

commit a8f23a461958479a773c482260899522b76efa81
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 18 14:54:41 2008 +0300

    Stuff legacy compat wrappers out of sight into header_internal.c
    - probably need to carry the compat stuff around for a while but get
      them out of the way...
    - headerFreeTag() and headerFreeData() are just the same thing, with
      headerFreeTag() supposedly being the "new" interface which never was
      used, duh :)

commit 9bbfae241a67ecb15889f7e100f6c5111486255b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 18 14:35:56 2008 +0300

    Start shredding legacy cruft out of header code
    - headerGetRaw() can now be done with just a flag to headerGet()

commit 335363733e79c80f3ce211bad7290f1b897e7dc5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 18 13:15:46 2008 +0300

    Convert signature code to use headerPut()
    - tags in signature header overlap with those in "regular" header,
      can't use rpmtdFromFoo() as that would give totally bogus types and errors
    - just add a wrapper around headerPut() that behaves like headerAddEntry()

commit 9432e7a9bc2478edca78ba1ac3f855e2f2dda7db
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 18 13:13:22 2008 +0300

    Convert parseScript() to headerPut()
    - AFAIR the only place in rpm where the type of a tag varies like this, ugh
    - any modern rpm can handle the type being either string or string array,
      but preserving legacy behavior for now...

commit 005c5663e3f9eac1c6af835dcec7d8074783be31
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 18 12:57:36 2008 +0300

    Convert rpmpsmStage() to use headerPut()

commit ddd39615fd1bb0603be5626f1441dafc99c7055d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 18 12:55:02 2008 +0300

    Convert rpmdbAdd() to use headerPut()

commit d6735b669ef412d6be9b5d129436f98a3a0f0c10
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 18 12:52:28 2008 +0300

    Convert rpmtsImportPubkey() to use headerPut()

commit d68f236bba1c72fd667a7c128ad1cba98180b177
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 18 12:37:42 2008 +0300

    Convert RPM_BIN_TYPE uses in build to headerPut()

commit 9b162540e17fbe09f8de7d9bed09295e8c2219e6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 18 12:36:20 2008 +0300

    Add rpmtdFromUint8()
    - handles CHAR, INT8 and BIN types

commit 7e8142f3cf7b3818b2d9a3a86ae0cd62eb25f772
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 18 11:32:59 2008 +0300

    Convert legacy retrofits & friends to headerPut()

commit feb037946435a82c80d9332bfbd2b3b530cd39e2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 18 11:16:04 2008 +0300

    Lose the remaining hae and hme vectors from rpmfi

commit 53edb2a11c9f96654218cbe9c395dcf1f866f478
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 18 11:13:58 2008 +0300

    Convert file relocation to use headerPut/Mod

commit 312ab62fe10b05874b107084c26e9110dd13e84f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 18 10:59:45 2008 +0300

    Add headerMod() to headerPut() etc family
    - just a dumb wrapper around old headerModifyEntry() for now

commit 19415eaf046cb8f1a96f0d8bee5ae2433dc1b707
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 18 10:15:37 2008 +0300

    Convert rpmcache to headerGet+Put

commit 513e86a7b552d1db9cea3772829d723228175008
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 17 16:11:49 2008 +0300

    Number of sources can't very well be negative...
    - also happens to avoid compiler whining about differing signedness

commit 61aa9083544315f76d049f36794cc0bf6e5eb9fa
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 17 16:07:55 2008 +0300

    Shut up incompatible pointer warnings from rpmtdFromStringArray()
    - it'd be far nicer to be able to append single strings to string arrays
      through rpmtdFromString() but it'll require some interesting hackery
      in header.c
    - use temporary const char* pointer where necessary to convert to
      const char ** compatible type before passing to rpmtdFromStringArray()

commit 2f55856c85d02a81318065dbb1dd06ff92c9004c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 17 16:04:12 2008 +0300

    Make rpmtdFromStringArray() data argument const afterall
    - produces somewhat less warnings as is, a bit surprisingly

commit 05540d121b2555a79877fe013689ce0a6ecea258
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 17 15:36:33 2008 +0300

    Oops, RPMTAG_FILEMODES is INT16, not 32 type

commit d6aded3128ae82bbea3c8701f3be4326c71767db
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 17 14:54:44 2008 +0300

    Convert rest of writeRPM() to headerPut()
    - can't use rpmtdFromFoo() on signature header tags so need to construct
      them manually

commit 15cc52e514deedb403cbe1d516b8ffa09c8edba0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 17 14:45:14 2008 +0300

    Convert the final pieces of rpmfc to headerPut()
    - more const-impedance here.. need to do something about it

commit 36a187ec3308276b52ccf4a5858c0f5273bcef1e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 17 14:37:55 2008 +0300

    Convert the rest of genCpioListAndHeader() to headerPut()
    - explicitly cast all platform dependent data like time_t, mode_t etc
      to the type used in headers always instead of having everything doubled

commit f4cdf53a7616f541d6f8e3523bb18b50d2063d5a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 17 14:24:16 2008 +0300

    Add rpmtdFrom() methods for 16 and 64 bit types

commit 457a3110708e9d6c3e06441a5900b722c4c874a5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 17 13:30:08 2008 +0300

    Convert the straightforward bits of build/files.c to headerPut()

commit a48e97b5f00ca2ae366d3fa5af3656aed4a19b1f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 17 12:57:59 2008 +0300

    Convert most of parsePreamble to headerPut()

commit f6a8d969de0a05f78bc32a4f27f5017c348a67ee
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 17 12:56:53 2008 +0300

    Convert build reqprov to headerPut()
    - also changes index type to uint32_t instead of signed

commit 85f0bbe8abedbc8c2f316f5680055cf700ca4f25
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 17 12:55:33 2008 +0300

    Convert parseSpec to headerPut()

commit a61a1785a66a908af1c4d91b428af9c7ddfced08
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 17 12:55:01 2008 +0300

    Convert parseChangelog to headerPut()

commit 2b6f72a7ee51e6a38b845ade303897e4f9194a2b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 17 12:10:01 2008 +0300

    Convert most of build/pack.c to use rpmtdFromFoo() and headerPut()
    - heavy on asserts() for now to catch out any problems
    - sigtags can't use rpmtdFromFoo() as the tag numbers collide with the
      regular tag table and would cause failures if we tried

commit 61f096c11ee62c423c72957373a9b7d38557d5fe
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 17 12:07:19 2008 +0300

    Add some new rpmtd creation methods
    - rpmtdFromString(), rpmtdFromStringArray() and rpmtdFromUint32() permit
      populating tag containers of compatible datatypes that can be used for
      headerPut() with some extra type-safety and sanity checks on the types

commit 8687797c1ce3d43181d1991e6413098b3dd95a17
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 17 09:07:37 2008 +0300

    Replace headerRemoveEntry() with headerDel() everywhere
    - this is a no-brainer s/headerRemoveEntry/headerDel/g operation as the
      interface didn't actually change at all

commit b6c2fd31d8f6b28ace55d1656d19eed3cd6f70f0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 17 09:02:41 2008 +0300

    Rename headerRemoveEntry() to headerDel(), add compat wrapper
    - no differences here except headerDel is a nice short name (lifted from
      rpm5.org) to go with the new headerGet() etc family
    - deprecate headerRemoveEntry()

commit fd3d65cd5bdccf20f98762d84fe65bbb1c72b8b8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 17 08:49:19 2008 +0300

    Eliminate last headerFreeData() in rpmfi

commit bb81679aa5296b18856f1672ae85b6c7ab627e97
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Jun 14 14:43:15 2008 +0300

    Preserve headerNext() allocation behavior
    - commit bca23d19af0a6d7371d18f6fd763e37e03f6e21a accidentally changed
      it to MINMEM which breaks signature checking, lovely...

commit 541cc0447af97f8193eb5d8e9f1c6a4c71b61ffc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Jun 14 11:31:15 2008 +0300

    Lose the useless tag data type fields from rpmds
    - we know names and evrs are string arrays, if they're not we're screwed
      anyway, and flags need to be freed if header is not around
    - there goes hfd() too...

commit 92f6c2d3bd87d9dd67678ea61bca9a8158ed4ade
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 16 09:24:46 2008 +0300

    Use PRIu64 for file size formatter in queries

commit 3b75f9d589bd732dfac142266cbdd785a4bbefb4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 16 09:06:15 2008 +0300

    Doxygen parameter name mismatch

commit d8b28213ae4f9927e4f82e131aaad1dbcd5e7dd5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 13 16:55:12 2008 +0300

    Deprecate headerGetEntry() and headerGetEntryMinMemory()

commit fb618299bb0709941b2a1ee0ac23f8e50b492c7b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 13 16:52:39 2008 +0300

    Convert rpmtsSolve() to new headerGet()

commit 7b8fc75dea12884d51e41e9298d575910b0d072c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 13 16:44:35 2008 +0300

    Kill off hge() and hfd() vectors in rpmfi_s
    - hfd() with FORCEFREE or STRING_ARRAY is just regular free()
    - hge() no longer needed...

commit 30ebb0ba57056bd2a45aa36929c8c250afc35ed1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 13 16:38:10 2008 +0300

    Lose the remaining hge() uses in rpmfi relocation
    - ugly hackery for now, otoh it wasn't pretty to start with...
    - always make a "sparse" dupe of directory names so we know how to free
      it even if just some of the directories were relocated (this was broken
      from the previous round of relocation alloca removal hackery)
    - take advantage of RPMTD_PTR_ALLOCED to correctly free in all cases

commit 2f00f421ee3f1b945d228f7ca5175597ebc2ee43
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 13 14:17:08 2008 +0300

    RPMTAG_FILEFLAGS is not used for anything in relocations

commit 1091951365d28cabb8c7188738602975a6fdc7db
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 13 13:59:04 2008 +0300

    Convert validRelocation parts of relocateFileList() to new headerGet()
    - one piece at a time, this baby breaks easily ;)
    - take advantage of iterating over td as possible..

commit 911b3151ae1819d18a19b8d032e81aac25249036
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 13 13:52:32 2008 +0300

    Unbreak relocations broken by 621e7518418a0e031d302e8681fe0bf81d0c4b72
    - stick scareFlags into rpmfi_s for now so they're available everywhere
    - subtle games with how fi->dil is alloced on scareMem behavior, don't
      blow up and don't leak memory either
    - stuff scareFlags into rpmfi struct for now

commit f32ad12ea8391fd892d696987157ba59a0028b8a
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Fri Jun 13 12:51:01 2008 +0200

    Use 64bit sizes by default in rpmcache

commit d6026decfb6b85ff04521c37dfcd7e9b1482cf0e
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Fri Jun 13 12:31:26 2008 +0200

    Allow spaces in Source/Patch URLs and filenames

commit c071f46c85c2a0ad93e6d7c272f593b1303efc2e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 13 12:23:46 2008 +0300

    Simplify providePackageNVR()
    - convert providePackageNVR() to new headerGet()
    - use rpmds to figure if we already have the provide instead of manually
      doing it

commit 973cff0efcd62856505ad123ac63bd693a1fd261
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 13 11:42:46 2008 +0300

    hfd with FORCEFREE is just _free()

commit d23f3a67c73ff7789c4b6dea6f05f8b394abe3c6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 13 11:42:06 2008 +0300

    Convert rpmpsmStage() to new headerGet()
    - grabbing fi->fuser and fi->fgroup there seem completely unnecessary,
      they're always populated from rpmfiNew() afaict...
    - hfd() with FORCEFREE is just the same as _free()...

commit fe87b4a5cf101919ad8d3aa16fb1561fe0b5e4d8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 13 11:27:13 2008 +0300

    Convert runImmedTriggers() to use new headerGet()

commit 1910af96d9695f3ad654d393ca03c2d2f237f268
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 13 11:17:00 2008 +0300

    Minimally convert handleOneTrigger() to use new headerGet()

commit 615fff42f78201f673c064096d83691c663dff2f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 13 10:27:46 2008 +0300

    Convert legacyRetrofit() to new headerGet()

commit 669af8b842a075878f5d654e89e98c338de8d212
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 13 10:21:40 2008 +0300

    Minimally convert handleInstInstalledFiles() to new headerGet()
    - fi->hge style doesn't matter as otherStates is just local temp variable,
      use MINMEM always to avoid unnecessary copy (DEFAULT wouldn't copy either
      atm but..)

commit a82694308c93e40461b10a8a9572349232c0a067
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 13 10:16:24 2008 +0300

    Minimally convert rpmdbFindFpList() to use new headerGet()

commit 39b71970952e5190b76f1ba5f4e5967a38318b19
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 13 10:09:20 2008 +0300

    Minimally convert rpmdsNew() to use new headerGet()
    - take advantage of HEADERGET_ALLOC to avoid manually duplicating Flags
    - otherwise just dumb minimal conversion for now, rpmds could make use
      of rpmtd iteration etc but leaving that for later...

commit 7f8485a61cc7863d0047f0828f66a0579aa2b768
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 12 17:37:33 2008 +0300

    Convert a few hge()'s in relocation code to new headerGet() & friends

commit 9879f8440d8bebd8890cac95602dc06c158e516d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 12 17:09:35 2008 +0300

    Make fssizes tag extension 64bit
    - type changing doesn't matter much as this is just runtime generated data,
      not from headers
    - also gets rid of the remaining old hge() use in tagexts

commit 7ddf1722a8a66c5197f68655e8859f88585ece86
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 12 17:03:04 2008 +0300

    Change rpmGetFilesystemUsage() size args to rpm_loff_t

commit 7dc709b51751c065ff9df8098a13f04a7c6fea52
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 12 16:41:31 2008 +0300

    Make rpmfiFSize() return rpm_loff_t type, fixup callers
    - Internally file sizes are still stored as 32bit, going to 64bit wouldn't
      make much sense (would only waste memory for nothing) as long as we're
      bound by cpio's 32bit per-file limit. Being "64bit ready" for filesizes
      doesn't hurt anything though...

commit ec378955f15325db5cfa826746ca9ff60c011bfb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 12 15:31:27 2008 +0300

    rpmfiNew() doesn't use old hge() anymore

commit 4a680922c08653b0a9d56146404fb8b6f8fececb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 12 15:29:40 2008 +0300

    Use long archive size always in rpmfi

commit 4c6f6d88b86839557066672ae9ab33f2a53f6ecc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 12 15:05:43 2008 +0300

    Add more 64bit size tags + tag extensions for converting from old values

commit 3c5a8a7046cb6437f7a94742f5531a64165db684
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 12 15:01:58 2008 +0300

    Move 32 to 64 bit tag conversion to helper function

commit 5f361fe6a184b4d3772e234043a4f7dd647e41c9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 12 14:16:28 2008 +0300

    Use HEADERGET_ALLOC instead of manual copy in longfilesizesTag()

commit 621e7518418a0e031d302e8681fe0bf81d0c4b72
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 12 13:35:09 2008 +0300

    Use HEADERGET_ALLOC instead of manually copying the data in rpmfiNew()
    - headerGet() knows better how to create a copy of the data, let it...

commit ebcdb5c57a1e1d570c53927901e45d5c8f8c178d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jun 12 11:04:33 2008 +0300

    Dumb thinko in header data allocation

commit 7705737c4731f9b7164c91b150e2a94053bd07ff
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Jun 11 16:13:18 2008 +0200

    Use rpm_loff_t in printFileInfo()
    - and use snprintf

commit bca23d19af0a6d7371d18f6fd763e37e03f6e21a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 11 12:26:52 2008 +0300

    Add headerGet() flag to force allocation of returned data
    - earlier there was no way for caller to know if returned data or parts
      of it pointed to header memory (other than "know" how rpm behaves on
      given types), this allows consistent behavior for callers (ie you
      always "own" the returned data, not depending on header) when needed
    - if HEADERGET_ALLOC flag is set, all returned data is malloced, instead
      of possibly pointing to header memory depending on type
    - HEADERGET_ALLOC overrides HEADERGET_MINMEM if both specified
    - adjust various header internal callers to accept flags instead of just
      minmem argument
    - TODO: make sure tag extensions honor this too (all but fsnamesTag
      currently allocate anyway)

commit 1ae76f69df0eacf8f7fc7aab6b920010a5dde25f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 11 10:31:14 2008 +0300

    Make rpmps generic number field 64bit, rename getter method
    - rename the parameters + internal use to more generic "number"
    - rename ugly and dumb rpmProblemGetLong() to rpmProblemGetDiskNeed()
      and return rpm_loff_t
    - the number field is used for some other purposes too (whether dep problem
      is for installed package), work around this for now by just defining
      rpmProblemGetDiskNeed() result value as undefined on non-disk related
      problems ;)

commit 55d3bbd4b87ea7eec48cfa573bdac9001a8cc4da
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jun 11 09:25:02 2008 +0300

    Enable rpmDoDigest() to pass large file sizes
    - should probably be off_t instead of rpm_loff_t as this is lowlevel io
      business and only verification code cares about the size reported by
      it, otoh forcing 64bit always avoids having to deal with variable
      off_t type on printing.. shrug

commit 6c4772537d39f413cf62b1eefdcc151ae3a4b056
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 10 18:11:32 2008 +0300

    Add new tag extension to retrieve filesizes as 64bit types
    - convert old 32bit sizes to 64bit if large sizes aren't present in header
    - if large sizes are present, we malloc + copy the result to have consistent
      behavior for rpmfi

commit 9130995df6d2c6461042f3470499b933c10e0211
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 10 16:00:33 2008 +0300

    Minimally convert rpmfiBuildFNames() to new headerGet()

commit b0885bd0f20362bd2a30f59659ce1b61735b074d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 10 15:05:31 2008 +0300

    Callers are now well known, require non-NULL arguments

commit 8f12edb8f8ac6b3e30c9680562ebf9b4ed8a00d9
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Jun 10 15:33:24 2008 +0200

    Add 64bit filesizes to header on build

commit ac7f7121eba87fa2717e82d5b27ef335e3486517
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Jun 10 14:42:50 2008 +0200

    rpmtsUpdateDSI now uses rpm_loff_t for fileSize

commit 96d9edabb3c97d14438cb7c705caf0b04c72eac7
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Jun 10 14:41:10 2008 +0200

    Use 64bit data types in diskspaceInfo_s

commit cb4bce5e24e736c2b7e3dc28d2e475077d286ef3
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Jun 10 14:37:52 2008 +0200

    Create RPMTAG_LONGFILESIZES to hold 64bit filesizes

commit 6b3e0cc97bad88009ea84d7951cede53b7f20dc8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 10 14:36:12 2008 +0300

    Eliminate rpmfiBuildFNames() from the API
    - headerGet() with extensions enabled provides a nicer interface to
      the data, no need to expose it

commit dd46457ab5bedcaad7f83782dc6057d97372ec50
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 10 14:31:43 2008 +0300

    Eliminate rpmfiBuildFNames() use in psm

commit a2df078607de4cb7fe34164f0f1806a84120eddf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 10 14:29:30 2008 +0300

    Add RPMTAG_ORIGFILENAMES tag extension

commit d6641e87c5f28b3fbfa7566a583f6bc9f3523865
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 10 14:24:57 2008 +0300

    Eliminate rpmfiBuildFNames() use in python

commit 00a9e06618a09d057fdf08d2608536da75e741db
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 10 13:40:51 2008 +0300

    Use rpmFileHasSuffix() instead of doing it manually

commit 54f40c0faf2fc203447dc1533a2e3a3456551567
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 10 13:35:20 2008 +0300

    Convert psm rpmInstallSourcePackage() to new headerGet()

commit 779c05972673122d5a34f76e42932af3ab9f2a84
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 10 13:08:11 2008 +0300

    Convert psm markReplacedFiles() to new headerGet() interface

commit b4b6e53183ffe928932a11bf30de0c186510e7b4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 10 10:04:52 2008 +0300

    fwrite() returns size_t, use correct type

commit af9518dce2d483efe3d9f7191f583662dae276e2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 10 10:02:52 2008 +0300

    Use correct type for strlen()

commit 4ffee98a652dbc4fcfeb04076111bedf043a7956
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 10 09:59:09 2008 +0300

    Silly signedness mismatch in pgpMpiSet()

commit ae2cd002e679c9fffb7ac8d21380e91408532339
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 10 09:54:51 2008 +0300

    Use nlink_t for hardlink count types
    - avoids some signed vs unsigned comparisons

commit 46d4daee4560da3499b4be5ce71ad6994fb279ab
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 9 19:00:37 2008 +0300

    Change tag extension returns codes to match headerGet()

commit e08c4251a5e0fac04f86c0b03de5766d5c0cbf42
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 9 15:42:24 2008 +0300

    FD_t contentLength field is useless for us too...

commit f2bdbe3d7df24e874bb9713354a5bad4a7b7c349
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 9 15:38:28 2008 +0300

    Lose bunch of unused fields from FD_t
    - leftovers from times of internal http/dav etc transports..

commit 45ac934fdcf0efc14147051d37bab789b00b41da
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 9 15:33:27 2008 +0300

    Use 64bit type for fdSet/GetCpioPos() in rpmio

commit 243218b9cdfe43a878677d1abd530df0ee9986da
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 9 12:58:43 2008 +0300

    Use 64bit type internally for printing size when reading sig header

commit 522c6d074ef9f8ddb54feeaf40f28a846f2ec840
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 9 12:48:34 2008 +0300

    Use 64bit type for rpmte package size reporting
    - RPMTAG_SIGSIZE limits to 32bit in practise currently

commit c5119e79005fb6097c7bc420f045ba2fc1ba5bea
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 9 12:36:19 2008 +0300

    Use 64bit type internally for total file size counting in build
    - ensure we don't overflow RPMTAG_SIZE which is 32bit for now

commit 161395648f72bd27d18fcf96bdea38d24c74b437
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 9 12:29:07 2008 +0300

    Use 64bit type internally for package archive size & offset
    - cpio "new" ascii format limits individual files to UINT32_MAX, no such
      limit on the entire archive though
    - RPMSIGTAG_PAYLOADSIZE is 32bit type atm, assert that the archive size
      fits to UINT32_MAX despite internal presentation being larger
    - FD_t limits the real archive to rpm_off_t still

commit a45f4a18038cd8f4f3f7f7a7a2ea54e7ad5fde93
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 9 12:00:12 2008 +0300

    Use rpm_loff_t for transaction callback amount+total arguments
    - ought to be safe as data is only ever automatically converted to larger
      type on notify calls

commit 94552b96256c3620b4be407c501d0d926c081963
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 9 09:31:00 2008 +0300

    Handle NULL in str2hge()
    - at least apt-rpm expects to pass empty version as NULL to rpmdsSingle(),
      don't blow up...

commit d4c04512a15733ebfababd42b67d957744e613b9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Jun 7 13:17:52 2008 +0300

    rpmtdCount() knows how to handle BIN types, use it..

commit 5e40a2eb7723e0f812f25dc2d8bcdab845ade391
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Jun 7 11:43:54 2008 +0300

    Couple of more 64bit format types

commit db964ebf78a0f6a5434631ff2715517934e243ca
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Jun 7 11:32:02 2008 +0300

    Remove unused variable

commit 5c9c0c5353ba125a075441964738ac9fcabfdc84
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Jun 7 11:31:32 2008 +0300

    Handle 64bit type in headerMergeLegacySigs()
    - not that any legacy tags have 64bit type but shuts up gcc..

commit be712bc960fe8a75fc55e724b1e7d4980281eeb1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Jun 7 11:29:16 2008 +0300

    Use inttypes.h formats in stringFormat()
    - add 64bit type while at it

commit 3d8e8c5e53ca60817395164b1e08593da82cd2da
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Jun 7 10:51:53 2008 +0300

    Teach rpmdb about 64bit types

commit 05db695ee6cec89b98db87e7b00ecbffb65eb634
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 6 23:47:37 2008 +0300

    Add access methods for 64bit integer types to rpmtd

commit d1c9e039a4a245edf470d78dd465f55388cd916c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jun 6 18:53:44 2008 +0300

    Fix typo in sqlite backend preventing it from building

commit 90c18fbeced891129828c15e67bb22c3b25b7726
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Fri Jun 6 15:15:01 2008 +0200

    Introduce 64bit rpm_loff_t

commit 18def173765c3b820e99fd502e4d42853746823e
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Fri Jun 6 15:10:57 2008 +0200

    Teach sqlite backend 64bit handling

commit e2427595aac7f3d121324d67c474b27d4c549edd
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Fri Jun 6 15:04:25 2008 +0200

    Handle RPM_INT64_TYPE in header functions

commit 49dfc587d5b356551d646e40d575378a737c331f
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Fri Jun 6 15:03:19 2008 +0200

    Introduce htonll()

commit 56a8f9d90811cf6afc5a2cbbfff561d572f44307
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Fri Jun 6 14:57:05 2008 +0200

    Introduce RPM_INT64_TYPE

commit 09d8cbafff94cde37accd9c43fcd8087457925cd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 3 14:26:34 2008 +0300

    Unexport file state machine internals
    - fsmGetTs(), fsmGetFi(), fileStageString(), fileActionString(),
      fsmMapPath(), fsmMapAttrs() and fsmStage() aren't needed outside fsm,
      make 'em static to avoid leaking to symbol space

commit 50f34a622edc4d737d671622401f7dad44a10d62
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 3 13:51:57 2008 +0300

    Rip doputenv() for good

commit ed859877a5bcaff3ff3ac40ed39a48d1149471a9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 3 13:48:57 2008 +0300

    Use setenv() for preparing scriptlet environ
    - setenv() allocates memory as needed, no need to leak it
    - setenv() is nicer interface for the purpose anyway

commit a7a9595d9b1885fb57a27b76f15bb9ce09203949
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 3 13:39:07 2008 +0300

    Rip dosetenv()
    - just use regular setenv() instead
    - we already carry setenv() in misc/ for platforms that don't have it

commit 6f010ec18d7f48b123b96bce77be43f84c8328e0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 3 12:11:05 2008 +0300

    Unexport rpmsqCaught now that nothing needs it

commit 0863a4420ba0569da1edb0dd8018ce7f62c25a28
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 3 12:10:02 2008 +0300

    Use rpmsqIsCaught() instead of directly accessing rpmsqCaught sigset

commit cd040ba59b00ce709fa97016df9eaf733adc1ccf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 3 12:04:57 2008 +0300

    Add rpmsqIsCaught() for testing if a signal has been caught

commit df206dbaaf59af1c4879ecd3739f2cf9a46e48af
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 3 11:44:20 2008 +0300

    Unexpose signal queue and direct manipulation functions
    - nothing outside rpmsq needs to fiddle with the queue directly

commit fe5ff47868f2b6349bc14fa5bb5484d54110e6a3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Jun 3 09:09:00 2008 +0300

    Don't leak memory on python fi.Digest() calls

commit d6380f029cfa6b39eb036fe1818737296737f6fd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 2 16:10:19 2008 +0300

    Change rpmTagGetNames() interface to more like that of headerGet()
    - instead of returning a newly allocated rpmtd, take container to use
      as argument
    - return number of tags found, 0 on error (pretty arbitrary, just to
      return something...)

commit 6962b29bd9e8e67134aa43a163dab29bb7650426
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 2 13:05:46 2008 +0300

    Remove bogus comment on rpmDoDigest() ignoring algorithm

commit 4dee28ba4c01f105ec3c55ea5a4d3540357b0b42
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 2 13:02:54 2008 +0300

    Permit non-md5 digests on --fileid queries
    - the used algorithm is not known at this point so we can only do some
      "seems to be in the ballbark" checks on the digest
    - dynamically allocate memory for binary conversion as needed

commit 26dbd57da584e422649184f751ebbf850f01e64b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 2 12:20:22 2008 +0300

    Check that digest type is actually supported before enabling it
    - we don't necessarily support all types listed in pgpHashAlgo_e, it
      depends on what NSS happens to support

commit a8d8d5a4f4b77316292b8a2276bb74886945b6c1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 2 12:18:23 2008 +0300

    Enable build-time configuration of non-MD5 file checksum digest algorithms
    - if not specified (or 0), quietly assume MD5
    - otherwise check for supported digest type, complain and fall back to
      md5 on invalid
    - if non-md5 algorithm is used, add rpmlib() feature dependency to package

commit 2cbe2c3f5a31262fe4ccfb43234392ff95712e59
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 2 11:48:22 2008 +0300

    Document rpmDigestLength() error behavior

commit acab1c3999a4205ed43344c89c11818d9ba5093e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 2 10:53:38 2008 +0300

    Eliminate ugly static conversion buffer on rpmdbAdd/Remove
    - dynamically allocate on heap when hex to binary conversion is done,
      add new parameter to td2key() to communicate this back to caller
    - adjust rpmdbAdd+Remove to free key data if signalled by td2key()

commit b8dea7798888cf79ef2a5f5558f974482e5f740d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 30 14:45:13 2008 +0300

    Make rpmfiNew() use header specified file digest algo if present
    - if not present or unknown algorithm specified, fall back to md5

commit a98970a41be2be86bf4265450072daa03ce48eef
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 30 14:14:17 2008 +0300

    Use RPMTAG_FILEDIGESTS instead of MD5S internally
    - no functional changes

commit f788ab42eb141a7dc3cb289fad0e6dcfa2c95434
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 30 13:57:47 2008 +0300

    Bring back RPMTAG_FILEDIGESTS now that we're a bit more ready for it
    - this reverts commit 6db2a8ec03c057850a0349537c705556068b9d32.

commit 05be2a2fdb44558d2b6bc1b785128c1decae2aae
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 30 13:16:23 2008 +0300

    We don't want per-file digest algorithms, just per package...
    - rename RPMTAG_FILEDIGESTALGOS to RPMTAG_FILEDIGESTALGO
    - change type from integer array to single integer value

commit b9c69d1bab3c5b932664a7d071aaf11ef1d48483
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 30 12:28:29 2008 +0300

    rpmfiFDigestHex() instead of manual hex conversion
    - showQueryPackage() and python rpmfi_Digest() at least..

commit 29bdb49f16361548d53de75b827a2c926d4d292a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 30 12:21:48 2008 +0300

    Add rpmfiFDigestHex() for converting the binary digest to hex
    - conversion (back to) hex presentation is done manually in several places,
      might as well have a function to do it...

commit db573c81ed17f6852221fab05799382d4fa6cb89
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 30 12:11:16 2008 +0300

    Mark rpmfiMD5() deprecated

commit e1724c3d4ac040cd3bb6124df94e764338557475
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 30 12:06:40 2008 +0300

    On a second thought, rename rpmfiDigest() to rpmfiFDigest() for consistency
    - it operates on current file of the set, not on the entire set so while
      changing API, name it similarly to other things that operate on current file

commit fad3c10b90cc547f4149dd5db70874a2ea1ea992
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 30 12:02:21 2008 +0300

    Add fi.Digest() method to python bindings
    - use same code for creating the digest object on iteration and method call
    - fi.MD5() left around as alias for fi.Digest(), dunno if that's the best
      thing to do but it's just a python string so returned size isn't that
      critical..

commit 55d5c497363d3b8892979c16259ca6d54a6bde01
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 30 09:22:21 2008 +0300

    Convert showQueryPackage() to use rpmfiDigest()

commit 7b696eb1a92769d807aa68c6fbe56535b427d595
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 29 20:41:21 2008 +0300

    Make rpmVerifyFile() honor rpmfi digest algorithm

commit a1e49ce26be58043c913d69547afa3632386e7a0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 29 16:39:11 2008 +0300

    Use new style headerGet() + rpmtd iteration for rpmfi digest population

commit ae5667cf06d882e2760b924585a8900dcdfcfe19
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 29 16:26:21 2008 +0300

    Lose unused fdigests array from rpmfi struct
    - only used in rpmfiNew() where it's converted to binary from the header
      hex presentation and free'd immediately, a local variable will do just as
      well
    - some useless references in fsm and psm, kill'em too

commit ea134adf81d22b26807d2c71895b0c23a7804d33
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 29 15:48:40 2008 +0300

    Make handleOverlappedFiles() honor rpmfi digest algorithm

commit 4aa7ab642237180fb5ab7cf0547425ae1c311d3a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 29 15:33:44 2008 +0300

    Teach fsm internals to honor rpmfi digest type

commit b851f93fc630ea5f16fada8e0ffce9ae8f1515df
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 29 15:26:12 2008 +0300

    Make rpmfiDecideFate() honor rpmfi digest algorithms
    - this gets into trouble if digest algorithm differs between packages,
      can't be helped (same is true of rpmfiCompare() too)

commit 5046e63f279b4be78279abcc327606b894dcb8c9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 29 15:20:05 2008 +0300

    Make rpmfiConfigConflict() honor fi digest algorithm

commit db4d3ea25558a26f560615d1df3d12d5286e036d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 29 15:16:45 2008 +0300

    Make rpmfiCompare() use rpmfiDigest() instead of rpmfiMD5()

commit 4c4c6d0423c182189fe65177554ad9179c786200
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 29 15:09:58 2008 +0300

    Make rpmfi initialization honor used digest length

commit 95e65d98ead5c354a0d257b9112a5cb14c88ebb4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 29 15:04:46 2008 +0300

    More preliminaries for non-md5 file checksums
    - add rpmfiDigest() method (similar to one in rpm5.org) for retrieving
      file digest, algorithm and digest size
    - make rpmfiMD5() just a special case of rpmfiDigest()

commit 479b74bebe1caf2c785d4e9ac5a5974904c05f3a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 29 12:29:23 2008 +0300

    Rename rpmfi internals to talk about digest rather than md5
    - no functional changes

commit f771b641d33f9388c952d3fc7ad4daba9d789543
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 29 11:48:00 2008 +0300

    Rename rpmfi and fsm checksum md5-sum fields to refer to digest
    - no functional changes
    - groundwork for supporting other than md5 checksums

commit beb573c2a02c20d593577c7b358d722ed1c021c4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jun 2 09:48:41 2008 +0300

    Unbreak handling of packages without epoch
    - rpmtdToString() returned NULL on empty data, rpmtdFormat() returns
      (unknown type) which isn't exactly appropriate value for epoch...

commit 27855634e99f66eef2fbe4ca8837a00b46964cc2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat May 31 13:28:05 2008 +0300

    More generic iterators replaced with safer string iterators
    - additionally ensure rasprintf() will never get passed NULL where string
      is expected in doScriptExec()

commit c275732c75e23501fb66fa30ee025c2a1a50a736
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat May 31 13:20:21 2008 +0300

    Use safer per-datatype iterator instead of generic one

commit 4bf224c0c99ba20440444660f9767190c62176d7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat May 31 13:18:19 2008 +0300

    Replace couple of generic td iterators with safer string iteration

commit 784fe494b772bcb54b23fdc6d017f301c606d0fd
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu May 29 09:48:16 2008 +0200

    Make Group: tag not mandatory in spec

commit 903c5d9fec67bc15695807fb507912206e039d40
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 28 09:53:30 2008 +0300

    Move key size fixups from rpmdbAdd & Remove to td2key()

commit c96d78d9727b1f5c8c473ab83d0d7d8d3a162027
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 28 09:42:31 2008 +0300

    Don't bother with rpmtdInit(), headerGet() resets index anyway

commit 827a800687140df56e446b4752ac9c7bd5613d0e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 28 09:41:34 2008 +0300

    Eliminate ill-indented block obfuscating the code

commit 63f24f2307426106a0bccde91c1b8cfa8bc5da36
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 28 09:35:13 2008 +0300

    Unify rpmdbAdd() and rpmdbRemove() debug logging
    - move the debugging out of the loop, we only print once for a given tag
      anyway
    - dumb helper function for prepping the actual output

commit 2a1c813d495c5b05576f2feff6679aed20e587d5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 28 09:04:31 2008 +0300

    Unify rpmdbAdd() and rpmdbRemove() tag data to key "conversion"
    - the tag container contains everything needed for converting
    - ugly static buffer for holding data from binary conversions needed,
      figure a better way...

commit 0f123cbfd951b6bca0bf5da104ae45e9a3746f83
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 28 08:20:19 2008 +0300

    dbi->dbi_rpmtag had better be the same as rpmtag here...

commit 1f32f4aa7773e247f55cb05e9b86f4d3e5009d02
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 28 07:52:19 2008 +0300

    Convert rpmdbAdd() to new headerGet() + friends

commit e4f021461d22880794c69e2974173ad03f640d5e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 27 21:49:02 2008 +0300

    Remove RPMTAG_BASENAMES special casing from rpmdbAdd()
    - no longer needed as we don't support uncompressed filelists in the db
      anyway (commit 30bb2ba4c8e481dd1c18e6ce8cd6c563fb42e23b)

commit 27557a00d57974a17b8af933ce8c6f1ebfeb76b5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 27 21:33:28 2008 +0300

    Eliminate ill-indented block obfuscating the code

commit f55661ff4d89ab46c1bfb332a3c8f29b548a0688
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 27 21:27:20 2008 +0300

    Convert rpmdbRemoev() to take advantage of new interfaces
    - use headerGet() and rpmtd iterator on the result
    - eliminates breaking out of loop kludgery
    - allows all filemd5s fiddling to be done in one spot

commit b10e0cc5ee430c77738f07509b1372c1c146cd2f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 27 20:16:17 2008 +0300

    Turn silly one-case switches into if's

commit 432eba406defa8cc0ecec85750450a9d891d4555
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 27 14:26:54 2008 +0300

    Minimally convert rpmdbFindbyFile() to new headerGet()

commit 0520845553cf949a384cfa05c84a8bd0c9c47da6
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon May 26 16:14:25 2008 +0200

    Remove unused pointer

commit cbf71aaf9d8abee453dacd28a325d73d2acae7f1
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon May 26 16:09:46 2008 +0200

    Remove legacy RPMSENSE_SERIAL.

commit f3703cbaf0e58c0c6eb567465284efbf1df6fe9a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 26 17:23:36 2008 +0300

    Similar payload compressor headerGet() conversion for rpmpsmStage()

commit d7b6c49df211eb775265c79922d608fb285c088f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 26 17:19:07 2008 +0300

    Convert rpm2cpio to new headerGet()

commit a9d7285cb0c984e42e0358a3e0d305ecb7d22333
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 26 17:07:00 2008 +0300

    Oops, don't free data we're supposed to pass around...

commit af8b41c64af39ce07d85fcd92fa78d566747d815
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 26 16:24:07 2008 +0300

    Use rpmds methods to check for duplicates in addReqProv()
    - simplifies things quite a bit...

commit 5267010b65524d697a93042b21cff33416c35417
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 26 15:28:03 2008 +0300

    Convert addPackageProvides() to new headerGet()

commit c4082255fd1db46fb720dbe48f57413784718e56
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 26 15:27:33 2008 +0300

    Leftover debug junk removal

commit 1814782eef6bc94300cb351111e08872442f1b49
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 26 15:25:11 2008 +0300

    Convert writeRPM() to use new headerGet()

commit 1315d9f498a38916a128ef95d3bbd7557fc1ebc3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 26 15:16:39 2008 +0300

    Convert processPackageFiles() to new headerGet() interface

commit 0537efeb77531d2db84deee1c0c872617f00caab
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 26 14:45:00 2008 +0300

    Convert addFileToTag() to new headerGet()

commit 5c2acdcdfb756192feeb0f98cc3772484af7ddfa
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 26 14:31:54 2008 +0300

    Convert timeCheck() to new headerGet() & resurrect functionality
    - it's been broken since forever as it used RPMTAG_OLDFILENAMES but by the
      time it got called, the filenames had already been compressed so it didn't
      find anything to check
    - largely useless but easy to fix, shrug..

commit 059de6fa03f8c2113ce60d6857a5bf0746c43e0c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 26 13:17:07 2008 +0300

    Convert rpmInstall() to new headerGet()

commit 3a32216d0299c6aa771ae849749c3f590db82a0d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 26 13:14:53 2008 +0300

    Convert rpmVersionComprare() to new headerGet()

commit 77c25969a97cc3fc31e7c8bdbc27a99481ae828a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 26 12:40:43 2008 +0300

    Convert triggertypeTag() to new headerGet()

commit efe1665024e374a8f8704d229b1d306af232c396
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 26 11:42:57 2008 +0300

    Convert compressFilelist() to new headerGet()

commit b5efc4ce63f3c166d551a8710ac1c6696f5adf49
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 26 11:13:30 2008 +0300

    Convert expandFilelist() to new interfaces
    - use headerGet() with extensions enabled instead of directly calling
      rpmfiBuildFNames() and flip the tag to OLDFILENAMES

commit c9f4cb186e2bfd66154e358fbff2abccb6e654dc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 26 11:01:25 2008 +0300

    Convert headerGetNEVRA() and headerNEVRA() to new headerGet()

commit 965f806bcee5ce615033d41a35ea8a29f3b7be71
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 26 11:00:06 2008 +0300

    Treat epoch as unsigned integer

commit 94d3103d7bd299f2dee31e3002cfb7c0ca030389
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 26 10:29:30 2008 +0300

    Kick out fpLookupHeader(), it's been unused since 1999...

commit ec3ab45ad2b867d27af5e5fc5e13a955933675e8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 26 09:32:57 2008 +0300

    Deprecate headerNextIterator()

commit bc84f12def0fba3ba4e7d3963a1de16a32c3b6da
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 26 09:27:02 2008 +0300

    Convert hsaNext() to new header iterator interface

commit 775f4217560ac2d136b5c53b6c691771af058244
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 26 09:22:17 2008 +0300

    Convert python rpmMergeHeaders to new interfaces

commit 276994ca46697f5660ff92a3b4329b2bc3c83eb6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 26 09:15:54 2008 +0300

    Convert python hdrKeyList() to new header iterator interface

commit 1071f0150253b38d1406b8a85a4f06d77de4ec5b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 26 09:04:28 2008 +0300

    Minimally convert rpmVerifySignatures() to new interfaces

commit 5f061ca214578f41dc93aa6c75a78cb4006f7e33
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 23 16:59:04 2008 +0300

    Unbreak DSA signature trailer handling
    - RFC 4880 states the following:
     V4 signatures also hash in a final trailer of six octets: the
     version of the Signature packet, i.e., 0x04; 0xFF; and a four-octet,
     big-endian number that is the length of the hashed data from the
     Signature packet (note that this number does not include these final
     six octets).
      ... but we were using size_t, whose size is  platform dependent, oops :)

commit 071692bb84ca70667f1b3d756145087baac7245c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 23 13:47:54 2008 +0300

    Convert rpmResign() to new header iterator/get interfaces

commit 7c0cea411a1f8889c58b1a0ed59b3a9984c1fbfe
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 23 13:46:21 2008 +0300

    Don't crash on package without RPMTAG_OS
    - damn pubkey "packages"

commit f00a37f33877c9cf59f1b0a90f3851eb4ffdd1e7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 23 13:11:13 2008 +0300

    Convert readFile() to new headerGet()...

commit 1e4e1a4bee41dee73b72ff93c2b7e9a5513bd030
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 23 13:01:47 2008 +0300

    Convert getSignid() to new headerGet()

commit e31d9e9e9449a31d96e98a8eb39304710e7c649e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 23 12:49:04 2008 +0300

    Convert the rest of rpmReadPackageFile() to new headerGet()

commit 2110a17195a51fdb95212d603a8fc47d63c47041
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 23 12:37:00 2008 +0300

    Start converting rpmReadPackageFile() to new headerGet()

commit a4ea9a0e83ce62f00e61015f312c219fee36ca7a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 23 12:25:51 2008 +0300

    Convert headerCheckPayloadFormat() to new headerGet() interface
    - otoh the function is just dumb and should die but...

commit 16c23df21b4879665d1733ed13a9a4e9344369c4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 23 12:19:16 2008 +0300

    Convert rpmReadSignature() to new headerGet() interface

commit 3eee3e56eca3058abe88e2e213d37a889af38db5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 23 12:10:14 2008 +0300

    Convert makeHDRSignature() to new headerGet() interface

commit 3a9ff8fc41dcf3fc5b4ca3eff42d389ec35c1089
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 23 11:54:08 2008 +0300

    Convert rpmtsFindPubkey() to new headerGet() interface

commit 9c209c8a2de5adb741f90c37a59c99d3578f4bdf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 23 11:49:56 2008 +0300

    Add rpmtdGetIndex() + rpmtdSetIndex() methods

commit 9979407567cb42df9873f83a4dca779073759296
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 23 10:31:45 2008 +0300

    Add rpmtdSetTag() method for setting (or changing) container tag + type
    - permit change on non-empty container to compatible type to allow things like
        headerGet(h, RPMTAG_FILENAMES, td, HEADERGET_EXT);
        rpmtdSetTag(td, RPMTAG_OLDFILENAMES);
        headerPut(h, td, HEADERPUT_DEFAULT);
    - empty container can be set to any valid type

commit 3249b289e8d9f6bcebf38e451e25b4540da3c63e
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Fri May 23 07:49:43 2008 +0200

    fix handling of omf dirs (rhbz#448033), thanks to Matthias Clasen

commit c227b9ebf9c942aea232d209dbb4d276227c7341
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 22 11:25:15 2008 +0300

    Remove some leftover doxygen @todo markers

commit c08dd6b7c0be0062a262d575c0643cab7dafcf5b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 22 10:14:40 2008 +0300

    Fix printing of patch numbers, doh :)

commit 973ffa42db3cc68ec1a5cd4d8e64b8f0e95e68cc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 22 10:03:04 2008 +0300

    Include tagexts in potfiles
    - nothing to translate there at the moment but..

commit f81f3f0fc5a95d8fcd6e7b1ca2a05f81a2ae4065
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 22 10:01:23 2008 +0300

    Add tag data container api to librpm + hacking docs

commit c4614940c04c6ce88dd4d5a3959ec1175a91fe0e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 22 09:18:48 2008 +0300

    Oops, rpmTagGetValue() expects tagname without RPMTAG_ prefix

commit d1ad87a7f868eec726c8fb2214090e783e171380
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 21 17:59:05 2008 +0300

    Eliminate rpmTagTable internals from API/ABI completely
    - everything is now accessible through other methods, no need to expose
      our internals
    - tagtbl.c is now included from tagname.c instead and not separately built

commit 058e94ec41130fa675aabd97088d9f286148286d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 21 17:26:52 2008 +0300

    Ditch the tagtable argument of (internal) headerDump() function
    - might as well ditch the function too I guess...

commit c06d688a3929742691eae2eb49643b2bd1d013c1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 21 17:13:51 2008 +0300

    Don't muck with rpmTagTable internals in python initialization

commit 2525d192304246d366e08974ccfabb6bfdfc5ff4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 21 17:03:15 2008 +0300

    Dom't muck with rpmTagTable internals in rpmDisplayQueryTags()

commit 82a7df65b5f173693a9be140704e75cc94a77b25
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 21 16:34:27 2008 +0300

    Remove double doxygen documentation

commit c62b69b4fa71416510aad06314eb29cf811abf02
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 21 16:32:37 2008 +0300

    Add rpmTagGetNames() public function for retrieving known tag names
    - use rpmtd as container to avoid inventing yet more datastructures and
      iterators

commit b6d8ac89b8018e8942fb63a8e6c73f9e3f817ef0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 21 16:05:29 2008 +0300

    Include stdlib.h for free()

commit a63c97eb123b54786c7b964906f7b034985ef100
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 21 16:02:10 2008 +0300

    Slight tweak to rpmvercmp() docs as suggested by Pixel

commit d2ef49038a12adb33329c5784b76d6b20e2e3ce0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 21 15:59:39 2008 +0300

    Include rpmtypes.h first instead of rpmtag.h everywhere

commit 3b59f236b333c171b8497363309daffea165878c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 21 14:51:25 2008 +0300

    Convert headerGetColor() to use per-datatype iterator

commit 35c0252857822faac53b7df9c9632d21bed926d5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 21 14:49:12 2008 +0300

    Add per-datatype iterators to rpmtd
    - just uint32 and string iterators for now, those are the most common ones
    - allows somewhat more convenient code constructs when type is known,
      eg no need to separately check for non-null inside loops

commit 3d8656f040d2c780c88d0cd63594898f0605a6fa
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 21 13:37:32 2008 +0300

    Convert rpmfiBuildFDeps() into fooTag() style interface
    - fileprovideTag() and filerequireTag() are just thin wrappers around
      filedepTag()
    - simplify things a good bit by not bothering to create a single allocation
      blob for the return

commit 0021f8a18fe941aaca56aeb0efa01f387ab9a42d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 21 12:45:34 2008 +0300

    Lose rpmfiBuildFClasses() completely
    - only fileclassTag() needs it, move it inline there
    - simplify a lot: as we now can deallocate regular string arrays too,
      don't bother creating a single allocation blob for the classes

commit 1c80801ab60e4f67b7b201dd983ed33d4297db38
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 21 12:41:31 2008 +0300

    Misplaced assert
    - we only care about non-NULL data if we're supposed to free the contents

commit 82b89c3b0c78b0df871879d6c9bfe4af1b5e3b9a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 21 11:50:27 2008 +0300

    Eliminate rpmfiBuildFClasses() and rpmfiBuildFDepts() from the API+ABI
    - both are only needed by tag extensions, move them there
    - headerGet() with extensions enabled provides a nicer API for accessing
      the data anyway

commit 7d38c6e2d80492ba32814e4ad1175247fab3d931
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 21 11:45:46 2008 +0300

    Split tag extensions out of formats.c
    - they have practically nothing in common now..

commit a668b84c95d04dc0c6cd70b93a58938e8e798b7d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 21 11:04:59 2008 +0300

    Minor fixes to triggercondsTag()
    - lose unused variable
    - remember to reset versions iterator too

commit bda56cc4376c34a87ac96df06be72d64e04d03ee
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 21 10:40:50 2008 +0300

    Resurrect triggercondsTag()
    - convert to use new interfaces and simplify

commit ce70ac871764ec835d5c83b3115d3a7236196838
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 21 09:22:38 2008 +0300

    Add rpmtdDup() method for deep copying of tag containers

commit 34151a9e35122d90bdc1863dc5dc5e1fec786afe
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 21 09:02:10 2008 +0300

    Turn rpmtd freeData into flag bitfield, allowing more precise control
    - headerGet() returns even array data in a single allocation block, but
      regular (string) arrays need to have the pointers in array freed too
    - set immutable flag on header data

commit 3a12552a3a9d1389bab08d3ee1161b5d1c654565
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 20 16:21:34 2008 +0300

    Convert mireSkip() to use new interfaces
    - let rpmtdFormat() figure conversion to string, simplifies things a lot

commit 7c3d5dc15d66367fc5fbf89815f514d381e3d16f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 20 12:38:40 2008 +0300

    Add arraysize format
    - it's a bit silly but avoids ugly special casing in formatting

commit a9920b929913ccd627b817476dddd45aea1fd6b7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 20 12:18:14 2008 +0300

    Fix up for RPM_BIN_TYPE td->count abuse just once, in rpmtdCount()

commit 5d10e43f8f8a1cb8041326bae8251f8e0ba311f0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 20 12:09:14 2008 +0300

    Drop unused variables from formatValues()

commit fc38b84c010fa8c4f2617b1351ceaccec9f5da60
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 20 12:07:38 2008 +0300

    Lose the useless rpmtdToString()
    - either use rpmtdFormat() or xstrdup() on rpmtdGetString() as appropriate

commit fae1edeb939e761a7fd070081d378a2a67a485f9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 20 11:57:30 2008 +0300

    Lose padding argument from formatters
    - formatPrefix already contains the padding, just use rasprintf() on that
      to get properly allocated buffer without messing with it manually
    - xml format doesn't honor padding anyway so don't bother with it

commit 845eb8fc70493da08e9cf31749796dbb976aa657
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 20 11:56:47 2008 +0300

    Avoid leaking memory on legacy header interface emulation

commit 86bd819cee03899bb35b5edc99ccc0831755f388
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 20 10:16:52 2008 +0300

    Treat all integer types as unsigned in formatting

commit efa130cee8c209ecd6bdf8ad51c3ad26579cf37e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 20 10:10:18 2008 +0300

    Simplify xmlFormat()
    - use rpmtdFormat() to convert all data, xmlFormat() only adds the
      xml markup now

commit 9b32e58e173799b750a12c1bac51a7d49aadd4d6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 20 09:40:24 2008 +0300

    Add rpmtdFormat() method and enumeration of supported formats
    - permits formatting any rpmtd data to our supported formats over
      iteration

commit d5381c16a794e03c0e4f202daa3c882dc1876f41
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 20 08:53:05 2008 +0300

    Handle RPM_CHAR_TYPE too in stringFormat()
    - used by filestates and nothing else apparently, duh..

commit 2bfc0fad3876bcee0143664cae9f0f0107a86a33
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 19 17:51:42 2008 +0300

    Add stringFormat() for barebones string representation
    - use for everything not specifically overridden
    - pushes practically all formatting out of formatValue() into formats.c

commit f12983793ec4683e5083917f97169d23292c7ef7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 19 16:54:29 2008 +0300

    Make formatter functions operate on rpmtd current item
    - avoids extra abuse of the element field and simplifies things a bit
      (eventually, not really yet)

commit a40e3266836131ba5213d3b97d0f7ae93dedb755
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 19 15:56:34 2008 +0300

    Push rpmtd a bit deeper into formatting
    - use rpmtd methods for pulling current integer/string values
    - allows collapsing string and string array handling into one case
    - CHAR/INT8 type aren't handled atm, but then nothing at all uses them...

commit 91c13fbf4f5a3dcdd19be82698157d08f522a11a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 19 15:24:24 2008 +0300

    Minimally convert formatting to pass around rpmtd

commit 246c2e2be254bc2f1014e55ccab91066b90aabea
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 19 08:38:53 2008 +0300

    Remove unused datafree handling from formatValue()
    - the data is always cached and cleared elsewhere..

commit 1a11aa66407ac40325d50e6a138c7f36672cc2ae
Author: Jindrich Novy <jnovy@dhcp-lab-186.brq.redhat.com>
Date:   Tue May 20 14:07:10 2008 +0200

    Ignore quotes in expression

commit 75f62d7c5fddf9d7e0cb588ac59d899aa426b60c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 19 08:27:10 2008 +0300

    Teach rpmtd about the raw i18n string (array) type

commit 405bbc571ff5273c5bebdec088631c4ac2c1b32b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat May 17 20:15:53 2008 +0300

    Move all i18n string munging into separate function

commit 7fedc2057030e46e1421fa5d14278c2d7bc49a87
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat May 17 09:50:39 2008 +0300

    Add HEADERGET_RAW flag to disable i18n lookups

commit 89b8f53621f0dcb9954bd167aa83fd60a95836a8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat May 17 09:37:32 2008 +0300

    Push header getter flags down to internal helpers

commit ea0cdcc0dd4ffa17e7f0675b9a4eaf542e2de46c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat May 17 09:28:32 2008 +0300

    Oops, thinko/typo in headerGetFlags

commit 0c75369d4681150d986df24e3414b86ccf1f4fb1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat May 17 09:25:04 2008 +0300

    Add typedefs for header get/put flags, use instead of int

commit 7be1b1bafcf1dd6a8464240d5fefa00cfd78c084
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 16 18:17:53 2008 +0300

    Remove unused hRET_s and HE_s definitions for good

commit 72e4b1ae7e5dfb67aec9c4dafb0e5bc9413d7c84
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 16 18:15:08 2008 +0300

    Use rpmTag enumeration for finding tag format func instead of names
    - comparing strings when we only care about the tag number is waste of time

commit 019ee37f6279cc6e114359224ace85cb2260f6e0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 16 17:46:49 2008 +0300

    Eliminate rpmHeaderFormats and rpmHeaderTagExtensions from the API

commit cbf1a479ee62a5538065b92dcd8d9eb8e842bdda
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 16 16:56:32 2008 +0300

    Push finding of formatter and tag extension functions down to formats.c

commit eef5417deabb40304cc225ded2ebc7bf48dc2c46
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 16 16:26:19 2008 +0300

    Make extension formatter functions opaque
    - tag extensions are only needed by headerGet(), formats only by headerFormat()

commit 696184533aa34b7ae29c5567e1255f00505d884f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 16 16:07:41 2008 +0300

    Eliminate the now unnecessary extension type fields

commit 44bdb776b7d557941d377d0adfdedd2940546164
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 16 15:55:44 2008 +0300

    Split format and tag extensions to separate tables

commit a81863d443f3916d9562009fb5374e20b7fbc70e
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Fri May 16 06:05:25 2008 +0200

    Update Portugese/Brasilian RPM translation (rhbz#446795)
    - thanks to Igor Pires Soares

commit 62f53dbc66fd4accb0fd1be1245f197ccfe52c2c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 15 18:52:18 2008 +0300

    Dumb segfault in %posttrans

commit 504ff9ce0677143a586c4e2baef5cff7d8bf81f1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 15 18:28:50 2008 +0300

    Unbreak conditional formatting

commit 7df706ac167907559bb3acf7ac8223534a2e3f05
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 15 17:17:45 2008 +0300

    Let rpmtdFreeData() do it's job...

commit aaa93e9008d56e14eae78b882c1396a117184995
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 15 15:24:16 2008 +0300

    Don't look into extensions when adding tags to python
    - tagtable knows all our extensions too, there's no need to

commit a06b172d8d76c16011ba74ec9b60640ff07439da
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 15 15:13:03 2008 +0300

    Change headerTagTagFunction() to take tag container as arg
    - minimal, crude conversion of all formatTag() functions for now

commit 25edcbcf4b5a420ef101c9622d0c4bcb6bcfd464
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 15 14:55:59 2008 +0300

    Minimal conversion of in-rpm python bindings to new headerGet()
    - hopefully preserving legacy quirks and all...

commit 5198c91280ba07959b3efe5605a1e1a697c17d40
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 15 13:42:12 2008 +0300

    Eliminate unnecessary extension fields
    - push default extensions yet deeper into format parsing
    - single bounds-checked function to retrieve from cache

commit 6f18834d4ebbea4af8c2b3be12f25d28fe4f4450
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 15 12:27:20 2008 +0300

    Permit tags down to HEADER_IMAGE in queries to match previous behavior

commit 445671ee82d95b7465663152a633c38bd3300d5e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 15 12:04:17 2008 +0300

    Unify header and extension tag handling of headerFormat()
    - retrieve both header and extension tag data through headerGet,
    - cache the return data for both types for an enormous speedup especially
      for array types

commit 91540f53598a1d53448c56b7153634d9c961f630
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 15 08:01:45 2008 +0300

    Convert rpmtsAddInstallElement() to new headerGet()

commit bbb566a261ae5fbafe770beabf9a0bcc5487e5ab
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 14 15:08:40 2008 +0300

    Commentary fixes
    - "return" is not our registered trademark :)
    - document the intent to make rpmtd opaque

commit 65100c8e1d8c4eea4c0b410a9427d1bb76cba8fe
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 14 14:36:36 2008 +0300

    Teach headerGet() to retrieve extension tags if enabled with flags
    - bit of a kludgery until fooTag() formatters have been converted
      to take rpmtd struct as argument
    - idea lifted from rpm5.org, implemented independently (and probably quite
      differently)

commit b94a187adcdb7c32983f5c6a5718fb4fa2f87bc5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 14 13:26:58 2008 +0300

    Simplify rpmDisplayQueryTags()
    - all relevant tags are now visible in tagtable, don't bother looking up
      extensions separately

commit 7bb210e0af1570d05eaa267b3b96344df45a4871
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 14 13:20:46 2008 +0300

    Teach rpm tag table about extensions
    - add rpmTag_e enumerations for all extension tags
    - add extension field to headerTagTableEntry_s for things that are "pure"
      extensions (instead of just extension overrides) so we have a chance to
      catch out anybody trying to insert such tags to headers

commit c0bdbb0946203ec2f53c6eff3f64558a2876e6a0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 14 10:29:07 2008 +0300

    Convert headerRegenSigHeader() to new header iterator interface

commit df23089b1f0e15e4e0adc4a97c143b7cf9680127
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 14 10:20:56 2008 +0300

    Convert headerMergeLegacySigs() to new header iterator interface

commit c66f0b5008ab0088e83a4ca69ddbe7d3ef9def55
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 14 09:46:04 2008 +0300

    Convert file color calculation in rpmfcGenerateDepends() to new interface

commit b31eef9e9b22a3842b8937f17aa2aae6a6e50dad
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 14 09:31:24 2008 +0300

    Convert headerGetColor() to use new headerGet() interface

commit 7e0f6408bac7ae3eac004c8224c97fd83cb021a7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 14 09:20:51 2008 +0300

    Convert addTE() to new headerGet() interface
    - streamline things a bit while at it...

commit 49a892fae1df0f7abdcffa387aa347c9e8b4a299
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 14 09:12:40 2008 +0300

    Convert rpmteNew() to use new headerGet() interface

commit abbf9a2c8d9504d917da2c7d53da5758cd55e4e9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 13 18:12:13 2008 +0300

    Simplify rpmInstallLoadMacros() a bit
    - use new style headerGet() and let rpmtd work out the string presentation

commit 30c4e29d33b9f7c4ef2011a1d84bbf7b3b156ea6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 13 18:09:35 2008 +0300

    Add some more rpmtd access methods
    - rpmtdGetUint() for 16 and 32 bit integer types, similar to
      rpmtdGetString() (equally usable with scalar types and arrays)
    - rpmtdToString() which converts "any" header data into string presentation

commit e7ec2339b42b9be6bad373d7ced4d351ce0db08a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 13 17:35:32 2008 +0300

    Convert runInstScript() to use new style headerGet()
    - simplifies things quite a bit when we can just iterate over the
      prog arg container, no matter if it's string or string array

commit 2d0a4329bddd8a644e87d8b00399096423cef2f5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 13 17:10:42 2008 +0300

    Simplify prefix shuffling in psm runScript()
    - use new style headerGet() and pass the entire tag container down to
      doScriptExec(), removing need for big bunch of local variables
    - use rpmtd (iterator) methods for processing the prefixes in doScriptExec(),
      it does the right thing for legacy non-array RPMTAG_INSTALLPREFIX too

commit 56878483b7814ef4ef544783dbea2d833af6758c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 13 16:12:56 2008 +0300

    Convert a few cases in rpmfc to use the new rpmtdFromArg() methods
    - safety asserts for the expected type for now...

commit 11980a42f28feddf33c0b2c74b6424b217404263
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 13 16:11:28 2008 +0300

    Add methods to construct tag containers from argv and argi arrays
    - basic type checking done based on tag (return) type

commit ae218d25aee90d967b3b928b332cb759f27cc22e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 13 15:05:22 2008 +0300

    Convert couple of headerGetEntry() uses to new interface in parsePreamble

commit bf84c980ed948f16a55ecd9469d8c579667ff3a0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 13 14:42:32 2008 +0300

    Use tag container iterator for ismemberInEntry()

commit 9de4354a1258abf81963f450a654275cfbc08c33
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 13 14:40:55 2008 +0300

    Add support for iterating over tag data container
    - rpmtd iterator init + next methods
    - string accessor method for RPM_STRING_TYPE and RPM_STRING_ARRAY_TYPE

commit 7c4cae7030d13ec43d1d31d16e598cd994410516
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 13 14:03:08 2008 +0300

    Convert checkForDuplicates() to use new interfaces
    - doesn't gain anything but ...

commit 223da12f24cc0e1e0a878ab7dc420df468d9578b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 13 14:02:40 2008 +0300

    Add rpmtdTag() method

commit 362f46a05a5a4a285d764f64dce05bd93eb6f537
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 13 13:44:04 2008 +0300

    Simplify initSourceHeader()
    - use headerCopyTags() instead of manually doing it for the main tags
    - use the new rpmtd interfaces for build restriction copying

commit ed23d35805cb42099ab22b6525bc6ad08e9a5967
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 13 13:29:07 2008 +0300

    Convert headerCopyTags() to use rpmtd internally

commit e2ea6a35bacad840c5455bd7c3e91f45dcbf2cde
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 13 13:25:14 2008 +0300

    Convert headerCopy() to use the new interfaces internally

commit 0ee2364ae2a06b6c176414f13f1297e9eec32cb6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 13 12:32:25 2008 +0300

    Add headerPut() for adding data to headers through tag container
    - tag container isn't that big a win on add/append operations as it is on
      retrieval, just use the existing headerAddEntry() family to do the
      actual work
    - headerPut() is a nice and short name, lifted from rpm5.org, API might
      differ
    - flags to control operation, just whether append is permitted or not for now

commit 29087b08a6cd4d6591ee238e0b650f6d41f6eddc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 13 11:59:19 2008 +0300

    Use a macro for copying td entries in compat wrappers
    - just to avoid tedious umphteen copies of if (c) *c = ...

commit 09807fed0353fe67cb7df4db4b21e544ebc118cb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 13 11:46:33 2008 +0300

    Convert header tag iterator to use rpmtd
    - new API function headerNext() for the new interface
    - headerNextIterator() is just a compat wrapper

commit 8cd92950bda4658c8d376c5c381237301013edfb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 13 11:38:38 2008 +0300

    Convert headerGetEntry() internals to use rpmtd for data
    - new public API function headerGet() that returns data in rpmtd, with
      flags to control operation (just minmem or no for now)
    - headerGet() is nice and short name, lifted from rpm5.org but our API
      differs as this takes a separate arg for the tag to get instead of
      (ab)using the container to pass data back and forth
    - internal compatibility wrapper to keep headerGetEntry() and -MinMem
      version working

commit 752343c2b7798b1fa225abc9ade94529a6175ad7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 13 11:14:50 2008 +0300

    Start converting header internals to use the rpmtd struct
    - turn copyEntry() into a wrapper around copyTdEntry() to avoid having
      to convert everything at once
    - copyTdEntry() takes rpmtd and populates with data and whether memory
      was allocated or not

commit a83cfce188a4545756a8fa5791f3b9cbe20190ed
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 13 11:02:45 2008 +0300

    New "tag data" container struct + some basic methods to deal with it
    - to be used for passing around header and extension data
    - inspired by similar changes in rpm5.org, details and implementation differ

commit 6fd646f238b402d2e47c8af8520d3ffa48ebd91e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 12 18:32:11 2008 +0300

    Remove tags from headerSprintfArguments
    - its always rpmTagTable so no point carrying it around

commit a670a393cadca5df927afda6e1543030a1990fc7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 12 18:27:36 2008 +0300

    Update translations again for headerfmt.c split

commit e89d20336111c4ae2089d0af4eb5a651bed3aca8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 12 18:26:11 2008 +0300

    headerfmt.c has translatable strings, add to POTFILES

commit e02523286c03fdad3b5f8df4639cfd0317e496a4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 12 18:03:47 2008 +0300

    Use rpmTagGetFoo() instead of local myTagFoo() versions
    - hsa tags will always be our default rpmTagTable so we can use the
      somewhat optimized rpmTagGet() versions instead of dumb linear searchs

commit 95386fe3318cc8cbd0dc240570dfae226afd6501
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 12 17:25:54 2008 +0300

    Push extension defaults one level deeper into headerFormat()

commit fe3ba9991019dcf5de7f92577cfd04872fe7584d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 12 17:08:12 2008 +0300

    Simplify header extension handling
    - remove the now unnecessary chaining to different extension tables

commit ecd8c6c44149b880e672accc416d593c551de72e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 12 16:53:26 2008 +0300

    Move headerVerifyInfo() to header.[ch] where it logically belongs
    - removes the need for copy-paste typeAlign array in package.c

commit f081d3da6658f7ea4ea7c3ba4355d2dcd34721fc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 12 16:31:27 2008 +0300

    Eliminate copy-paste versions of hdrchkFoo() macros
    - just stick 'em into header_internal.h, both header.c and package.c
      need it anyway

commit ea9b00dab87a5d44730e6b51f72a9ed3a5432164
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 12 16:20:13 2008 +0300

    Convert internal uses of headerSprintf() to headerFormat()

commit 67b4f3a13de4012eaf0caf4b62c93cf001d9cdc1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 12 16:16:37 2008 +0300

    Add simpler headerFormat() function, deprecate headerSprintf()
    - headerSprintf() is just a dumb wrapper around headerFormat() now,
      tbltags and extensions are not used at all
    - baby steps towards eliminating the the tag and extension tables from
      the API/ABI...

commit 05b70ba53c12ee70975586e1ed0368cd3442d05c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 12 15:48:31 2008 +0300

    Split header sprintf formatting to separate source
    - header.c is now just low level header handling, headerfmt.c is the
      formatting engine and formats.c has the actual formats
    - also move bunch of formatting-related stuff from header_internal.h to
      headerfmt.c, nothing else needs it

commit 0ea9805e2dd21a1b05343f8aa81ab91232ceda48
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon May 12 12:46:37 2008 +0200

    fully translate to CZ after latest po files regeneration

commit 5dcb97da2bfcee68108824a37c7a788193e3004d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 12 12:54:36 2008 +0300

    Move header "default" formats to formats.c along with the rest of 'em

commit 32bf84651419d071ec1d940dac1dfd3f9c807422
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 12 12:31:54 2008 +0300

    Update translations after the big shuffle

commit c57c62b94aaf8032b144a3766f311d3893d7bc8d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 12 12:30:00 2008 +0300

    Adjust pkg-config now that librpmdb is gone

commit 90e10708f824ff8e86b0f030d12e14329b1cf2c0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 12 12:28:12 2008 +0300

    Fixup paths and Makefiles for rpmdb/ elimination

commit a68a5a049e98bc8406c0e7015ad5dadbed39972d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 12 12:17:01 2008 +0300

    Start (re-)joining librpmdb and librpm
    - the split is largely arbitrary and mostly just causes, well, arbitrary
      issues and limitations
    - just move content from rpmdb/ to lib/ for now

commit 962c7b49d2c290678afad9a9893f76450e655108
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 12 10:45:33 2008 +0300

    Rip old, bit-rotten librpm test programs

commit 84b29174e6a8da103e8afbffa48dc10958cc2e62
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 12 10:44:07 2008 +0300

    Rip old, bit-rotten librpmio test programs

commit 2f890bda8b8685167433bf04c515d488555b1436
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 12 10:40:14 2008 +0300

    Rip old, bit-rotten librpmdb test programs

commit e13edf837af856a76aa6e7ea7171a2c86546e39f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 12 09:14:09 2008 +0300

    Remove internal copy of Lua for good
    - we can do just fine with externally linked lua, no need to carry
      this around

commit 88e2895bec3be69772ca79c30bd417cd59cbe668
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Sat May 10 08:55:04 2008 +0200

    Fully transale RPM to Czech
    - convert the po file to UTF-8
    - translate all strings

commit 06a0d228d2c190a8bc59fe897ca980e5f98ef8bd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 8 19:03:12 2008 +0300

    Caught pants down with strlcpy() vs strncpy() semantics
    - rstrlcpy() calculates source string size regardless of length limit,
      but cpio fields are fixed length character arrays, not strings -> kaboom
    - wondering if zero-terminating strncpy() variant might be more fit to our
      purposes than strlcpy()-clone

commit 6a8ddd9bc22cd8bf7d5100da9915cdae91549a3b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 8 17:10:10 2008 +0300

    Eliminate static print buffer from prDbiOpenFlags()
    - unlikely to actually overflow but it'd be all the more embarrasing
      as it's just diagnostics/debug code...

commit aa3ab021a8b7c633ddacca71adc1b86038bff3ff
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 8 15:24:46 2008 +0300

    Avoid entirely unnecessary static temp string buffer

commit 4ad7f40395be970b7d84cb9782ce6c7717db9d37
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 8 14:37:29 2008 +0300

    Rip useless ROOT_GROUP checking from configure
    - it was only used for an ancient RH-specific %_fixgroup anyway which
      in turn only ever did anything when building as root - bad bad...

commit c9b6fc7e2d3d7395346acf3e3092d22474d82d55
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 8 14:36:08 2008 +0300

    Don't run fixowner and fixgroup at end of %prep
    - they'd only work for root, and we don't want to encourage people to build
      as root...

commit c5696f60193c31d98f1e313145b3611de3d34906
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 8 13:40:26 2008 +0300

    Oops, matchpathcon() wants st_mode, not the entire struct stat

commit 8ecf69ed6236c308f183125301fa571c4e21491f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 8 13:35:27 2008 +0300

    Block signals while moving the rpmdb to final location after rebuild
    - interrupt there would be pretty disastrous..

commit 0af31c915e01b08f6efb044ac24a799c6b375410
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 8 13:30:16 2008 +0300

    Lose the silly db argument from block/unblockSignals()
    - unused, and unusable anyway as signals are global, not per db thing...

commit 485ee9bbb4ff3e9c3b4cb65052055d4e8d47e78f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 8 13:06:34 2008 +0300

    Restore rpmdb selinux contexts to default after rebuild (rhbz#442149)

commit 461fd000d22d76d97ad0ac76263477d364691d32
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 8 12:40:37 2008 +0300

    is_selinux_enabled() returns 0 if not enabled, not -1

commit 6ca5373aa1f58ccc60e844bd8fd006a738670883
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 8 11:53:42 2008 +0300

    Check for transaction root sanity in rpmtsSetRootDir()
    - only regular paths starting with / permitted
    - change prototype to permit return codes

commit 6006265faf97316d9685b0dc958770f50b954a3e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 8 11:38:52 2008 +0300

    Set ts rootdir to /, not empty if rootDir arg is NULL
    - lets see what break... ;)

commit 42d0c3bd685ddbc093c3e9c8449c4cb015923f25
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 8 11:27:58 2008 +0300

    Only append trailing slash to ts rootdir if rootdir != "/"

commit 8a024c20e79c76bcf02f6af3f6348cc25e62b646
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 8 10:20:39 2008 +0300

    Fix up some fuzz in Finnish translation

commit b5a6e815863a10d953ad3e857dc495d51334cc47
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 8 10:07:31 2008 +0300

    Fix up / remove some fuzzy German translations

commit 2c15e8409325f9179a68f7cc0da3668eee093198
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 8 10:00:04 2008 +0300

    Updated Serbian translations from Igor Miletic (rhbz#441101)

commit 93208f1552bea2a5802ec4b3b03c4014ab680da4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 8 09:56:54 2008 +0300

    Updated Polish translation from Piotr Drąg

commit 41a7b93e6b764e9b2966ee61cb9802b6fd7064af
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 8 09:53:36 2008 +0300

    Updated German translation from Fabian Affolter (rhbz#438178)

commit cb36d48351ea944d445db29635750042f445ec3d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu May 8 07:48:51 2008 +0300

    Check rpmAddSignature() return codes (rhbz#442761)
    - headerGetEntry() on RPMTAG_HEADERIMMUTABLE already caught the corrupted
      package, we just didn't paying attention
    - make the hge failed error message a bit more meaningful

commit 79dd50f94d94158e03fd57a72a896f8fc20f61ec
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 7 12:15:22 2008 +0300

    Oops, don't access variable after freeing

commit 31724be2f63b286d8684d72d62a7a19d1233143b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 7 11:33:33 2008 +0300

    Oops, plug leak introduced by previous commit

commit 15e9e1ff64aaf986ab0dd97b9b12c3d4fb22a484
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 7 11:12:17 2008 +0300

    Simplify doPatch()
    - use %{uncompress} macro to pass files to patch always
    - gains support for lzma patches for free

commit fc36f322e123a7421b798f12358cc701e28afbe0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 7 09:48:39 2008 +0300

    Plug another memleak from genCpioListAndHeader()
    - free fuser + fgroup info instead of NULLing, add comment on side-effects

commit 16d0a2fa29df88689fc56ef271a2538a5ca1cba6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed May 7 08:32:36 2008 +0300

    rpmfi actions is already alloced by rpmfiNew(), don't leak memory

commit e907b912f9e237bc840d714e4e03f4d26deafe0c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 6 12:48:09 2008 +0300

    Eliminate nasty hackery with "special" doc + docdir format
    - figure out docdir early on in spec parsing, store in package struct
      instead of abusing parse buffer
    - fixes memleaks from docdir_fmt and headerSprintf too

commit 41147e5b18ad09b04697c3aa729539689547281c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue May 6 10:00:07 2008 +0300

    Eliminate static buffer in lua findkey()

commit e5df40e683552c4f5b36f029ff5bced39058452b
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue May 6 08:47:18 2008 +0200

    Nuke EXPRBUFSIZ, it is now useless

commit 90eec724a834c26131be11145a62084efe3bafb8
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue May 6 08:46:50 2008 +0200

    Don't use static buffer for parsing an expression

commit 697f27bbbab27a2f353803bf765f334b8668b59d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 5 13:52:19 2008 +0300

    Indentation fixes

commit 3e991bc40c4f7818d720649901993cec95056540
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 5 11:53:48 2008 +0300

    Eliminate static sized string buffer in parseScript()

commit 39168e47aba5c953ecd3bde7134c577b31871bf9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 5 11:37:08 2008 +0300

    Trap RPMRC_FAIL returns from spec parse with abort() for now

commit d0ea38a4cae75a6c69536e3949e230f3999ae7d1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 5 11:31:09 2008 +0300

    Make readLine() return PART_ERROR on errors

commit 557c2673f7e7ff222f8abcc35ecd601007374b6f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 5 11:27:49 2008 +0300

    Fix (hopefully ;) signaling for line continuations on spec parsing

commit 3c5f581df6be0776b1d49e94d518c58e93a886c2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 5 11:20:13 2008 +0300

    Indentation fixes

commit 7b80971078c7c278d80243bfde3afaa7ce540eb8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 5 11:15:12 2008 +0300

    Make parseSpec() return PART_ERROR on all errors
    - also fix memleak from buildroot errors

commit 2ecee0c334d95fafb9ad423e998482cd1209d0e4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 5 11:00:45 2008 +0300

    Make parseFiles() return PART_ERROR on failure
    - streamline exits by assuming failure

commit d1e30d48756f3cbbc8cd4d6470880b3c089eab00
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 5 10:56:55 2008 +0300

    Make parseScript() return PART_ERROR on errors
    - streamline exists by assuming failure

commit 64527b4dafec9064f2964dbef949a5af3854636d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 5 10:46:43 2008 +0300

    Make parseDescription() return PART_ERROR on errors

commit 6e0450da248157ae0c88fbd6ffba842cd461c9f2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 5 10:42:41 2008 +0300

    Make parseChangeLog() return PART_ERROR on errors
    - streamline exits, fixing memleak from stringbuf on various paths

commit 4e1e3b4aa9e2a3c29d67d889461284994feceffa
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 5 10:34:33 2008 +0300

    Make parseBuildInstallClean() return PART_ERROR on errors, streamline exits

commit 6e1e700a251cd91b1177b56ae74fc52ea18b0432
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 5 10:24:22 2008 +0300

    Make parsePrep() return PART_ERROR on errors

commit e0787e828b0a73b96d8928586d5a15d30a3c61cb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 5 10:08:33 2008 +0300

    Streamline parsePreamble() exists
    - assume failure, single point of exit
    - fixes memleak from NVR
    - use PART_ERROR for error exits

commit dc61ca3b65f74e850184ead6e6985842ae7b767b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 5 09:51:14 2008 +0300

    Start cleaning up the rpmerr removal mess of build/ return codes
    - define PART_ERROR for error returns from various parse bits,
      teach parseSpec() about it
    - streamline parseSpec() exists

commit fc728e489f707ad1823e55b311944f1dff8ec65e
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon May 5 08:12:04 2008 +0200

    Remove obsolete parseForRegexLang()
    - no use for it since _langpatt is now gone

commit 25de12b124120748c832e5167f6d2551a26937e3
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon May 5 08:05:50 2008 +0200

    macros.in cleanup
    - really deprecate _langpatt
    - obsolete _package_version
    - use paths more conforming to FHS
    - fix typos

commit 8e1cb253db82d4f26eb0455f985771924eb5ecca
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 5 08:38:57 2008 +0300

    Use fprintf() for writing PGP passphrase (like we do for GPG)
    - avoids having two separate write()'s to check
    - ...if we actually checked the result, gcc doesn't whine about
      unchecked fprintf() return...

commit f78fa317377e889ac59635abc556a92828b5c794
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 5 08:27:07 2008 +0300

    Check for pipe() returns on GPG and PGP signature creation

commit 12bb7024fd900ef5fc4468293a41b5cf163b8cdf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 5 08:23:24 2008 +0300

    Check for pipe() returns in rpmfc helper output reading

commit b15e3e3ca1f09163fcb392a15fefb36581b101e2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 5 08:15:17 2008 +0300

    Silence bogus warning on memset()
    - gcc doesn't realize shrank == 0 would terminate the process via error()
      before reaching memset(), work around...

commit 961b6c4b2f9c2cd4663f85fdb38906765e3640aa
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon May 5 07:37:50 2008 +0300

    Check for pipe() return code on --pipe

commit 25136233dbd286c23b21689b9300001a3137f378
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat May 3 12:44:16 2008 +0300

    Add missing include

commit 015d99725f98e0bcce86faf20309a167689ac0e4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat May 3 12:43:37 2008 +0300

    Initialize nb before use

commit 8c1038edc355f63e89588424bf23b9cee8a26a4f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat May 3 12:38:02 2008 +0300

    Create signature target rpm in same directory as original
    - rename() doesn't work across devices, oops...

commit 0ee67fede81600dad90da5139f6109399d6926e7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat May 3 12:34:19 2008 +0300

    Temp file handling tweaks
    - rename rpmMkTemp() (back) to rpmMkTempFile()
    - rpmMkTemp() is now a lower level thin wrapper around mkstemp()

commit 87ea239ffc5763b1d85ce1b0d35759d9bafe1cb5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat May 3 10:36:00 2008 +0300

    rpmReSign() doesn't return rpmRC codes, make it more obvious
    - cleanup "rc = RPMRC_FAIL" copy-paste leftovers

commit 5f607d565d583b26347d3560e15ab3258596220a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat May 3 10:27:16 2008 +0300

    Eliminate static sized string buffer frpm rpmReSign()
    - use rpmMkTemp() for creating the temp file instead of insecure
      close(mkstemp())

commit e5972a715ad84a71b372e167fe0dadb74e6d9ab6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 2 11:48:27 2008 +0300

    Check for arch compatibility on multilib too

commit 3e6009a7440ea5fc3edc21bfe299288018e4e3ea
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri May 2 11:40:29 2008 +0300

    Ensure default SIGPIPE handler for --pipe (rhbz#444389)

commit a306a04a895a1632ad80753c61e57864f0019441
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 30 14:38:09 2008 +0300

    RIP alloca()

commit c089a45bf6100b9df9cf0452759c7c46ea3bbc5b
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Apr 30 13:24:51 2008 +0200

    Rewrite xmlFormat() and remove useless xmlstrlen() and xmlstrcpy()

commit 995ebf4b588e092ed008c2757ed84713d9e1e61a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 30 13:34:28 2008 +0300

    Kill of alloca() from tufdio tests
    - yes it leaks, who cares

commit d6964535c9647ee74972e053b9d0607f9b4b0693
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 30 13:30:54 2008 +0300

    Lazily copy basenames in relocation if needed
    - couple of tiny array helper funcs to push stuff out of relocateFileList()
      instead of adding yet more gunk
    - use the helpers for dirnames too

commit b7aab95d58110fcab8dc9fe34d36be0f86b8398f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 30 12:45:16 2008 +0300

    Simplify relocation directory name manipulation

commit e454180a3f6bc16431ba710f5a16aca83f1c5396
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Apr 30 11:39:49 2008 +0200

    Don't use alloca in rpmal.c

commit 355c673016632c1218860b5df82c689ec535cf9d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 30 12:29:40 2008 +0300

    Allocate relocation dirnames on heap, not stack

commit b28bea588b533d310b877f641e0fd9ae6ca143d4
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Apr 30 10:32:33 2008 +0200

    Don't use alloca() in fpsCompare()

commit d4d4a29f8dd2497607fd375f4966abb48a1c522c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 30 10:32:57 2008 +0300

    Begin untangling relocation alloca()'s
    - allocate rpmRelocation paths on heap, not stack

commit 657dc38ec051d3333dad5cb15c6cb7a884c151e4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 30 10:05:48 2008 +0300

    Kill the remaining alloca in fp doLookup()

commit 782c4335c350d96695b1193a9fa1bb394c155037
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 30 09:47:25 2008 +0300

    One alloca() down in fp doLookup()
    - single point of exit from the point where buf gets allocated

commit 2bd9d3401efca50e1ff99f694c6beaa57504b71b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 17:38:21 2008 +0300

    Plug a memleak in lzclose()

commit aaa7fa0f943817d344fab0a95340744f6acebf9d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 17:09:07 2008 +0300

    Plug memleak in parseSimplePart()

commit 485d19548834a68db3c57082b6bdc57fa58cf5f8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 17:04:08 2008 +0300

    Plug memleak in macro doFoo()

commit 586268cc94edf8dd8af5668f3c9000d02facf57b
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Apr 29 16:08:19 2008 +0200

    Don't use alloca in rpmGenPath()

commit f05198ca5ec06fe48805b55ec51ef600612804cf
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Apr 29 15:40:20 2008 +0200

    Calm down compiler warnings due to fsm->patch being const

commit 67f9ca7e1404fc85ad367d5eb9b90f8c5501ebad
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Apr 29 15:01:05 2008 +0200

    Avoid general ETXBSY and allow it for HP-UX exclusively
    - for more info: http://www.uwsg.iu.edu/hypermail/linux/kernel/9808.3/0883.html
    - remove the rest of alloca()s

commit 498677dad2abb860ca6acd5aaa7aa0ef6d288267
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Apr 29 14:59:18 2008 +0200

    Remove useless alloca_strdup()
    - and yet more alloca usage in writeFile()

commit 4bb802d6aedef83a0d356d522e107697fed1c369
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 15:50:43 2008 +0300

    Remove unused alloca_strdup() define

commit 010706ecbd4bdabc954fd9f4499a99772b92fea7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 15:08:47 2008 +0300

    Allocate directory names on heap in compressFilelist()

commit ba763344d4116912e616337f7a8c305a29c48c36
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 14:44:28 2008 +0300

    Use rpmGetPath() to construct full dbpath in db3open()
    - another alloca() bites the dust...

commit bd9e4ed8d7dec6ab7a247298fa8ce5be67e98f58
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 14:17:09 2008 +0300

    Oops, one more alloca() in rpmtsOrder()

commit 4fc15424d815bc9dc22644d16ea00db06136b8b9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 14:13:59 2008 +0300

    Remove alloca() from rpmtsOrder()

commit 08314921ea4ba4c333838da8682f01c846ff744b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 14:02:06 2008 +0300

    Dead code elimination

commit 2cef5b46458f2d19825cdc40b21bb591b69f87f0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 13:49:53 2008 +0300

    Remove alloca() from rpmvercmp()
    - use variable array instead, versions had better be short enough to fit
      on stack safely

commit b5568386c2a16a56a5b290a4e6f498229d7e695f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 13:33:09 2008 +0300

    Avoid alloca() in headerLoad()
    - REGION_TAG_COUNT is a constant so local array would do even without
      C99..

commit 3be356cb0f583c68069150421a7bdd5a8f65d555
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 13:17:33 2008 +0300

    Avoid alloca() in findTag()
    - instead of putting RPMTAG_ into strings where it's not present,
      do it the other way around
    - avoids a string copy and redundant string comparisons on the RPMTAG_ part

commit 69e3540ed0c875221147d9af5ab90089d55e5fa5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 12:59:20 2008 +0300

    Avoid alloca() in shescapeFormat()

commit 226c4d16279702593ff814b1573e79e33e17fd72
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 12:45:51 2008 +0300

    Avoid alloca() in headerAddI18NString()

commit cde5bfe5180906783ca26257e05349d82f8d9272
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 12:40:03 2008 +0300

    This code has been commented out since -99, time to let go...

commit b2e875be7822947a5a9db063a234e67fd5f2261c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 11:54:44 2008 +0300

    Eliminate remaining alloca() in rpmAddSignature()

commit 5e775226b6d11fc81022971c06d14ffda3b2e6cb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 11:53:13 2008 +0300

    Plug memleaks in signature generation

commit 059719a8fe75ade2d4c12f880cc5d9fb3a1cd140
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 11:35:30 2008 +0300

    Avoid alloca() in rpmtsSetRootDir()

commit 4114a51af8e2bbdbe0a243639be1bd496c469885
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 11:26:00 2008 +0300

    Avoid alloca() in rpmDetectPGPVersion()
    - the whole detection is hysterical and could/should probably be killed
      but what the heck...

commit e80494b132dfd3c48bff1e03ad5b2f0f8900274c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 11:24:25 2008 +0300

    Avoid alloca() in makeGPGSignature()

commit f438a06cb286fe36ce7cd5887fe49271134b10cf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 11:16:46 2008 +0300

    Avoid unchecked stpcpy() to static sized buf from alloca()

commit cdcda7408ab9f56bfc0d39cf553f92a92390a83a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 11:12:15 2008 +0300

    Avoid alloca() in rpmtsInitIterator()

commit dee823bf5022bde263da195458f7670581aefef5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 11:02:41 2008 +0300

    Read "canonvendor" macros by default again
    - the whole vendor etc stuff needs rethinking etc, but for now this "works"

commit ad9d933514a6ab5015bccf7a7022f0d2319deab7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 10:40:56 2008 +0300

    Avoid alloca() in legacyRetrofit()

commit 2c12725620193da7078693cb893c3dab152a0dfa
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 10:34:54 2008 +0300

    Use rpmioSlurp() for reading in rpmrc's
    - have doReadRC() do the opening of files instead of caller
    - always expand rc filename on entry, simplifies includes too
    - streamline exit paths

commit 10b2d1300b25e47a525666fb2fc4564fa4cfe13e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 29 09:59:56 2008 +0300

    Use rpmGlob() for expanding rcfiles globs
    - avoids static filename buffer and simplifies things quite a bit

commit eff99d08ac16fd37d53bb1baf808aadaa1e99743
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Apr 28 17:30:25 2008 +0200

    Rewrite rpmioMkpath()

commit b1c39e29cf1a1ad464a349aeb93b2e5d6ca2b89f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 28 22:25:47 2008 +0300

    genCpioListAndHeader() indentation fixes, no other changes

commit 4201d351b44deea108bffa74ca61d4aa4c0a611e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 28 21:58:26 2008 +0300

    Allocate buffer for %doc filenames dynamically

commit a0797f0237e42556400f7107108f1687bcf097e2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 28 21:37:47 2008 +0300

    Remove unnecessary block which just messes up indentation

commit e3a2ab3269706336b4a5692ccd4563a4bc32a730
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 28 21:35:26 2008 +0300

    Set processingFailed just once

commit bae8f0ba0558436ea3bcc7a4f646204333e619ca
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 28 21:18:00 2008 +0300

    Don't leak memory on failure in checkForValidArchitectures()

commit 04d0b1857ce2f15490938bb8608ecc0d6b6472f5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 28 20:05:07 2008 +0300

    Remove unused alloca_strdup() macro

commit b22712816cff6ee572056cb55c5d256a33ae1778
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 28 19:10:25 2008 +0300

    Remove alloca() from strntoul() helper in cpio
    - only small (< 10 chars) buffers from struct cpioCrcPhysicalHeader are
      ever passed in, tmp buffer on stack is safe

commit bbc89fbd350a27e296171f6eb973143662c88214
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Apr 28 15:59:46 2008 +0200

    Get rid of string hackery in rpmdb-py.c

commit e22ee464765a40549e35fa754f1c4b018b5901b2
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Apr 28 15:42:04 2008 +0200

    Remove alloca() from argv.c

commit 476d19ca290d04c05a454873d67f45a8f498e98b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 28 16:22:33 2008 +0300

    Oops...

commit e1a99d0e12003705376d065d1e471b5221d231de
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 28 16:03:44 2008 +0300

    ..and for parseForLang()
    - return rpmRC like the others
    - avoid alloca()
    - streamline exit paths

commit c310e2b6a6cdd57b2ce83b197f2d93a5562e147b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 28 16:00:16 2008 +0300

    ..and the same for parseForConfig()
    - streamline exists, avoid alloca()

commit 1b61e468ad75e86bf75d34e78d7f59ca1b5e26a6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 28 15:57:32 2008 +0300

    parseForAttr() tweaks
    - streamline exit paths a bit
    - replace alloca() with xmalloc()
    - avoid messing q value while checking for white space
    - return rpmRC like the others

commit ea60f8ccb7de9302b84f2e8578ece00570756553
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 28 15:50:15 2008 +0300

    Avoid alloca() in parseForVerify(), streamline exit paths a bit

commit 27ad1cfe07206b4bd7fed3631571ed7ef6d8fdd2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 28 15:41:55 2008 +0300

    Avoid alloca() in parseForDev()

commit 18bf8b8b8e246f2612ec16f56defbdbe21a45404
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Apr 28 14:08:35 2008 +0200

    Oops, simplified it a little bit much, fixing

commit 98259f8edbe3997f4153cecb4ce681bb64ef71d7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 28 14:47:12 2008 +0300

    Avoid alloca() in rpmGetFilesystemUsage()

commit 61f144522e830ac714679f2ed860d91ea387c3ee
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 28 14:42:10 2008 +0300

    Avoid alloca() in getFilesystemList()

commit 75a21293916e7b93b5059045a254ab8bb3ecc96f
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Apr 28 13:41:45 2008 +0200

    Remove remaining alloca() from rpmdb.c

commit c729f98edf54306501096b7f45c55ca1f96373f0
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Apr 28 13:40:56 2008 +0200

    Simplify dirName handling in rpmdbFindByFile()

commit 811e7be604bc5f557ed5f3b7c711cc2322438761
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 28 13:56:31 2008 +0300

    Drop va_copy check, it's already done by C99 checks

commit 48f19e86b56be0ae220ff1114711684dd04480bd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 28 13:53:35 2008 +0300

    Require C99 capable compiler in configure
    - also enable C99 mode if not on by default

commit cf767c71c57491e3fb159762da9ae86ebb828ef0
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Apr 28 12:50:24 2008 +0200

    Remove alloca() ussage from rpmcache.c

commit 39dcd57e7fbcc56baed1ac40f518f510b3ca7c32
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Apr 28 12:50:01 2008 +0200

    Remove unused alloca'd variable

commit 324eb182d71af8195df6a2dc52f39f4e1b62db55
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Apr 28 12:38:32 2008 +0200

    Remove alloca() from build()

commit dbaa9393a20cbbba39f7fa6ee47d375f4ef8862f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 28 12:31:51 2008 +0300

    Eliminate pointless alloca() of hsa struct

commit 4b737f54118a0cae4cea75b667c568a8d24c2c33
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 28 12:25:25 2008 +0300

    Remove bogus/misleading return value of hsaInit() and hsaFini()
    - they don't (de)allocate anything, just modify contents of hsa arg

commit 2e833ba2c1eeb2113264c056435ae906a8c05a3b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 28 12:22:33 2008 +0300

    Remove useless hsa != NULL checks
    - hsa has been already dereferenced before it's even checked for NULL in
      both hsaInit() and hsaNext()
    - they're just internal interfaces with limited number of callers, we know
      hsa can't be NULL

commit 4727c2a0aa9b489d187ce9ad3e035bf460071b59
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 28 10:59:42 2008 +0300

    Remove unused variable

commit 80cb43fbec02340ec6a56e5a6a2c2a273aea9898
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Apr 26 14:41:58 2008 +0300

    Eliminate pointless alloca() of entryInfo and indexEntry structs

commit 4d0345558534bbeda0b6e3ed960192e0b9bd11e1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Apr 26 15:05:50 2008 +0300

    Eliminate pointless alloca() of local DBT structs

commit 64b573d16682e61b134e287d74a18b22a669419b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Apr 26 15:02:33 2008 +0300

    Eliminate pointless alloca() of struct fileIndexEntry struct

commit 46409035b658d22baa01456a227ee9839d06a1da
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Apr 26 14:59:09 2008 +0300

    Eliminate pointless alloca() of availableIndexEntry struct.

commit 88c06ccbf7ea2544ea38675646626532e6547f17
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Apr 26 14:51:20 2008 +0300

    Eliminate pointless alloca()'s of local dirInfo struct

commit 2a51abedaf8cc83fee37dccdf28888cc9baae1fb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Apr 26 14:27:30 2008 +0300

    Eliminate pointless alloca() of entryInfo and indexEntry structs

commit 7ee6ca9761b417e87f97cc8b4c20a3b4de1a5ef7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Apr 26 14:21:22 2008 +0300

    Eliminate pointless alloca() of local DBT structs.

commit e96d841a591356dd8fc76c75b87684a1c2e69d18
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 25 20:29:14 2008 +0300

    Revert bunch of alloca() -> xcalloc() changes in rpmdb routines
    - reverts the following commits:
      0b0dcd114028e1e2a00870917cf07a27858a30b1
      71018d6d2d67b0e17c737e7a90cc6bb704e224ba
      f6134083c29c6cbb3be255fadae47a03d35b09f1

commit f3ead2bb94d4d1451d35f52c4dede180f9895019
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 25 15:39:04 2008 +0300

    Replace alloca() + memset() with xcalloc() in runImmedTriggers()

commit 8e4a5f88cecb6c9f577cff91317851bc94abd8e9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 25 15:37:09 2008 +0300

    Avoid alloca() in markReplacedFiles()

commit 8660bd61f6bbfa20635ef74827b36864d9ec0e7b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 25 14:52:03 2008 +0300

    Reduce alloca() use in relocateFileList()
    - dirindexes and filelists can be *huge*
    - filenames and such still allocated on stack, fix later...

commit d49aaeb3b94a7e2e53137e09a517534b080ac444
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 25 14:34:36 2008 +0300

    Replace alloca() + memset() with xcalloc() in skipFiles()

commit ccb9647d4b5c6255ff1b734d565b4873c67e0445
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 25 14:07:31 2008 +0300

    Avoid alloca() on links in printFileInfo()
    - also avoid unnecessary nulling of perms, no point ..

commit b046a85409cecc5d7efdb75185fe155e0e3e4293
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 25 13:54:59 2008 +0300

    Avoid alloca() in rpmInstall()

commit 12ad8a5194168cf5213dc359283a8997c5501999
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 25 13:39:44 2008 +0300

    Simplify setPathDefault()
    - use rpmGetPath() to get sanitized path, avoid alloca()

commit a2addde0f041531a47dbb20ba8d8cc5a5de1488f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 25 13:28:51 2008 +0300

    Avoid alloca() in rpmteColorDS()

commit eabf5f3583f10bd0f52c15ecedefec88efd2c670
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 25 11:55:14 2008 +0300

    Single point of exit for dateToTimet(), avoid alloca()

commit f9a95483bb09672c9efdeba82c015c12dd6c7291
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 25 11:45:35 2008 +0300

    Simplify lookupPackage(), avoid alloca()

commit c636733f3628ebf2e10cbf10dfbe7a52cc5e3fea
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 24 17:36:31 2008 +0300

    Remove outdated comment, add a todo reminder
    - callbacks for pkgs without files have been faked for some time now..

commit c55f6b0a939cb63a07cc9bf227b5b25372dee59e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 24 16:48:32 2008 +0300

    Might as well compare against tag shortname now that we have it

commit d97ba0bcdc77c304726b96e9faa86b7c319e8346
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 24 15:40:36 2008 +0300

    Update translations for a change

commit 14f39232e6c995b51e5d1b1e946306cbb3e63a5b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 24 15:38:45 2008 +0300

    Remove bogus const from rpmfi pre- and posttrans members
    - they're always duped, scaremem or not

commit 16a5cd02fd99cd2f6c82c550fae3afec769c217d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 24 15:23:35 2008 +0300

    Eliminate rpmtsi refcount debug junk from the API

commit 3a74895f03e9fb0c217813aaec8e3b4c1e04a0b4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 24 15:16:03 2008 +0300

    Update doxygen docs for rpmteNEVR() and friends

commit 365ffe20a4f7a3bc3a25f313075f70233f854b6b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 24 15:08:09 2008 +0300

    Remove meaningless extern on function declarations in headers

commit f4d12868e2a45b20fe9b3f763ce35a30623f3633
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 24 14:56:27 2008 +0300

    Add rpmteEVR() function
    - seems like a useful thing to have
    - use in ensureOlder()

commit 4d35eeca1a1f203b404e5878d6ae6a4bec35966f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 24 14:51:58 2008 +0300

    I'm out of words for this one.

commit 90ca5e5989ec289a51d2e1c7c8caa59063a6fb70
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 24 13:23:22 2008 +0300

    Check for chroot() result in rpmtsRun()
    - bail out early if it fails instead of blindly continuing and potentially
      messing in real root (chroot can fail for priviledged user too)

commit 0ec3352e8b1b1f622e1577c67d0a3303422d4b77
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 24 12:49:58 2008 +0300

    Make verification with --root reliable (rhbz#434150)
    - use chroot() instead of vain path manipulation games, otherwise we'll get
      bogus results from users and groups not matching etc
    - non-priviledged verification is unreliable anyway but running rpm under
      fakechroot (unpriviledged) gives much better results than just prepending
      root to paths

commit 39b75d24cfe50e65ca940cc5466beb03c801d033
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 24 12:03:45 2008 +0300

    Force preloading of name service libraries early in initialization
    - this used to be done in psm chroot operation, but we can end up chrooting
      before that and without psm being involved
    - don't bother with getpwnam() etc, a single gethostbyname() loads all
      that we need

commit ae929240b7643927bc8a317e7992272f69a00fe0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 22 16:08:43 2008 +0300

    Relax spec permissions in tar build mode (rhbz#253648)
    - mkstemp() can create the spec in overly protective mode (rpmlint whines),
      relax up to what users umask permits

commit 6e090dc5c5bdfad6f8a38d2a1efb47b361d98c7e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 22 15:23:38 2008 +0300

    Simplify rebuild path generation

commit 338c9be61eef414ea361ac849bc1f926429957f6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 22 13:55:41 2008 +0300

    More path manipulation craziness eliminated in rpmdbRemoveDatabase()

commit 542cfb834305208beeb26e817de62ce26de9103a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 22 13:44:10 2008 +0300

    Eliminate wtf dbpath shuffling in rpmdbMoveDatabase()
    - Just use rpmGetPath() to construct a clean, sanitized path from the
      components.
    - Previously it went to great lengths to ensure trailing / on dbpath,
      catenate the paths with %s/%s/%s and then rpmCleanPath() to get rid
      of the extra slashes just created. Eh.

commit dfa18246308ca5c84fd642a8a3f4a6a0cf252cce
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 22 13:32:52 2008 +0300

    Don't use alloca for struct stat, there's no need to.

commit 045cef3286f164e47bcb38e6518861b07222ee9d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 22 13:30:52 2008 +0300

    Add rpmdb internal helper for cleaning up db4 environment, use it

commit edc5c0ef84a93d19acecaaf9a07f4ea3e9d8fe39
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Apr 22 12:41:08 2008 +0200

    Recognize lzma payload from its magic
    - old lzma formats without magic aren't supported

commit fa5c3f2e1025d1d1d34f3f09879a6dec75d2325e
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Apr 22 12:39:55 2008 +0200

    Prefer lzma's default -7 compression

commit 94958a3bfb89393179b7af19ea62d4b1d25d32d2
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Apr 22 12:38:51 2008 +0200

    Main lzma rpmio part based on SuSE patches, thanks for them!

commit 3626f25462e1e63c226134dec254a19799dfced7
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Apr 22 12:37:18 2008 +0200

    Handle lzma payload

commit 3d8e0ef13a389c263e3867c20ce32ae088332c7b
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Apr 22 12:36:27 2008 +0200

    Check for lzma library in configure.ac and reference it in Makefile.am

commit 341266ae26e6685d1d870b28533469c653355bb7
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Apr 22 12:32:59 2008 +0200

    Allow lzma payload in writeRPM()

commit 91123cd04354afadcf62ecf2b6aa069c8bee86d2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 21 12:39:18 2008 +0300

    Rip i386-specific RDTSC support from rpmsw, use gettimeofday() everywhere
    - we don't need accuracy beyond what gettimeofday() offers for bleeping
      debugging benchmarks
    - we especially don't need hw-specific magic asm voodoo to get unreliable
      timing results (RDTSC isn't reliable with multi-core/hyperthreaded CPU's
      etc etc)

commit 4fc06030b80f3b7c2f6d5dd80d65091c19b70b22
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 21 11:13:13 2008 +0300

    Avoid %patch0 getting applied twice

commit b80340761ff064469fa099a541e2f2128ca482c6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 21 10:59:31 2008 +0300

    Retrofit RPMTAG_SOURCEPACKAGE to source rpms.

commit d2406a7e1754e7a7dc8136d9132929df36e5f230
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 21 10:33:02 2008 +0300

    Use argvAppend() instead of manual reallocs, plug a leak from rpmGlob()

commit fe42db7ab8647b534cd9d080e72944ba7e1fb938
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 21 10:18:14 2008 +0300

    Use argvNew() instead of manual xcalloc()

commit 29faf5715517090a36e6597c4a4c3a12f0d2df9f
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Apr 21 08:19:23 2008 +0200

    Allow setting fuzz factor  to 0
    - use --fuzz=num instead of -Fnum or -F num to be sure
      /usr/bin/patch won't complain

commit 3dad7b6f0f463d23518f01543da9b958c6fc12f0
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Sun Apr 20 12:17:37 2008 +0200

    Don't use static buffers

commit cee73cdc0b165a4044795accde68bf37a7684125
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Fri Apr 18 20:46:56 2008 +0200

    Fail if it is impossible to chdir to directory to run helper script in

commit 12d3a07710543a72cdc1d46a287051dda53f38a3
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Fri Apr 18 19:45:10 2008 +0200

    Fix find-requires (rhbz#443015), thanks to Bill Nottingham

commit 3d7db31479c1cf72fdec31bffe207b2a82068210
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Apr 20 22:07:53 2008 +0300

    Sanitize + simplify %patch "macro" parsing
    - Use popt for parsing %patch options, give accurate error messages for
      invalid options instead of bogosity like "bad %patch arg: %build"
    - Store all the variants we can get patch numbers into argv to avoid
      special cases when passing to doPatch(), this has the added bonus
      of removing the 1024 patches per line-limit (as if somebody cared)

commit d35cec6ab24b1d52bfaf953e3174b53f3d1170de
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Apr 20 16:38:29 2008 +0300

    Single exit point for doPatchMacro()
    - assuming failure avoids several redundant free() + return RPMRC_FAILS

commit 36cfff298b22ea852f379766dd7cf92ede486e3b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Apr 19 18:56:38 2008 +0300

    Remove ugly hardcoded docDirs list, use a macro instead
    - %__docdir_path contains a colon-delimited list of documentation dirs,
      macros expanded before use
    - dynamically allocate docDirs as needed using an argv array, removes
      the hardcoded MAXDOCDIRS limit on docdirs (not that it matters much)

commit 099d00de3b59052800b4e78665ccafc896d7cbc4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Apr 19 17:09:43 2008 +0300

    Avoid entirely silly rpmio FD_t use in processPackageFiles()
    - since we just convert it to FILE anyway, no much point in using
      Fopen() then
    - additionally remember to close the file if expandMacros() fails

commit 1b972af8a6698b62a0c0d0818c6798aaad386b70
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Apr 19 16:58:55 2008 +0300

    Change rpmGlob() argvPtr type to ARGV_t to make it obvious what it is
    - also fix incorrect doxygen commentary, it wasn't a contiguous blob
      before either

commit 2072bd1334846e4d005c2e07949439f391ea6173
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Apr 19 16:53:29 2008 +0300

    Use argvFree() instead of manual free in loop

commit 9ae911eefedf9618202b6a76d670914119784464
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Apr 19 16:42:56 2008 +0300

    Further simplify rpmInitMacros()
    - Use argvSplit() for splitting the macro path to components instead of
      manual pointer-parsery.
    - If URL's are to be supported or accepted at all (previous code attempted
      to skip them), ':' is a very poor delimiter character.

commit e336146a9ed6b28016e0dff68ea6979c3c9e5866
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Apr 19 16:04:11 2008 +0300

    Use argvFree() to free, well, an argv, instead of manual freeing

commit 5ad7fbda008cba771e09fe305a9b9db3dc5c6f06
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Apr 19 15:58:47 2008 +0300

    Simplify rpmGlob() a bit
    - use argv*() routines instead of manual argv manipulation

commit cebd0812bc5b1a6d13a6e0937081f728750d016c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Apr 19 13:42:04 2008 +0300

    Dynamically allocate buffers in rpmGetPath()
    - avoid unbounded copying to static sized buffer
    - use rpmExpand() instead of expandMacros() which requires preallocated
      buffer when we've no idea of expanded size

commit 601b4d0e6aff2630fa17269d063106d8d921d4b0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 18 16:14:15 2008 +0300

    Avoid entirely silly rpmio FD_t use
    - rdcl() converts to FILE internally anyway, might as well use fopen()
      to begin with...

commit 38a975b157b504f1234b37c17bc072dd1eb291f6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 18 15:39:07 2008 +0300

    Remove unnecessary xstrdup() of dynamically allocated buffer
    - leftover from when target buffer was static...

commit 94e64fec18f874383f9cbc8eda568664899d658f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 18 15:15:05 2008 +0300

    Cosmetics: group operations more logically...

commit 2d8a27ed1677da5c6a633a4fb757fc43bef4aeb9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 18 14:54:58 2008 +0300

    Simplify macro option + argument processing
    - eliminate ninja-voodoo pointer-hackery parsing, replace argvSplit() +
      argvJoin() and friends
    - eliminate static sized buffers from the processing

commit 2c0f7b54ad8e820ec357c2cad08b8d31fa601ead
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 18 12:36:02 2008 +0300

    Add argvJoin() for combining an argv array into a string

commit 8058de50b24f548df74df0943cb3a29c77fbde61
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Fri Apr 18 08:53:43 2008 +0200

    Put '\0' to the end of xmalloc'd string to avoid surprises

commit 94ff22b129aeb31c38848231e40f87aa4a5613a1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 17 21:56:54 2008 +0300

    Eliminate splitString() and freeSplitString()
    - use ARGV_t instead, it does the same and more...

commit 7800911bb0281ab568af767c74a664b9c1215464
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 17 19:21:00 2008 +0300

    Accidental typo introduced to rpmrc.c
    - seems to be one of those days, sigh

commit 9cf9ad7bc919451f65881c2dd8bc772ba722d7fa
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 17 19:19:07 2008 +0300

    Fix dumb off-by-one errors introduced by rstrlcpy() changes

commit d4247aadff9dba401dd00f0e4e06969323b9b6f0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 17 18:41:53 2008 +0300

    Bunch of strncpy/strcpy -> rstrlcpy() converts in rpmrc

commit 834d2d2691fa3b645902bbcd1db9310c14778a69
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 17 18:15:34 2008 +0300

    Add missing include

commit 85c6b8e5f9007e1cdf77e043079c774792d8aa52
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 17 18:13:52 2008 +0300

    Use rstrlcpy() for various parseForFoo() functions in build
    - just avoids manual termination, no new fixes here..

commit 567c0abc7eb558318077feed34a1edde9fafcaf0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 17 18:04:28 2008 +0300

    Use rstrlcpy() for parseRCPOT() version + release
    - just avoids manual zero-termination

commit ba85f103e59db924f4dfba7de56bcba8ae311e3c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 17 17:39:08 2008 +0300

    Use rstrlcpy() for printFileInfo() username + group fields
    - just avoids having to manually ensure zero-termination

commit 805e294fd638bd3585400e317be81207e4a04388
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 17 17:35:57 2008 +0300

    Make sure lead->name is zero-terminated

commit 119b912773e3c41a95f31656ccbd59bbc307c915
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 17 17:33:57 2008 +0300

    Add rstrlcpy() string helper function
    - adopted from strlcpy() function of OpenBSD originally developed by
      Todd C. Miller <Todd.Miller@courtesan.com>

commit 02939c15fa5441aff46c5f2cbf28086fe67d4d24
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 17 17:33:30 2008 +0300

    More accurate filtering of .rpmnew etc when reading macros
    - strstr() can give false positives, use rpmFileHasSuffix() instead

commit 64e0af29bd38a1f8ca6573ec9734ae475ed8909b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 17 11:09:03 2008 +0300

    Have gcc check for NULL terminated arglist in rstrscat

commit 9a6be3e53feb1a32e446419c6fef910eeca74f01
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu Apr 17 09:20:20 2008 +0200

    Use sizeof() instead of direct element count in rpmluaGetVar()

commit bf3a8e6cd9253d2321a5d8c3665fa591eca3aae8
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu Apr 17 09:01:28 2008 +0200

    Don't copy unknown number of args to buffer of fixed size in initGlobs()

commit ea8029e4e95e4106ccb6f9b6407c3856a281f312
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu Apr 17 08:57:43 2008 +0200

    Introduce rstrscat()
    - concatenates NULL terminated strings to newly (re)allocated memory
    - purpose is to prevent overflows caused by writing unknown-sized stuff
    to static arrays

commit f829c42f075bfac82b08859a476579446d594518
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 16 15:54:40 2008 +0300

    Fixup leftovers from lua/local -> luaext/ move

commit 2eb20ad3596131420a1f90695f4bedf73067816b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 16 15:52:06 2008 +0300

    Document external Lua requirement

commit 2ecd8fc22d897cffe433820205036b493fd012f2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 16 15:35:09 2008 +0300

    Convert to Lua 5.1 library loading style
    - luaL_openlibs() to pull in all standard libs
    - local extensions need to be registered by calling through Lua

commit 1c979525dc1341c7fb812bf84d167467094de1ec
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 16 15:25:26 2008 +0300

    Add modemuncher + lua init stuff to extra dist
    - modemuncher.c gets included from lposix.c, urks

commit d48e6c8380ebd1f8d4d4f2c80237dc1488da4721
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 16 15:13:25 2008 +0300

    Start phasing out internal copy of Lua
    - don't build internal copy of Lua
    - move 3rd party extensions (posix and rexlib) to toplevel luaext/
      directory, built by default (unless --without-lua specified)
    - auto*foo checks for external Lua
    - minimal tweaks to lposix.c and rpmlua.c to get them build with Lua 5.1

commit 23f6917b10b6579f39199e4cf831bb13fc596824
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Apr 16 13:40:05 2008 +0200

    Don't use mktemp(), we are sure mkstemp() is available now

commit 58fdb16e75b2200ba509f726f938493eb43d54ea
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Apr 16 13:01:06 2008 +0200

    Require mkstemp()
    - mktemp() and tempnam() are dangerous

commit 0662b2f5127df7c07e379c117085236f90f49b42
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Apr 16 11:16:31 2008 +0200

    Kill queryHeader(), it is not used and buggy

commit 004c0e50b42ba10141f20eb3bf2756fd70282f2c
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Apr 16 10:24:02 2008 +0200

    Kill useless initialization to "(unkown error)"

commit 9e7d602e85440dd79054dfc440aa14bba841a830
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 16 11:20:03 2008 +0300

    Plug a memleak from pgpHexStr() in showQueryPackage()
    - fmd5 is only needed in a single place, move alloc+free there to avoid
      having to deal with the umphteen continue cases

commit 36cbcd454eed7a4158f49cfa55e2f817d5e22bd9
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Apr 16 10:11:37 2008 +0200

    Don't try to print buf if it's empty

commit ce12b9a112ea067d8fb2e3960d95a4468555e33b
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Apr 16 09:37:55 2008 +0200

    Remove insecure hacks in query.c
    - remove useless flushBuffer() function
    - update printFileInfo() to not to write to unknown sized buffer,
      but to print stuff which is its sole purpose

commit e50ff40c22e5cc4e7cdfecbb9808ad5ce3d6b77c
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Apr 16 08:28:11 2008 +0200

    Remove useless statements from showQueryPackage()

commit 9a2b4cba0501235345ea42e068bff86bd17d15d3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 16 10:17:19 2008 +0300

    Make headerNVR() just a special case of headerNEVRA()

commit 72b200fe591bf13ecd238c1b3bae785a2907a0fc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 16 10:10:25 2008 +0300

    Oops, only free if allocated ;)

commit 9ebe81639ab6f5449ad88818fdd111c68245a101
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 16 10:06:56 2008 +0300

    Plug a memory leak in armorFormat()

commit 3ffbbeba1977caec65a826e408a433dddb7d2fbd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 15 18:52:42 2008 +0300

    Doxygen parameter name mismatch

commit ecaa34a9dc564848f4f1bf3cd413b27da2b9c9ff
Author: Jindrich Novy <jnovy@dhcp-lab-186.brq.redhat.com>
Date:   Tue Apr 15 17:35:13 2008 +0200

    Get rid of string hackery and overflows while rpm -V

commit 0097e6fe952a6f39b99e4761ab613f99d34c142c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 15 17:26:15 2008 +0300

    Plug a memleak in pgpArmorWrap()

commit d43ee53dff96926d3ac61c9a3cdc9251326e2790
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Apr 15 16:30:58 2008 +0200

    Introduce rstrcat()

commit f68f58975b27b724f363c86e11699af9fc05bee7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 15 13:27:43 2008 +0300

    Make find-lang.sh --with-man brp-compress friendly (rhbz#440994)
    
    Patch from Ville Skyttä

commit 21b6c0af17bf95cd00b63b2a935eb3bdf848ff7f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 15 13:13:40 2008 +0300

    Don't exempt *.jar.so.debug from buildroot checks (rhbz#442264)
    - jar-kludge no longer necessary...
    
    Patch from Ville Skyttä

commit 627384f9e1e5342d09e1b06b9ac1822f76972d80
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Apr 15 09:47:38 2008 +0200

    Remove rhnplatform, cvsid/svnid, serial and copyright tag from rpmbuild

commit 55ebf2c079aa992f891ecc87d3f7aba0a844650d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 15 09:16:05 2008 +0300

    Abort build on unknown payload compressor
    - explicitly check for supported rpmio names, bail if not supported

commit a2bb439b3ca2789f21b1e6efcbe0ffbaad8aca48
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 14 19:32:51 2008 +0300

    Avoid unnecessary rpmio_flags alloc+free in psm

commit 180a34737203f0cc1e0d843915387ce9ffeb894c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 14 18:15:39 2008 +0300

    Fix indentation, no functional changes

commit 9309a4bd097971fe68b6bfa50d73614baa4844fe
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 14 18:13:06 2008 +0300

    Remove glibc upgrade-helper kludge, only necessary for some ancient RHEL

commit a4bbf08f4d365f2d5ff746917cb8cfa66c1d3c51
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 14 18:12:18 2008 +0300

    Lose useless variable

commit 7401ab6a7b0de775e16b371a1c98989cd3eb2cd5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 14 17:45:09 2008 +0300

    Split scriptlet execv() preparations to separate function
    - runScript() is way too big and still is, but this helps a bit...

commit d6de2803fc0a2be2da06e1f6f0c2b197510a8a1a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 14 16:22:28 2008 +0300

    Use argvAddNum() for passing package count numbers for scriptlets

commit 900cbb7d79a18fdf48964c395e038eabc38db43e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 14 16:21:40 2008 +0300

    Add argvAddNum() for adding integers to ARGV_t arrays

commit d858e4ffa5937aed5bbccf0d4667e1eebd55c218
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 14 16:01:06 2008 +0300

    Use ARGV_t for passing around scriptlet args psm internally
    - simplifies allocation and freeing while getting rid of alloca()'s
    - runScript() needs to add arguments anyway, might as well pass in something
      that *can* be modified without having play games with partial const,
      partial alloca() contents in the argv
    - removes the need for passing argc around
    - (unrelated) move fi->h NULL check in runScript() to beginning *before*
      we trie to access it, duh...

commit 08375fa2bce124043954feb8c9cc8023004586bf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 14 15:58:10 2008 +0300

    Add argvNew() to create an empty argv array
    - yes it's silly ;)

commit 4e34fcc2bda4b6d18287981c1a053e9e4380caba
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 14 14:35:34 2008 +0300

    Avoid redundant script name formatting
    - remove two points of return to simplify allocations..

commit 8444c6f997dd0eb2a26b6870768b5bd049247db9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 14 14:33:23 2008 +0300

    Reduce #ifdef clutter

commit 647ff16b50dab2172d25b62dedb7db15a38811ea
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 14 14:31:51 2008 +0300

    Streamline lua scriptlet name formatting - just do it once

commit 1eb29bc44d01192dc050fa9bc4a4b0a6a25726d1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 14 14:29:07 2008 +0300

    Move legacy prefix compat thing out of the loop
    - additionally fix indentation, move block local variable def there...

commit 25f15fca9e1c25bbceee362be7016108a26eeaeb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 14 14:27:29 2008 +0300

    Don't bother calculating max prefix lengths
    - no longer needed, prefixBuf is dynamically allocated as needed

commit 2c2d98a08dc3a9e21014b3173102f73064fc39ae
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Apr 14 13:07:58 2008 +0200

    Prevent buffer overflow in readKeys()

commit 641046e4da826b5845703cdee23227dc344b0c7a
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Fri Apr 11 15:52:56 2008 +0200

    Use dynamic allocation in rpmlibNeedsFeature()

commit 3928bbc25bfe554767590abe009bc40b4776d49d
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Fri Apr 11 15:44:04 2008 +0200

    Don't warn about BuildRoot being "/" while it is actually empty

commit 674194a9e7f7e0430f73f438f42472b6983afe83
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 11 11:45:35 2008 +0300

    Simplify url retrieval in rpmInstall()
    - use rpmMkTemp() instead of manual kludgery to static buffer

commit 769bc86d6d9372c1b3e804309227a0e8554d72d1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 11 11:18:25 2008 +0300

    Simplify url io a lot
    - move urlhelper call to urlGetFile()
    - have urlOpen() call urlGetFile() instead of going the other way around
      and causing local copies and all sorts of craziness in the process

commit 78f38f9be08270170bf58e066db263bdde54f5da
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 11 09:38:44 2008 +0300

    Remove #ifdef NOTYET junk

commit fde961e25b09806e501dd214059ced1b07c1b49c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 11 09:05:05 2008 +0300

    Rewrite rpmMkTempFile() for sanity
    - Actually use mkstemp() for creating the temp file and return a FD_t
      dupped from the file descriptor returned by mkstemp().
    - Simplify the interface while at it.
    - Change callers for the new interface.
    - Yes we now require mkstemp() to work, mkstemp() is in POSIX.1-2001 and
      this is year 2008...

commit 48ff62a5291458ed1181cd6c31dcadb193ad2f8e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 10 19:36:03 2008 +0300

    First crack at adding ISA provides to packages (rhbz#235755)
    
    - Horrible kludgery to get the isa names and bits into platform specific
      macros from installplatform script. That beast needs to die. I mean really
    
    - In build, add provides: name(isa) = evr automatically when it makes
      sense (similarly to name = evr provides). ISA consists of ISA name and
      bitness (or wordsize). This can be used to correctly
      express multilib dependencies without resorting to (expensive!) file
      dependency kludges, eg for dlopen()'ed libraries where automatic
      dep extraction doesn't force dependency on 32bit vs 64bit version, you
      can now use:
          Requires: foo-plugin%{?_isa}
      This expands to foo-plugin(x86-32) for i?86 packages, foo-plugin(x86-64)
      to x86_64 etc, and permits spec to be shared with older distros which
      don't have ISA provides.
    
    - The same could be expressed with "canon arch" just as well, but
      using the ISA to differentiate from %_arch and the like:
      eg i386 could be used instead of x86-32 but it's overloaded with meanings
      (the actual i386 processor vs i386 compatible cpu family etc)

commit af4eea99e311171e3dd771c9b2ff59faa964d208
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 10 16:07:47 2008 +0300

    Avoid exporting providePackageNVR() in the library

commit d39cc1dfda9a18a3f3254e8d9ce240ace48c6856
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 10 16:04:12 2008 +0300

    Just call legacyRetrofit() instead of combined compress + provide retrofits

commit c01ef8228d373f9232b47ac0915bb5786e5266cf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 10 15:52:36 2008 +0300

    Add a static, far simpler addPackageProvides() to build
    - use instead of providePackageNVR() which needs to deal with all sorts
      of hysteric "rpm 3.0.2 did this but 3.0.5 didn't" issues

commit 61dff28fa4d9169fd71ccd7a9f83c1186037c3db
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu Apr 10 14:42:50 2008 +0200

    Call rpmReadConfigFiles() in rpm2cpio so that rpm2cpio won't complain like:
    warning: Unknown system: (null)
    warning: Please contact rpm-maint@lists.rpm.org

commit f350b07096e872eb135df68f2906eae2aa306726
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 10 11:25:59 2008 +0300

    Remove unnecessary providePackageNVR() from writeRPM()
    - packageBinaries() already calls providePackageNVR(), no need to do it twice

commit de8250d2dfe74cdd0637dda3000b0ff6d7e730be
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 10 11:19:40 2008 +0300

    Unbreak srpm generation broken by 6c4b0cbd93709a0fa18af1ef07af44da585bb5f6
    - cpio path in source packages is the basename of the file, make this
      obvious by using basename() instead of pointer games

commit 61b008b5662652ac8ea05ea76611b73efe9cf656
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Apr 9 13:05:45 2008 +0200

    Don't use stack allocations in rpm2cpio when not needed

commit 0b0dcd114028e1e2a00870917cf07a27858a30b1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 9 13:54:39 2008 +0300

    Replace alloca+memset with xcalloc() in rpmdbRemove()
    - return value is suspect: should return "ret" instead of 0 at exit
      depending on what happened in removal but preserving previous behavior
      for now...

commit 71018d6d2d67b0e17c737e7a90cc6bb704e224ba
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 9 13:48:00 2008 +0300

    Replace alloca+memset with xcalloc() in rpmdbAdd()

commit f6134083c29c6cbb3be255fadae47a03d35b09f1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 9 13:47:29 2008 +0300

    Replace alloca+memset with xcalloc() in rpmdbCountPackages()

commit 2dad3ee0a9ff1ad4a247d863e82c12cd66df3d89
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 9 13:33:23 2008 +0300

    Tonne of indentation + other formatting corrections to rpmdb.c
    - no functional changes

commit b54d8bc2432d2fb85fe5c768df4f2cdcbe22b73e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 9 13:19:26 2008 +0300

    Remove dead code

commit 3a5a6ef0fa1577ca9147a846982cd9f48eb52a82
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 9 13:17:01 2008 +0300

    Report full nevra in rpmdb remove debug message

commit 2191e9fe72d18202a8a9e28c818b25f10ed635ab
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 9 12:47:03 2008 +0300

    Plug memleak from signature verification now returning malloced msg

commit 03c6b4e934600fbe488c12d9271b72f676a4803a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 9 12:36:38 2008 +0300

    Simplify addTE()
    - eliminate wtf parsing of NEVR string which doesn't actually handle epoch
    - use NEVR and NEVRA formatting from headerGetN*() instead of doubling the
      work here

commit aa04d0cef3854ddc3d2f6eb316c1b7780dc1b53a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 9 12:17:36 2008 +0300

    Make headerGetNEVRA() return "src" as the arch of source packages

commit e7ed105c6723d7820ffd9fa917de1fdac47590a0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 9 11:17:43 2008 +0300

    Initialize pointers, lowercase local variable

commit 994ccffd9dee5d25da4fe4be1fa4d4b9baeea990
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 9 11:16:24 2008 +0300

    Rewrite headerGetN*() for simplicity
    - eliminate stpcpy() games
    - reuse evr formatting from headerGetEVR() instead of doing it over and
      over again everywhere

commit 7a64fb564a7c79e47a3ad86d17b5c671a64e44c4
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Apr 9 10:10:17 2008 +0200

    Fix buffer overflow in verifyDSASignature()
    - caused by assumption that sizeof(size_t) is always 4 (credited to jbj)

commit 3d717d5c45ac17dcd3e0ae82f8f7d9846b1a5d97
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 9 10:37:34 2008 +0300

    Replace bunch of wtf evr formatters with headerGetNVR()

commit f380230a0316c02fd6864b6bba688e2e340b1d7d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 9 10:36:29 2008 +0300

    Add headerGetEVR() convenience function
    - return formatted evr string + optionally pointer to name

commit 99322592d79cda46ad4e8f62d9613e2e588dc523
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 9 10:32:44 2008 +0300

    Make headerNEVRA() actually return epoch too
    - ...but epoch is an integer type in headers, change proto to match that

commit 2cfe2d3a3b73889584a0b165b07e8088a312e5c0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 9 09:00:44 2008 +0300

    And one more url->path rename for local variable...

commit 6c4b0cbd93709a0fa18af1ef07af44da585bb5f6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 9 08:58:46 2008 +0300

    Rip the last remaining disk/file URL stuff in processSourceFiles()

commit 7c0b42146764c9ece60039de9cc1be36adb0f71c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 9 08:56:27 2008 +0300

    And more url ripping in build internals... (addFile)

commit e92d22e279fe01d84e339a45755eb9e8d989388d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 9 08:37:15 2008 +0300

    URL -> path renames continued... (build internal)

commit e89e7ba72fa9e710cef780ed5931ed2c2d0603a8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 9 08:25:46 2008 +0300

    Rename FileListRec_s disk + cpio path variables
    - on-disk is diskPath, rename previous fileURL to cpioPath to make it
      more obvious it's the path in the package cpio
    - neither is an URL

commit 29674e476a9477cdf8c7c38743b6441eacb5f011
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 4 20:37:48 2008 +0300

    Yet more url ripping - buildroot is never an url
    - this is a bit nasty as it changes public interfaces but rename to make
      it explicit they're just plain paths, not urls

commit ddf46496b38cd30df54950816f6379239259d2b9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 9 08:17:48 2008 +0300

    Remove remote host stuff from build doScript()

commit 0e74fc55c7202eb975a8de878c4731b587d7a994
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 4 19:53:40 2008 +0300

    More url-path ripping in build/
    - processBinaryFile() and processMetadataFile() don't need steenking urls

commit d121a41db20392f24f85126a99f7736f614a660b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 8 17:15:32 2008 +0300

    Whoops.. gpg-pubkey "release" on import got busted accidentally

commit b2cfc89de58d856c4c1c26a3a24b91610b38b243
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 8 15:36:46 2008 +0300

    Simplify rpmtsImportPubkey() a bit

commit 1721614d96499d0cf7bc1a31a133fa1c777aa841
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Apr 8 13:48:39 2008 +0200

    Rewrite pgpArmorWrap(), the former one was insecure/hackish

commit 090782c3c73f1b95fa6a033f73ce3cb4de38b539
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 8 14:11:47 2008 +0300

    Collapse unknown / unsupport RSA sigtypes into one fallthrough

commit 45f5e135aee276185eda68642406ec6917315006
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 8 14:06:07 2008 +0300

    verifyRSASignature() buffer & other fixes
    - avoid stpcpy() into static sized buffer
    - avoid potential sigp NULL-dereferences

commit 08b4a09eef6da8f11f992f2461e67fdc8f256fc9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 8 13:27:59 2008 +0300

    verifyDSASignature() buffer & other fixes
    - avoid stpcpy() into static sized buffer
    - avoid potential sigp NULL-dereference

commit 7fe19f0a2a1a8bc6e2940d042b0478282b4887ce
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 8 13:23:01 2008 +0300

    verifyDSASignature() buffer & other fixes
    - avoid stpcpy() into static sized buffer
    - avoid potential sigp NULL-dereference

commit 2bb4618fa107dc6b02a4eb64e42d0021eca2a9e9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 8 12:35:36 2008 +0300

    Eliminate four copy-paste versions of nibble(), stick to rpmstring.h
    - the one in signature.c wasn't even used...

commit 6f619206fd13c6f0d90f143ec829312da4f81c48
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 7 16:13:29 2008 +0300

    Eep, fix bunch of segfaults and other fun from previous commits
    - leftover stpcpy()'s to non-malloced addresses and other "minor"
      issues in verify*Signature() stpcpy() elimination ;)

commit 69c71c2f61a487f3350afb7190dad5df01280eda
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 7 15:07:08 2008 +0300

    Avoid stpcpy() to static sized buffer in verifySizeSignature()

commit 4441084171df6a70e27321e9b371cd8f662a67ad
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 7 15:01:49 2008 +0300

    Avoid stpcpy() to static sized buffer in verifyMD5Signature()

commit 5a0880a5de5f9c00447d200846cf887d32755df5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 7 14:55:36 2008 +0300

    Avoid stpcpy() to static sized buffer in verifySHA1Signature()

commit 5ef8b53108f41220602febbd2e10f3b76ab32b82
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 7 14:04:00 2008 +0300

    Eliminate several copy-paste hex converters, use pgpHexStr() instead

commit a41079c692aa75e5e995aaf1e5c355edad5552d9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 7 13:51:20 2008 +0300

    Uh, return allocated pointer, not the end of string...

commit 9bb88f4088de792c2844357ad447470e76a7281a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 7 13:37:56 2008 +0300

    Remove static print buffer from pgpMpiStr(), return malloc'd data instead
    - convert callers
    - only rpmpgp.c internal needs, move it there and make static?

commit 41a32765bde2cc346bba93da364e7bdeb985fcf9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 7 13:29:48 2008 +0300

    Merge pgpHexCvt() into pgpHexPrt(), removing it from public headers

commit 8909c054344834eb45f2b09c0abfcfa1be214ec4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 7 13:26:46 2008 +0300

    Minimally convert all pgpHexCvt() users to use pgpHexStr() instead

commit a3a1a45c5131fd1dd885cd73bb9f6a7ac1ddbfba
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 7 13:02:18 2008 +0300

    Remove static print buffer from pgpHexStr, return malloc'ed memory instead
    - inlined to get xmalloc() from system.h for consistent malloc fail behavior
    - convert callers for new behavior, apart from some debug fprintf()'s

commit a70823b99814cbef36ccb1a056f9ef1e97bd7ae2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 7 12:35:37 2008 +0300

    argvPrint() doesn't modify argv, make it const

commit 4e0ab74d8d7a1ee976b7010bd80135cf4a1c7d7a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 7 12:34:07 2008 +0300

    Add ARGI_const_t, use where appropriate

commit c29adcee80825d79e7b522ee651360d1b9542539
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 7 12:11:16 2008 +0300

    ARGV_t const pendantry
    - const on typedef'ed ARGV_t doesn't mean a thing, add a new,
      (exec* compatible) ARGV_const_t and use where appropriate consistently in
      argv.h, rpmcli.h etc
    - popt's argv const is the wrong way around for exec() family, add explicit
      ARGV_t casts on popGetArgs() uses where needed
    - compiler silence, aaahh...

commit d728ec28dcb51a139859e4fed48da894a8363a51
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 7 09:31:21 2008 +0300

    Remove ARGstr_t typedef, it serves no purpose
    - argv strings are just regular strings, the structure is what's special

commit a5533d9604770d16942e44d8c68cd28228bc3d2b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 7 09:29:06 2008 +0300

    Remove bogus const from ARGV_t

commit 4a466050480fb3a83712270e2f2a46ea912fc440
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Apr 7 08:25:17 2008 +0300

    Missing rpmstring.h include

commit b05c0ebdf8ccd647df14538f6d8980c797344daf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Apr 6 19:34:59 2008 +0300

    Eliminate the uuuuugly and unsafe static tagname buffer
    - generate the "human friendly" strings at build time, store into the tag
      table structure
    - also avoids copying of strings and lowercasing strings at runtime

commit a413ad7de22601e59e7a89f19e4986eaf70c8fbb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Apr 6 19:18:44 2008 +0300

    Move the tagtbl.c generator monster awk out of Makefile.am
    - stick it into a separate script, reformat so it's possible to actually
      read it...

commit 14dcc3034d3d252d20f80daa9b708f88cc45dbaa
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 4 17:00:38 2008 +0300

    Oops, no newline at end of filename please...

commit 52788e096d89c340efc359c3b9dc52e1ff05fb95
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 4 16:26:55 2008 +0300

    %{_builddir} is not an url, ever

commit 0787a058f11fade66fad8d78186ae2b533e20f45
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 4 16:23:56 2008 +0300

    We only ever deal with local, regular files in doUntar()

commit c9bc7c32e0f260312f74b448c859c3acb9aab0b7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 4 16:21:58 2008 +0300

    We only ever deal with local, regular files in doPatch()

commit ed3eb18dc26f87d8bffdd1a05d26dd962b7f6f2d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 4 16:08:37 2008 +0300

    Rip the url-craziness in rpmMkdirPath()

commit bde22653bc7ef058933af9e814d9644fcf28abb2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 4 16:03:59 2008 +0300

    Get rid of the horrid zcmd hackery in build
    - rpmExpand() with %uncompress macro etc simplifies things a lot

commit 3c1126b2a550e2d9f51f46e0948ab5925fe07089
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 4 15:52:52 2008 +0300

    Missing decompression arguments for bzip in uncompress macro

commit d3f73d3ba08da6fa2906764f9a7504385801816b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 4 15:19:28 2008 +0300

    getcwd(NULL,0) has undefined behavior, use rpmGetCwd() instead

commit febc6c0db9b39b7ec5e588f2ff9ebf7d48529c61
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 4 14:59:55 2008 +0300

    Stop pretending spec can be an URL or stdin, neither works
    - simplifies buildForTarget() somewhat
    - while at it, eliminate remaining unsafe string operations

commit 384a31b4a355719d88d0472c0fd451874be1a4e9
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Fri Apr 4 13:48:02 2008 +0200

    Escape '%' in format

commit 6a766c00bfd488bb2c703910efc386bf95e6af05
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 4 14:33:45 2008 +0300

    Wait for popen() to finish before passing to isSpecFile()

commit 61c0a71059d5e5a5782018505c70d88299ac3ede
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 4 13:52:59 2008 +0300

    Move spec extraction from tarball out of buildForTarget(), simplify
    - use basename() and dirname() instead of manual pointer manipulation games
    - static buffer only used for reading tar output, sizes checked

commit a375e78dc61ccf171188db58f2acfb6224d553dd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 4 12:12:29 2008 +0300

    Oops, add --wildcards back to spec extraction
    - shouldn't really have to hardcode such stuff but...

commit f548819c94d9049a7dd850546286f8583e5a47e0
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Fri Apr 4 13:17:53 2008 +0200

    Simplify formatting

commit f6b0cf178a74f193677bbf9b2b345fea6ab0fff9
Author: Jindrich Novy <jnovy@dhcp-lab-186.brq.redhat.com>
Date:   Fri Apr 4 13:01:50 2008 +0200

    Simplify package comparison and format creation.
    - also avoid static buffer usage

commit baf9e88fc341d6b7bac32fb8e22768791bd596af
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 4 12:06:06 2008 +0300

    Assume failure in buildForTarget() to remove redundancy

commit 50a020acca37f48a074f5e2a5e09f6dcb16ec387
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 4 11:54:21 2008 +0300

    Refactor spec extraction on tar build to address several issues
    - simplify the code
    - use %{__tar} instead of tar as the extraction command
    - avoid sprintf to "big enough" buffer
    - basic spec sanity check while extracting, fixes rhbz#281391

commit 8aa121da832c1f66b1900f9568955ce4e32e49f2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 4 11:53:07 2008 +0300

    Spec of size zero cannot be a valid spec file

commit 02de5d75af66f9d27edd8e12cc6004b15b22ce55
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 4 09:55:30 2008 +0300

    Remove copy-paste currentDirectory() from sqlite, use rpmGetCwd()

commit a5a3cfc3ed67083eaca0af0ffa48565a3e99bf57
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Apr 4 09:53:17 2008 +0300

    Rename currentDirectory() to rpmGetCwd() and move to librpmio

commit 0bd84c940930b3752eedf967f8c842c270cb33f9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 3 20:48:17 2008 +0300

    We don't support url's in db paths, rip.
    - file:// vs / is silly and anything else wont work, so pretending
      url support just plain bogus

commit 110d354c726ca5e4a66ccbf185334514d452abf2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 3 19:44:49 2008 +0300

    Replace silly static rpmioFileExists() in rpmdb.c with plain old access()
    - accepting url's there is not only useless but just plain bogus

commit c7b5cb2dc4a884aa080a23e1ba0772047ca566e4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 3 19:28:37 2008 +0300

    Remove ancient db1->db3 conversion cruft

commit 1a269ccd3cc3be1bb5763f23621c63b5ac199c92
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Thu Apr 3 13:47:05 2008 +0200

    Don't use static buffers in parseForRegexLang()
    - use dynamic allocation instead to avoid overflows
    - also update addFile(), which is actually the only function
      that calls parseForRegexLang()

commit 58b9fb501d09ade85eae216d26cf51e31bb71e9d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 3 13:17:57 2008 +0300

    Lose the remaining static buffer from rpmVerifySignatures()
    - Simple and stupid: catenate previous buffer + latest message
      with rasprintf() over and over again. Not the most efficient way but hardly
      matters here...

commit 2dc456a10c0dab0d0a00afb0a1d15bf96b74a0d1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 3 12:53:11 2008 +0300

    Typo...

commit ef8669e5bed26597fb2200af54ef93abb6fe138b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 3 12:46:33 2008 +0300

    Refactor signature check results formatting
    - Document what it's supposed to do, should be more obvious from the
      code too now.
    - Remove one level of stpcpy's to static buffer.
    - Simplify the theoretical case where many missing/untrusted keys are
      found (this shouldn't happen in practise ever) by only showing last of them

commit babade159482170bbcb710970277a90e994695fe
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 3 09:59:46 2008 +0300

    Enable stack protector by default if gcc supports it

commit 85571b25e97eed423a9931377aa40bb3fc42fb76
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 3 08:37:20 2008 +0300

    Remove unnecessary check for "keyid:" in result
    - nothing in signature.c emits such string

commit 40686de0a9de6751443f608d77260cbf071c5ff2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 3 08:07:00 2008 +0300

    Start eliminating static buffers from header/signature checks
    - Push msg buffer allocations down to the lowlevel rpmVerifySignature() and
      internal verify*Signature functions, nothing above them knows how much
      memory they need for messages.  Use rasprintf() where obvious,
      stupid malloc(bigenuf) for now otherwise.
    - Changes public API but can't be helped, printing to an unchecked buffer(s)
      of unknown size from one of the more security sensitive pieces is just
      .. not ok
    - Minimally convert callers to the new allocation scheme

commit a0f0da6858782f6a4980437ec05ca37e244ee578
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Wed Apr 2 18:14:47 2008 +0200

    Simplify message key creation in i18nTag()

commit 999248ab03133f3a2b92b64885e691d133a38b36
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Apr 3 08:04:22 2008 +0300

    Shut up a few silly compiler warnings that have crept in
    - missing cast in python header subscript
    - (false) warning about uninitialized variable, unused variable in rpmfc
    - missing include in rpmio

commit 62b46347770c31c6702983777b47d986c1db756b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 2 16:52:44 2008 +0300

    Further streamline / simplify rpmVerifySignatures()
    - Use more meaningful variable names than res2 and res3
    - In verbose mode we just dump whatever rpmVerifySignature() returns,
      error or not. Otherwise the output depends on the result. Make this
      obvious in the code.

commit 8229aa671d972a465440c16647d04664b5265707
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 2 16:30:21 2008 +0300

    Make signature verification result not depend on rpm verbosity level, ick

commit aa3c11428c71a5737dc6ac0cd67d6238cee74645
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 2 15:42:21 2008 +0300

    Further redundancy / copy-paste coding elimination

commit 30d094653ea1834a90f6a74d9b43863b0364c475
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 2 15:36:00 2008 +0300

    Collapse the huge sigtag switch in rpmVerifySignatures() a bit
    - move the tag names from a helper functions
    - collapse similar cases to just one

commit e04c47822773b36d30487c7075880e0dfa7794b3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 2 15:02:31 2008 +0300

    Simplify missing/untrusted keys logging

commit d3c3d4b902cee9cbdcf59aa96848063cc571a2f6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Apr 2 14:25:51 2008 +0300

    Revert "Disable signature and digest checks on query modes (rhbz#438625)"
    
    This reverts commit 5a684f5527bc7f32fca9ba7878802519bb515e93.
    
    All the sigchecking on database queries is mostly waste of time, but
    this disables signature checks on queries of non-installed, untrusted
    packages too. Separate vsflags for db / non-db operations or such needed
    before this can change.

commit 38b81493b044565271a277aa2c7a56408d3c8565
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Tue Apr 1 16:37:42 2008 +0200

    Fix off-by one in urlOpen()
    - the former implementation didn't take the ending '\0' into account

commit aabf8d59854c43f6bf1a11fee2b5100a0917b90c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 1 15:40:07 2008 +0300

    Updated German translation from Fabian Affolter

commit eca89ea370d2c52442d2817a13bc50023f3294f6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 1 13:35:16 2008 +0300

    Enhance error message on failed ts lock acquire (rhbz#427064)
    - adds strerror() output to the failure log message, this should give
      sufficient hint as to what's wrong (eg "permission denied")

commit d1839289023dd4b7cd66f43f8d83b39bb5243ee3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Apr 1 13:32:57 2008 +0300

    Convert rpmlock.c to general rpm indentation style

commit b8c440deade99bee36ba7f2475b85703d896925c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 31 17:06:47 2008 +0300

    Avoid entirely unnecessary static buffer + string copy.

commit 3e4b229e7188a5ee42770089bb9a78dc158b5302
Author: Jindrich Novy <jnovy@redhat.com>
Date:   Mon Mar 31 14:13:59 2008 +0200

    Make ExcludeArch/ExclusiveArch/ExcludeOS/ExclusiveOS actually work

commit e36758060cbdbf463ded17b9b0143b719575244a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 31 13:31:03 2008 +0300

    Avoid static sprintf buffer in rpmReadHeader()

commit 560ee3c8db839495dcbae124f7d08ba77badcc47
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 31 13:12:29 2008 +0300

    Avoid static buffer for error messages in rpmReadSignature()

commit dccc08ab66c044beaba806006302c88c42ad458e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 31 12:07:37 2008 +0300

    Make rpmLeadCheck() return error msg instead of direct logging
    - allows silent checking of lead and differentiating between
      non-rpm files and incompatible rpms, avoiding bogus "not an rpm package"
      message on manifests

commit 7ddab3ab8d55a6d763c823d4b3c090ccf53eb5a4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 31 10:44:23 2008 +0300

    Dead code removal (macro stuff)

commit 802900a230322b8b0c7c9af320dc83a9f6355c6e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 31 10:41:07 2008 +0300

    Dead code removal (popt switches)

commit 33b3da3ec1fb05588a9e7fc54c06c49196867244
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 31 09:56:24 2008 +0300

    Simplify rpmdsThis() code
    - use rasprintf() to construct [e:]v-r string instead of stpcpy games
    - use str2hge() for N + EVR

commit 58afc25abb0ebb7826c00b688fb1825d5b0087c0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 31 09:25:18 2008 +0300

    Copy N + EVR into ds in rpmdsSingle() instead of just referring
    - previously a ds created by rpmdsSingle() would turn invalid as soon as
      the N and EVR strings were freed by caller (or went out of scope), ick
    - convert the N + EVR strings passed to rpmdsSingle() to similar construct
      as is returned by headerGetEntry so headerFreeData will free both the
      pointers + contents.

commit ed07bf20d3bffaa4d99b4b6db2460b1046d7ead3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 31 08:42:33 2008 +0300

    Bunch of formatting & indentation fixes to rpmds

commit de4c3a999fcfa2bcf6772b4165b7e0ff1c8edb54
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Mar 29 15:22:45 2008 +0200

    Fixup for rpmfcSaveArg() no longer freeing key

commit 6cdf14e01d74fd0a65cd3c37ebc805bf43c3740a
Author: Jindrich Novy <jnovy@dhcp-lab-186.brq.redhat.com>
Date:   Sat Mar 29 11:18:30 2008 +0100

    Don't use static buffer ptr as return value in doUntar()

commit 6dfb63cd599145e9c720708041d617187ab18091
Author: Jindrich Novy <jnovy@dhcp-lab-186.brq.redhat.com>
Date:   Fri Mar 28 17:49:56 2008 +0100

    Don't use static buffer in rpmfcELF(), make code more comprehensive

commit 7ff4c18275ab6f5d9e2c88825e13d437fc6f4b91
Author: Jindrich Novy <jnovy@dhcp-lab-186.brq.redhat.com>
Date:   Fri Mar 28 17:21:47 2008 +0100

    Remove unneeded buffer operations

commit 813ab8ca8eb4e39d3fa457755256d42f84edcc59
Author: Jindrich Novy <jnovy@dhcp-lab-186.brq.redhat.com>
Date:   Fri Mar 28 16:15:48 2008 +0100

    Add rpmfcAddFileDep() to handle file dependencies securely

commit 21fee2eb610c0daec72cfdcf12c9bcdcce428431
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 26 16:32:30 2008 +0200

    Simplify changelog entry additions

commit 9c05a57f236ed445fa2dcb5af2b722d8ea45bfec
Author: Jindrich Novy <jnovy@dhcp-lab-186.brq.redhat.com>
Date:   Wed Mar 26 11:52:07 2008 +0100

    Remove unneeded free()

commit e27f6c2bfc05dc80168c5426dac87b297dce2167
Author: Jindrich Novy <jnovy@dhcp-lab-186.brq.redhat.com>
Date:   Wed Mar 26 09:21:30 2008 +0100

    Use rasprintf() in rpmfcHelper() and rpmfcGenerateDepends()

commit 6a7ef512e26c28f28673930649084e7144c4c2a0
Author: Jindrich Novy <jnovy@dhcp-lab-186.brq.redhat.com>
Date:   Tue Mar 25 15:21:47 2008 +0100

    Don't use static buffers to communicate between funcs, avoid buffer overflows.

commit 087c90bfe6b12bebeb17ba2e15a39d5fcc58417a
Author: Jindrich Novy <jnovy@dhcp-lab-186.brq.redhat.com>
Date:   Tue Mar 25 14:57:37 2008 +0100

    Use rasprintf() for NVR construction to remove BUFSIZ limit.

commit 721933384e3e82dc409bbc27d3534543bb10da12
Author: Jindrich Novy <jnovy@dhcp-lab-186.brq.redhat.com>
Date:   Tue Mar 25 13:54:36 2008 +0100

    Avoid artificial limits on tag name.

commit 5a684f5527bc7f32fca9ba7878802519bb515e93
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 25 07:49:47 2008 +0200

    Disable signature and digest checks on query modes (rhbz#438625)

commit 3e76c5767afb48fd81a743bd0bc102eba0b2b611
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 24 23:58:43 2008 +0200

    Avoid couple of static buffers in psm script functions

commit 2a909bdd028ddc2c46c00ddaa3b06003b0329ce9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 24 23:47:41 2008 +0200

    Avoid static buffer for sql commands

commit e501ac2e77ab4f145c75558aa456b7f5f481d254
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 24 22:42:34 2008 +0200

    Remove bogus const from rpmRelocation struct

commit 3e57c865c5d3b65fea8bcddb376e0bc859b598e3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 24 22:25:26 2008 +0200

    Remove bogus consts from rpmte_s

commit a600230b7d614ba8a012bc35a9502949c1722944
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 24 22:23:01 2008 +0200

    Revert rpmhash changes from previous commit
    - rpmhash key + data actually sometimes point to const data.. pff

commit b43b6d6df1027cccdb39d23d2742c66eaafb65c5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 24 22:13:45 2008 +0200

    Remove bogus consts from rpmdb and dbi structs

commit 9a4a5c9ef8f03aca05f7eeed3e7afb297ac72a6b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 24 21:05:05 2008 +0200

    Remove bogus consts from pgpDigParams_s structure

commit e1e66978a1c5871614091d15dfb84439ba90a09e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 24 20:44:48 2008 +0200

    Remove bogus consts from urlinfo_s structure

commit c4659498e3d5833de32c494d83567c9c15e13700
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 24 19:03:20 2008 +0200

    Eliminate bogus const's from spec structure

commit 2cf64e95f2dc2ecac05be776e1e5e3a232d56948
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 24 19:03:00 2008 +0200

    Eliminate bogus const's from OFI_t and spectags structs

commit 4110664044beec1278b512e73f377d4b06f05b90
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 24 19:02:44 2008 +0200

    Eliminate bogus const from expression value

commit fa7e15c2abeaa5a0832a21c75084d2f406fd93ea
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 24 19:01:29 2008 +0200

    Use regular stream functions instead of rpmio in spec sanity check

commit f35277c8a7443d96ae2708abbf1ad74110b5cf8a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 24 19:01:12 2008 +0200

    Eliminate wtf pointer games with rpmGetPath result.

commit 7ae38cdd44f599cd64890580d65d25bffcb543d5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 24 19:00:48 2008 +0200

    Avoid entirely silly rpmio use in addFileToTagAux()

commit 3531bc5d8c356c56cc40b2a82b7a323981537995
Author: Jindrich Novy <jnovy@dhcp-lab-186.brq.redhat.com>
Date:   Sun Mar 23 13:51:53 2008 +0100

    Fix buffer overflow while creating patch script, avoid fixed sized buffers.

commit a3ceec504b4af47b96f90d18dce9df861e0cb93c
Author: Jindrich Novy <jnovy@dhcp-lab-186.brq.redhat.com>
Date:   Sun Mar 23 13:50:07 2008 +0100

    Use rasprintf() to prevent overflows and remove BUFSIZ buffer limit.

commit ca6f318f7a692a8d9f7f404fd6b138e040e1a2b2
Author: Jindrich Novy <jnovy@dhcp-lab-186.brq.redhat.com>
Date:   Sun Mar 23 13:48:49 2008 +0100

    Fix writeRPM, use dynamic allocation, remove bogus const.

commit f779b62276205a116814bde15613fa7383063967
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Mar 20 10:59:17 2008 +0200

    Adjust maintainer makefile for hg -> git switchover

commit 53bc909971af88b463ce39f4b30244ceb36e0c3e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 19 09:30:33 2008 +0200

    Remove bunch of double consts. What have I've been thinking?

commit 5c7cccd3b6631009060fae1511b5de2ae987daa0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 18 13:22:15 2008 +0200

    Erm, off-by-one in rasprintf()
    - should probably change the comment to "simple, stupid and buggy" ;)

commit d07bcfece15f9e8352ea077e1a7fe8a082bc2eab
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 18 10:28:57 2008 +0200

    Bump up sonames, we're not exactly ABI compatible with 4.4 anymore..

commit 5c48f597c82c92a02f573056e26bd21fe17aeda3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 18 10:18:08 2008 +0200

    Use rasprintf() in rpmps to properly fix mdvbz#31680

commit 053cf9a25e341dec8e1beb4811ecb1e9b11715fe
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 18 10:17:03 2008 +0200

    Add custom asprintf() clone to rpmstring

commit c908277be21c27a0894404821605134958521fd6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 18 09:10:13 2008 +0200

    Minimal namespacing for locale-insensitive x*() string functions

commit 9df54a8622b6640e54288835e62c6c0d0b23c748
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 18 08:55:05 2008 +0200

    More xstr[n]casecmp to rpmstring along the other string-stuff...

commit 5cde7cac9619c90a83d2f63f9d3760d0a352226e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 17 16:03:51 2008 +0200

    Refactor running of pre/post scriptlets out of rpmtsRun()

commit 5dca5fac8fa29d50d92a5939fe60cbdcc34c0cf4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 17 15:48:25 2008 +0200

    Add missing popt include

commit d9c780d19edeaf81427411205a9620c9516902c1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 17 14:36:43 2008 +0200

    Remove rollback and repackage from documentation

commit db37b74a2bfebf067f421b891d3837baaca7fc6f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 17 14:27:03 2008 +0200

    Nuke leftover rollback test program, update translations

commit d89482a7c7777d379ab415ebba747b6af19babdb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 17 14:10:11 2008 +0200

    Remove repackage support to finish off with rollback
    
    The two main classes of rollback/repackage need are
    1) Gimme back the previous version, this broke XYZ on my system.
       This is better handled by downgrading to the previous version of original
       package instead of repackaged garbage. We should maintain real package
       (version) history somewhere.
    
    2) Upgrade messed up my configuration.
       Instead of repackaging everything we could be far more intelligent
       wrt config files, stick them into real version control or at least have
       hooks to do so.
    
    RPMCALLBACK_REPACKAGE_* and RPMTRANS_FLAG_REPACKAGE definitions left around
    to avoid needlessly breaking everybodys callbacks (for now)

commit 37543ee0405c38a9d54a12bf0d54698773a3933c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 17 11:53:27 2008 +0200

    Remove rollback support
    - not possible to do reliably within rpm
    - effort is better wasted on investigating fs-level snapshots, which
      is way beyond rpm scope except for hooks to interact with the snapshot
      mechanism to communicate beginning/end of transaction and such

commit ed78cef2c3e14df29d95b476f00d51203fbe098e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Mar 17 12:04:39 2008 +0200

    Remove support for automatic rollback on failure
    - doing this reliably from rpm is simply not possible as there's no way
      to undo script actions, might as well not pretend we can
    - for a feature that's not generally usable it complicates mainline code
      way too much

commit 874f6849d9f5f71b3815e358c6c5a049f6a4fa66
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Mar 14 14:55:44 2008 +0200

    Make arch conditional in default queryformat
    - gpg-pubkey's dont have arch, argh...

commit 6db2a8ec03c057850a0349537c705556068b9d32
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Mar 14 14:27:01 2008 +0200

    Revert RPMTAG_FILEDIGESTS -> FILEMD5's for now

commit edec73e2128976c344d31953a3d2de698ff7de33
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Mar 14 14:18:43 2008 +0200

    Type missing for RPMTAG_HEADERI18NTABLE

commit 8e64df0e83029d08dfbd9a09ef76bed817d1ccf3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Mar 13 09:20:24 2008 +0200

    Don't process objdump sections after "Version References"
    
    autodeps/linux.req expects the 'Version References' section to be
    the very last section in objdump -p output and nothing else to follow
    it, but at least on Linux/ARM, you can get output a la:
    
            Version References:
              required from ld-linux.so.3:
                0x0d696914 0x00 06 GLIBC_2.4
              required from libacl.so.1:
                0x05822450 0x00 05 ACL_1.0
              required from libgcc_s.so.1:
                0x0b792655 0x00 04 GCC_3.5
              required from librt.so.1:
                0x0d696914 0x00 03 GLIBC_2.4
              required from libc.so.6:
                0x0d696914 0x00 02 GLIBC_2.4
            private flags = 4000002: [Version4 EABI] [has entry point]
    
    This causes the built package (if the external dep generator is
    used) to end up with bogus dependencies.
    
    Issue found by Rabeeh Khoury <rabeeh@marvell.com>
    
    Signed-off-by: Lennert Buytenhek <buytenh@marvell.com>

commit f1f6510ceecc99fb4d3fc0c840e5b87232566174
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 12 15:34:49 2008 +0200

    Remove rpmio internal copy of poptParseArgvString() & friends
    - the old included version segfaults with funky input and whatnot...
    - link rpmio to real popt instead

commit ef84de8868a6547ecb8e668d035d4f47bc8cd4af
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Mar 12 13:20:47 2008 +0200

    Refcount dbiTags on db open+close, automatically free
    - avoids rpmdb_internal include in rpmqv.c and memleak in API clients
    - avoids having to export dbiTags(Max) at all

commit dd2f0eb3c8a94485098cfd4412a1a14f28335d02
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 11 11:33:04 2008 +0200

    Update translations

commit e9d194444eee9e533c722a6e67ccb100ad99e868
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 11 11:32:37 2008 +0200

    Dust off Finnish translation

commit bbeb5148f2d8d6b9a401ef7c375638a69b60f59e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 11 08:42:34 2008 +0200

    Updated Polish translation (Piotr Drąg)

commit 35c7ef6d6bc776977248958dee052812841c4a5b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Mar 7 15:07:06 2008 +0200

    Remove support for automatically importing pubkeys from keyservers
    - causes hideous load on keyservers and autoimporting without user approval
      isn't a very good idea, especially as long as imported == trusted
    - nothing wrong with supporting hkp lookups on manual --import though...
    - curl doesn't know about hkp:// but http on appropriate port works just
      fine, lets re-enable it

commit 2414d522bcfbd0ba9e6115eb7e2a3ba056ac2795
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Mar 7 14:05:28 2008 +0200

    Remove BDB copy from the repository, it doesn't belong there

commit 501197e5ef5ce8687aaf8bd4352f296bb7a5c0e8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Mar 7 13:34:20 2008 +0200

    Stop insecurely covering up for lack of snprintf and vsnprintf
    - we require C99 anyway
    - snprintf is used unconditionally all around the tree
    - if absolutely needed for portability, having insecure wrappers littered
      over the source is not the best way

commit de4085814e7bf93042f2fba06ca24a4e63a4b27a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Mar 7 13:04:46 2008 +0200

    Drop librpmbuild from rpm.pc
    - next to nothing actually uses librpmbuild, avoid dragging it in needlessly
    - if needed, add a separate .pc for rpmbuild later

commit 0c0d812b0da674f237ad5932383a44befcf7f3a6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Mar 7 12:36:22 2008 +0200

    Remove unused RPMUSER, GROUP etc from configure.ac

commit a08ececd832de2fb4de426d8894abccd45a7c3df
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Mar 7 09:47:51 2008 +0200

    Log message adjustments
    - dont waste translator time by having them translate obscure debug messages
    - adjust some suspect looking logging levels
    - mark some non-debug messages for translation where missing..

commit e97e5619e9ddaf8b0e78568ca8848e2c2840dd00
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Mar 7 08:29:03 2008 +0200

    Further constants made readonly...

commit f686d184df613c74df678cbf3a878c25ccd4fec8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Mar 7 08:21:55 2008 +0200

    Make _cacheDependsRC static, not needed outside depends.c

commit 3be0a4b2f863e213af45b0495254108c13415a1f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Mar 7 08:20:03 2008 +0200

    Another pile of read-only constants marked as such..

commit 2e3972aebf3db7a6fce83619ea106b80145ed923
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Mar 7 08:15:03 2008 +0200

    Mark another pile of header_magic duplicates as read-only

commit 8e7bc88eda21acdee7552980738cbded4831f950
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Mar 6 20:47:17 2008 +0200

    Make yet more build tables readonly
    - preambleList and partList are modified runtime for silly reasons,
      reminders to fix later..

commit ba16c1bf2fdd1d70b5c8fed1ba88d967f941aa50
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Mar 6 20:25:43 2008 +0200

    Make headerCopyTags() tag list const, the list isn't modified..

commit 82a6fac62cf4dbc9f131d80271fab01cc55fd700
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Mar 6 20:17:20 2008 +0200

    Mark bunch of static build-related tables as read-only

commit 16686e9663d9538abb720db955e559f2b35da7b1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Mar 6 19:57:01 2008 +0200

    Make rpmfc token and apply tables readonly

commit 403dae827c313760cb1968556c1aeaf395236e98
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Mar 6 19:51:06 2008 +0200

    Mark rpm tag tables read-only as far as we can..

commit 146f12c05beff77c5e1a19989448ca3d12446a06
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Mar 6 19:50:41 2008 +0200

    Mark various header tables and definitions as read-only

commit 27f5763cfeb155450b6edf9a3e5dafce07024b99
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Mar 6 19:49:08 2008 +0200

    Make the rpmdb vectors readonly

commit 8fde231cac3a597a9199fd36c5c597555895644c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Mar 6 19:27:19 2008 +0200

    Mark bunch of misc. tables & such read-only in librpmio

commit 17fc6663ef4bd647934876638f3aa2edf34d3ce4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Mar 6 19:25:46 2008 +0200

    Make rpmio vectors read-only

commit 5b32005575079da789d49869870066e8784e6924
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Mar 6 19:24:56 2008 +0200

    Further read-only tables in rpmpgp..

commit 84ff15177d91f8d06615edccb851ff40f0eddb12
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Mar 6 18:56:38 2008 +0200

    Mark various pgp value tables read-only as they should be

commit 2f833b0a05b5823edeb935196bbcc3faf5d68228
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 4 09:26:27 2008 +0200

    PreReq -> Requires in ordering debug output (Pascal Rigaux)

commit 61189cb868c508c143a0a6c4f1a73de2bf0aeb30
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Mar 4 08:16:02 2008 +0200

    Add super-H arch support (rhbz#432496)
    
    Patch from Masaki Chikama

commit 28cba016c813a50b913a6ec484d4a428026c96a7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Feb 28 12:46:45 2008 +0200

    Add missing leading underscores to names passed to rpmMkdirPath()
    - mostly cosmetical, leading to slightly misleading error messages

commit 18d01d8f905178c79aae028e92ee5d8433ac857c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 26 22:42:46 2008 +0200

    Another unnecessarily exported internal variable..

commit 05129f83dc4d07c4bc507f66496ac3fc068deeb2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 26 22:24:28 2008 +0200

    Make bunch of macro internal variables static

commit b010266b0ea24ddbdb087d37db9d7ee7d2a34e83
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 26 22:04:17 2008 +0200

    Remove hidden --nolibio cli switch, avoid exporting noLibio

commit 2579e086c2be79d2d3f7e0e8ee3c25c25edd9a7b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 26 21:46:38 2008 +0200

    Un-inline fd*Digest()
    - avoid leaking nss + digest internals all over the place

commit 47a1e0515012d4aaf4f95fd71c4bc6de526e8d86
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 26 21:02:21 2008 +0200

    Make the rest of rpmio vectors static inside rpmio
    - no external needs, gets them out of exported library symbols

commit edc6b13d19d832ec56b63007ccc84a0f43cf2629
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 26 20:53:38 2008 +0200

    Remove the last remains of unused and unnecesary rpmio rpc vectors

commit 41ebd1cd50c780c247877bf9174aacdd55ca4923
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Feb 25 20:16:43 2008 +0200

    Eliminate rpmdb refcount debug junk from the API

commit 89f59abcd306aa7cff9b48936469d7e46bb655e1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Feb 25 12:33:19 2008 +0200

    Remove some ancient dead code

commit 98c5039009406609f661f98623776d6c18a04c2c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Feb 25 12:24:22 2008 +0200

    Add %{sources} and %{patches} macros to build
    - push full patch + source paths to global lua tables
    - %{sources} and %{patches} are just easy-access wrappers to accessing
      the lua tables
    - "git-am -p1 %{patches}" and enjoy...

commit 2b39ddc25d9773c793c4ceaef01de0b9df0c06ad
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Feb 25 11:03:38 2008 +0200

    Enable variadic checks for rpmExpand(), rpmlog() and rpmGetPath()
    Suggested by Michal Marek

commit c46876daa4358849de73a059d944f753a7584013
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Feb 25 11:02:07 2008 +0200

    Add new header for collecting utility macros
    - atm consists of portability wrappers to various gcc extensions like
      __attribute__() to provide additional warnings and checks
    - copied from glib to avoid an extra build dependency just for the silly
      macros, renamed to avoid namespace clashing

commit fe35727d3f55f830d21b71b183ea24974a02696f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Feb 22 15:39:24 2008 +0200

    Remove ancient commentary from header.h
    - history is interesting but public headers aren't the place for it
    - "implemented soon after" from 2001 comments are hardly useful :)

commit e59b3817bbbd530b4d6a003bc02a7ee8166eaa72
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Feb 22 15:37:16 2008 +0200

    Remove useless ifdeffed typedefs
    - rpmdb & iterator are typedeffed in rpmtypes.h already..

commit 03dcaf61eb48b12bc65322b40b30dc06e8a2ccc0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Feb 22 15:30:51 2008 +0200

    Rip ancient, unused setperms + unsetgit make targets

commit d0feb99eac9404648a9bde49f09042ade7fcdfbb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Feb 22 15:29:04 2008 +0200

    Use %__tar instead of hardcoded tar in build prep

commit 4dcc991ecc30e33f6f468885d8e92479d6dd3143
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Feb 22 15:08:41 2008 +0200

    Don't use legacy gzipbin etc macros in build

commit 3b7c968326190d2997ff3866f592b7003e889b3a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Feb 22 15:05:16 2008 +0200

    Make %patch use %__patch instead of hardcoded patch.. patchy huh?

commit 82f39bc8888a765dd4efb1b3539cc2bbac684ed9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Feb 15 15:21:22 2008 +0200

    Shut up another case of FUSE mount stat whining (rhbz#432907)

commit e93d76665d1ca0f7388ce22f1da1c38ce0900dac
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Feb 14 15:02:18 2008 +0200

    Improve header i18n locale matching
    - Get best lang from rpm HEADERI18NTABLE, instead of getting first fuzzy match
      (eg: zh_TW matches zh_CN whereas zh_TW entry is available)
    
    Patch from Pascal Rigaux

commit 9f8aa63490ac5091e77b9f2b7427f58a299a0eb4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Feb 14 13:10:51 2008 +0200

    Fix a bunch of string format mismatches found by gcc's variadic checks
    
    Patch from Michal Marek

commit 7585967f4707f4540517fca498b9c8683cf683cc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Feb 11 20:45:15 2008 +0200

    Play nice with xargs
    - xargs considers exit status 255 special, avoid it. The exit status is
      unrealiable anyway so 254 vs 255 shouldn't matter
    
    Patch from Michal Marek

commit 9dc19803d5495b598db0edb4b650d672c2a1f024
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Feb 11 20:12:32 2008 +0200

    Updated sr translation + new sr@latin version from Miloš Komarčević

commit 36e6679eb09bd93782098f653771e75939313c72
Author: Peter Jones <pjones@redhat.com>
Date:   Fri Feb 8 14:58:26 2008 -0500

    Don't add syntax errors while I add sorting...

commit ee6b38a9d11b8737f692aa5f954ac198bf4c46f1
Author: Peter Jones <pjones@redhat.com>
Date:   Fri Feb 8 14:54:29 2008 -0500

    make gendiff sort the file list, so comparing the results is easier.

commit 82b547aa26aac6e0580dfaedbbc84b99c558f586
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Feb 8 13:45:56 2008 +0200

    Oops, package type in lead was reversed...

commit d718b12f5bf5b94c418b8235e45625cbccca6220
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 5 17:42:19 2008 +0200

    Mass convert (back) to rpmTag as it's usable everywhere now
    - cast away a few cases where the enum usage causes ridiculous amount
      of compiler warnings from unhandled switch-cases

commit 53ca9c775f5f3954f140f462d22bb28cde02c6cd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 5 16:35:44 2008 +0200

    Use rpmSigTag enum for signature tags everywhere, more or less..

commit 130118d594ef2dfb02e5a2938fe4e1da1b3e6955
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 5 16:14:34 2008 +0200

    Use rpmTagType always now that it's available everywhere
    - RPM_MASK_TYPE is not a datatype, move it out of the enum
    - add missing cases python header handling to shut up gcc

commit bbcf8f5ff09a11ba9dcf688c923afcfd62dd9d4d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 5 15:32:33 2008 +0200

    Couple of missed type abstractions

commit 36e9f885e161395e2dfb033dcd7db1667764f273
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 5 15:29:37 2008 +0200

    Yet another typedef for transaction ID, use where spotted
    - most remaining [u]int16|32 uses are indexes and such, ie actual numbers

commit ceb96638f6ceba45f353c132b104543404515fed
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 5 15:15:41 2008 +0200

    Use rpm_flag_t for header generic bitfields (file flags etc)
    - additionally use rpmfileAtts / rpmVerifyFlags consistently where
      appropriate

commit bb72efaaaa6dad878877eeb12cbad67383a0cf35
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 5 14:50:42 2008 +0200

    rpm5.org: free 5 bits for rpmfc coloring expansion
    - also fixes an overlap between RPMFC_LIBTOOL and RPMFC_MONO which we
      had and were out of bits already

commit 455c428c066801431e98d70d5f57bf3ca3f37174
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 5 14:28:06 2008 +0200

    Couple of int -> rpmVSFlags type corrections

commit 4391fce5e16459ce7bf52eb601c2518f3d456d41
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 5 14:25:16 2008 +0200

    Doxygen bits for rpmtypes, much to do still

commit 5d38313e7c26bf628389382b78ef8f0db138e787
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 5 13:53:19 2008 +0200

    Adjust potfiles from rpmlibprov removal, update...

commit 9849c79d8d9c26f52831fd361bc3569926605a92
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 5 13:52:30 2008 +0200

    A missed color type in rpmds header

commit 50e2d4c26f3534baa6ec9be354e710a6e0ad6d88
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 5 13:49:13 2008 +0200

    Split non-cli verify bits from rpmcli to separate header

commit 9b0967588abde9599751a6ede176fa429202dd72
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 5 13:09:13 2008 +0200

    More rpm-headerdata types: rpm_dev_t and rpm_ino_t, use where spotted
    - both hardwired to 32bit internally

commit e5116e43d753114f9966ff42f507310a61f620eb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 5 12:58:05 2008 +0200

    Add rpm_rdev_t type for rdev entries
    - rdev is 16bit whereas dev is 32bit, eww...

commit e7ac5fb183c4c0e5762d7f18c1c19c03fcdc2b3c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 5 12:31:32 2008 +0200

    Add rpm_mode_t type, use where spotted
    - bring the types under some sort of control, previously mixture of
      int16_t, uint16_t and unsigned short was used..

commit 630a5a4c1a34f12ae264cba870005de0858353e6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 5 12:12:31 2008 +0200

    Add rpm_color_t type for rpm color bits, use where spotted

commit b12cb085232a586ebca8673ee766d9d6e826c8dd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 5 11:43:38 2008 +0200

    Port some rpmds improvements from rpm5.org
    - rpmlib provides as rpmds, lose the special case hackery
    - new functions rpmdsRpmlib() and rpmdsSearch() + python bindings for them

commit e1ca3be052b4f0bbc417489c3f97a00323630469
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 5 10:08:44 2008 +0200

    Rip the horrible (ia64) autorelocate kludgery. RIP.
    - we don't support relocated x86 on ppc in qemu either...

commit 5786cf78db37d9b129470cefdce5c0faf2c5c183
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Feb 5 09:41:11 2008 +0200

    Use helper function to eliminate some copy-paste code

commit 21ff694d9ccacd86c1e570fdc3af030e98d6da10
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Feb 4 12:33:02 2008 +0200

    Remove double documention for exported header functions

commit df1467c357aa18b4bc6a43df4832460d302054a4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Feb 4 12:28:02 2008 +0200

    Rip the useless header method vectors

commit 1df890dd0230ffceba86832b320d377b4d505cfb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Feb 4 12:00:30 2008 +0200

    Make check-programs build again, assorted type-fiddlings..

commit aca6d4a2c4cb5ce76031884933661115c80ca6c1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Feb 4 11:50:39 2008 +0200

    Add rpm_time_t type for time types (in headers), use where spotted
    - easy to spot, easy to change...
    - time_t size can vary, header data needs fixed size

commit e886043a1027fabf975f5e6542437e412a137e67
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Feb 4 11:27:00 2008 +0200

    Add rpm_off_t type for file size types, use where spotted
    - preliminaries for bumping up the max size, use rpm-specific type
      as off_t size varies, header data needs fixed size

commit f53c32337238b84bb2c555ca12cdc07f6a963361
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Feb 3 16:24:07 2008 +0200

    Make rpmtag.h standalone, include from rpmtypes.h
    - permits header.h to use the tag types
    - revert other includes back to header.h, more crazy #include churn...
    - rpmTagTable etc in header.h for now, don't really belong there but...

commit abce05919d417b026bedd81d3a08f91625392530
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Feb 3 16:09:20 2008 +0200

    Use rpmTag etc enums for rpmTagGet*() functions

commit cf34aa5a37eedb79af6ab0f0cbf86a15c56dcb86
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Feb 3 15:41:52 2008 +0200

    Stuff signature tags to rpmtag.h along with other tags, add enum typedef

commit db3f67ed5b53452a749427a16254a2d231522063
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Feb 1 20:28:31 2008 +0200

    Fix platform dir creation, doh

commit 7ffa128552d62cab163b534d1c7a8408b2baf6c2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Feb 1 20:23:25 2008 +0200

    Update translations for a change..

commit 7a77e1fd9fcd657fec32704628b77f6104cf5103
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Feb 1 20:18:23 2008 +0200

    Move platform specific macros under RPMCONFIGDIR/platform/
    - reduce the clutter in RPMCONFIGDIR, nicer for packaging too..

commit 15f33b33e62a703105671dff675689265da88b16
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Feb 1 19:53:04 2008 +0200

    Ditch the "system" %_topdir, default to ~/rpmbuild instead
    - building as root should be discouraged and /usr/src for building stuff
      is horribly wrong anyway

commit cc9880edc2ddf4d390f59b8b111222d42f73a816
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Feb 1 17:27:22 2008 +0200

    Don't close GPG/PGP stdin to avoid harmless warning (rhbz#197602)
    - patch from rpm5.org

commit a23502798fb1091b5ede3f1886c2dc8ed3e9e8cf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Feb 1 13:25:18 2008 +0200

    Avoid silly fixed size buffer in urlhelper

commit cc69dd20b8d0afdc09957c84eae0957ab1552618
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Feb 1 13:00:40 2008 +0200

    Allocate macro buffers dynamically
    - avoids some entirely needless overflow cases
    - somewhat similarly to rpm5.org but use heap instead of stack

commit 56405587f4b3890b34243904629cfa55cb345ce6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Feb 1 12:11:03 2008 +0200

    Allocate various macro buffers from heap not stack

commit 54ab76019e8656950c9f9bf144edd549805dee0b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Feb 1 11:36:13 2008 +0200

    Define + use MACROBUFSIZ for macro expansion buffer size
    - double the buffer size while at it...

commit d3d60512bd1df6ed5027a9dedf1d1ab97f969ad6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Feb 1 11:09:13 2008 +0200

    Macro shell expansion fixes (rhbz#431009)

commit fbc9b689fb7c30cabe51ec7b18bd692451249df1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 30 17:20:09 2008 +0200

    Add rpmtypes to doxygen, with todo reminders..

commit eb5dc35c196c2b18e934a046627ee59d54e175d6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 30 17:05:29 2008 +0200

    Include spring-cleaning
    - put some consistency into include ordering
    - everything (apart from bits missed ;) is now ordered like this
      1. "system.h"
      2. other system includes
      3. rpm public headers
      4. rpm private headers
      5. "debug.h"

commit d2efb5d77c0a3a2f5919d40efe5e163254422b6f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 30 16:04:40 2008 +0200

    Use rpmsenseFlags type for dependency flags everywhere (hopefully)

commit 87e01f336262f3738211c0426304590ed16284f8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 30 14:27:41 2008 +0200

    Use headerIsSource() instead of doing it manually

commit 99faa2735b4b52104a724c6a2f8a7a806bc9c7e5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 30 13:53:51 2008 +0200

    rpmlib.h mass eviction
    - explicitly include what's really needed instead
    - document remaining uses

commit 7751f84bf03076eb6ad90ad0d10144846726bdfb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 30 13:03:54 2008 +0200

    Move HGE and related prototypes to header.h

commit ecabf137102a980528f166b9690efb8adbcef95f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 30 13:00:26 2008 +0200

    Move rpmTagType stuff to rpmtag.h now that they can move

commit 6af8116e7268eb20d5127796d61e55d96360171a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 30 12:59:38 2008 +0200

    Stick rpmCallbackType to rpmtypes too..

commit 118db841210eb3809b6640a95b03214b0f3bc648
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 30 12:51:09 2008 +0200

    Add missing system includes

commit c482427fbfd604d665e7fd28d77cd58a22b089ae
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 30 12:42:07 2008 +0200

    Uninline headerFreeData()

commit 78f1f6378e3d18c1b46dc22125c6d3e0808f26af
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 30 12:21:59 2008 +0200

    Stick relocation struct to rpmfi, typedef to rpmtypes...

commit 90fbda71036fe0614558750fb4fa862c3b8f1413
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 30 12:20:59 2008 +0200

    Move FD_t to rpmtypes.h, used all over ..

commit e97aefc89bdd10616a8bd0fce642a93148ab357a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 30 11:55:06 2008 +0200

    Move rpmRC definition to rpmtypes
    - lot of places drag in rpmlib.h just for that...

commit 73ebfe9ad49e56230393ff2ffa818436b40fdedd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 30 11:51:28 2008 +0200

    rpmgi.h doesn't need rpmte and rpmts headers, implementation does

commit 458bf3b208e3efdd880db42c62f3b12b4eed4da0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 30 11:49:25 2008 +0200

    Remove bunch of unnecessary rpmlib.h includes
    - replace with rpmtypes.h where needed

commit e4d43147ee34310a7753383ffde0e24737fddc1c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 30 11:02:39 2008 +0200

    Move header and -iterator to rpmtypes

commit 8f4000b00b9e1a4c7db2eeca71fec1b47e3e0b63
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 30 10:57:15 2008 +0200

    Move rpmalKey and rpmal to rpmtypes

commit ebbcd155e71ef8ccfe4d3a4f6b6f00f67f089cda
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 30 10:54:04 2008 +0200

    Move fnpyKey typedef to rpmtypes, its widely used...

commit 3952ef87858c56094d8287d542f3aedd4cef6987
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 30 10:33:32 2008 +0200

    Move rpmts and other major data typedefs to rpmtypes from rpmlib

commit 3564b3a5ed2f2a6fe42fcab5af96f3cbc244fe96
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 30 10:23:47 2008 +0200

    Add missing c++ ifdefs to rpmtag and rpmtypes headers

commit 66345bfc5a141f30625746d087ce9d721746c8b8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 30 10:16:14 2008 +0200

    Move header tagdata types from header.h to rpmtypes.h

commit 0760e7389fc6c574cbaabb5788e29a69d3fa9dfb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 30 10:13:09 2008 +0200

    Add stub rpmtypes.h header stub
    - place to stick in common typedefs to untangle header cross-include jungle

commit 2adf19b3d3910b9ed579f89144dba0263e277c60
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 30 09:36:02 2008 +0200

    Remove fts flag API symbols from help messages
    - a --help message isn't an API reference...

commit 265b1f017fa7bbe7c6e2f92c910465c38eba8aa6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 30 09:24:43 2008 +0200

    Updated Polish translation (Piotr Drąg)

commit 87458bd399a8696cd0b6e981bb60a2f4de1c48d8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jan 28 15:10:14 2008 +0200

    Make rpmbuild -bs and -ts imply --nodeps (rhbz#235507)
    - better fix might be creating srpm header early but until then...

commit e39b131750525e0f5a172ebee55154db6b71be7b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jan 28 13:43:32 2008 +0200

    More missing system includes for size_t & friends

commit 80ca9f368f5ed417c57f524e88fa75f2b2ccc991
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jan 28 11:45:38 2008 +0200

    Add missing system includes to rpmmacro.h

commit 339f85db55c08d6cc28ec0076d4e8680813f15fc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jan 28 09:20:33 2008 +0200

    Fix printing of filenames with whitespace in check-files (rhbz#430428)

commit f2c6bef540faf4ba9b7a15b284a5d64f9c8b00fa
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jan 25 17:33:35 2008 +0200

    Urgh, add geode to x86_64 secondary arch subst-crack...

commit 43636bc513cc6f2a5ebf92015e1d74be981a2225
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Jan 27 17:28:32 2008 +0200

    Document libmagic requirement

commit 95b03441ccd5483503d9f5455655c256d0932007
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Jan 27 16:39:40 2008 +0200

    Rename isCompressed() -> rpmFileIsCompressed() for namespacing

commit 8e2965886f6347850b6e769cb27f4fb3ada7677d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Jan 27 16:13:47 2008 +0200

    Couple of missing doxygen groupings

commit a0a972702a940175b3c3ca2491c6e4d0a9461e63
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Jan 27 16:11:46 2008 +0200

    Detect lzma magic if it exists, otherwise dumb check for .lzma filename
    - Newer lzma-utils make a magic header in archives, current stable versions
      don't. Guessing based on common compression flags used by current lzma
      versions is feeble and futile...

commit c693169deeabf8f9a65a824699cc0c98cedd491f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Jan 27 16:00:40 2008 +0200

    Use the suffix checker in rpmfc instead of inline defined version

commit f0f392cce6c07fd4a41a392bc912a76238ee14eb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Jan 27 16:00:06 2008 +0200

    Helper function for checking given suffix on path

commit c63f859b78b7465aed64a85635a3d1d5e0f7b9d7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jan 25 13:54:39 2008 +0200

    Quote find-debuginfo builddir to avoid whitespace issues

commit 70e793db8944a9db0e6f49ce012c28dbda973369
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jan 25 12:24:55 2008 +0200

    Don't reset target when reading included rpmrc's.
    Doesn't seem to break anything and fixes rhbz#232429 AFAICT...

commit acf2fef3084c23e13751c39f3eb48dca12af947a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jan 25 12:20:56 2008 +0200

    Hackery to get secondary arch macros included on x86_64 (rhbz#194123)

commit 0d7bf59ad10cffb3d99fb309fd23083721ade106
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jan 25 12:14:23 2008 +0200

    Bandaid patch for truncated error message in russian (mdvbz#31680)
    
    Just up the stupid static reservation "high enuff" for now.
    asprintf() would be nice here but it's not at all portable, revisit later.

commit ee655d91a2dfd5e8863dca582077d1bd65cb240b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jan 25 12:01:16 2008 +0200

    Permit conflicting ghost files
    Patch from Pascal Rigaux

commit 86c2aa8cd638be53d35a22b4fc1302be434b8851
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jan 25 08:30:32 2008 +0200

    Print regular provides on build too (mdvbz#36672)
    
    Patch ported from rpm5.org by Pascal Rigaux

commit f62311615186365733b37681a8c010f12bea79f6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jan 24 09:34:33 2008 +0200

    Missing space in russian translation (mdvbz#36974)
    Patch originally from Andrey Borzenkov

commit eec03952be8cea647d1e4a2147f49e41a3051983
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 23 07:56:51 2008 +0200

    Make uncompress macros match reality (_ vs __ prefix)

commit 240b2d3441ce0711b0eacdc5345b3a4d9b77a716
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Jan 19 11:40:59 2008 +0200

    Add support for Geode CPU (rhbz#428979)
    
    Patch from Dennis Gilmore

commit 735c83ad5291858cc981f739c3db5165d50598d5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jan 10 13:39:00 2008 +0200

    Define and use UGIDMAX instead of hardcoded 1024 all over

commit 76d8c768363731b89da2c5db7625d2369adc5175
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jan 10 13:35:03 2008 +0200

    pgpHexStr() returns pointer to static buffer, return const char *

commit 6243ccd1e67f23e3ff07438c34ce18d5893b51df
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jan 10 11:48:31 2008 +0200

    And *really* skip the leading % in --define, double duh :)

commit 9672ca6925f51d4c6dc623ad1d6eda83a20d5461
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jan 10 08:44:21 2008 +0200

    Fix thinko in macroname underscore conversion, duh

commit 927865ad4a370f316b58c6322e7e9bfc8a64553d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 9 18:56:34 2008 +0200

    Document internal vs external BDB usage in INSTALL

commit 924007789d9ae450610cb839eb3f244580105d36
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 9 19:07:20 2008 +0200

    Drop internal db from dist tarballs
    - tarball down to almost 1/4 of earlier size
    - check if internal db is present in configure, point to INSTALL if not
    - distcheck needs --with-external-db now to have any chance at working
    - leaving to hg for now for easy reversal if getting second thoughts ;)

commit d4ec2ad87f544e18be164964ab8b04cb5ed3ad0f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 9 17:50:25 2008 +0200

    Dirty hack to determine BDB .la name automatically

commit b28ad6b47bda0c0ac49fdd10bf5f4f0c5e09964f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 9 11:11:49 2008 +0200

    Implementation of triggerprein scriptlets.
    Ported from rpm5.org work of Jeff Johnson to 4.4.x by Pascal Rigaux,
    further adopted to fit HEAD...

commit 478ec787f36ddac045dceb5727a699b7359bbbe7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 9 10:39:55 2008 +0200

    Permit files with glob characters in *.rpm packages (#147383)
    Patch from Alexander Suvorov

commit 88c020f771f477aa70b7251dde762cd8fb6fbf0d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 9 10:34:08 2008 +0200

    Avoid access(2) quirks querying symlinks, lstat(2) instead (#60288)
    Ported from rpm5.org work of Jeff Johnson

commit a97100ba1dae7849571f30075d7564fb0124dc91
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 9 10:21:01 2008 +0200

    Handle spaces in file path arguments correctly (#217258)
    Ported from rpm5.org work of Jeff Johnson with some differences:
    - place into rpmfileutil instead of rpmgi as it's more generic than just rpmgi
    - rename rpmgiEscapeSpaces -> rpmEscapeSpaces
    - return char *, not const char * as the return string must be freed by caller

commit 4906b9c55e97a9db8f5db4691a09253ddfadd976
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 9 09:28:09 2008 +0200

    Convert '-' to '_' within --define macro names (#124995)
    Patch derived from rpm5.org of Jeff Johnson

commit ce9ba5a7da2a53795f7fa71a593c645debf8f74a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 9 09:22:31 2008 +0200

    Last 8 bytes of RSA modulus is keyid for V3 pubkeys (#205080)
    - ported from rpm5.org work of Jeff Johnson
    - minimal fix suitable for 4.4.2.x too, full rpm5.org changes require
      api changes which would be ok for HEAD but that's a bigger task...

commit 02824f0e9d9c818c46682c13502a7eab5a4ebb0b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jan 7 16:09:20 2008 +0200

    Set a default 0022 umask value always (#83006)
    Patch derived from rpm5.org work of Jeff Johnson

commit c80bd01564129d531107f9f8458fc7c07e64bba5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jan 7 15:37:43 2008 +0200

    Print --usage to stderr instead of stdout (rhbz#164021)
    Patch derived from rpm5.org work of Jeff Johnson

commit fcf6b50378eaeac4c1f7ca215b33586b4d41072f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jan 7 14:48:57 2008 +0200

    Don't unlink non-temporary file in FSM_UNDO (rhbz#223931)
    Ported from rpm5.org work of Jeff Johnson.

commit 3f9cde75df2c4bf7c1de3dd309d3d43fe22d0df4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jan 7 14:25:24 2008 +0200

    Fix field-width specifiers when no space between (rhbz#323221)
    Ported from rpm5.org work of Jeff Johnson.

commit 37bff306ae3ca6e7402939748fc9b3ea062851fb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Jan 7 13:57:10 2008 +0200

    Handle "use vX.X.X" in perl dependency extraction (rhbz#140597)
    Patch by Steven Prichard

commit a93afe303547408a6a477c08e7e7b8e53b3f91d1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jan 4 12:04:49 2008 +0200

    Saner RPM_FORCEFREE_TYPE definition

commit 243e97302da9e9029f417fa3c698a2973a4e65b7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Jan 4 12:00:51 2008 +0200

    Remove now useless nested function in debugedit

commit d50839b8b3c4793a5c98c36f6e487aa12b9be3ba
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Jan 3 12:37:03 2008 +0200

    Lose unused pgpMpiHex()

commit 083fde4d7847e6d2fab3ef0978dc4458482c3633
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 2 17:47:30 2008 +0200

    Revert rpmfi bits from 01cd03ea8f8a
    - nasty breakage from new signedness mismatches..

commit 68e1d60be3e53c51d142dfca7b7a1501e867b681
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 2 17:02:06 2008 +0200

    Signedness consistency in rpmfi interface vs internal structures

commit bb70bbb7734a2a069bf97b08a354e04fb3debfdf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 2 16:18:55 2008 +0200

    Avoid compressFilelist() stack overflow in pathological cases
    - allocate dirNames etc arrays on heap, not stack
    - the arrays can be rather large and alloca isn't very friendly when it
      fails...

commit 505e8c352cdcd389b014e2276df30fb06e5984cd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 2 15:02:18 2008 +0200

    Teach (unused) headerFreeTag about RPM_FORCEFREE_TYPE

commit a9490bf57a52818fd67437d8389804641d8f875a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 2 14:44:58 2008 +0200

    Gah, nss doesn't use size_t for lengths...

commit 84386510030999daa5416eb17816bec9f2aa384b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 2 14:37:16 2008 +0200

    Assorted int -> size_t string size fixes in build code

commit 2a27a8fcdb33c69ead5df3701239c8d97c8c6e70
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 2 14:10:25 2008 +0200

    More assorted int -> size_t uses

commit 519eaaf236866602e1f05cf138ebb263b465f2b7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 2 13:55:16 2008 +0200

    Assorted int -> size_t corrections for string lengths

commit d772111b9131a9fc7ad4da1008ba6fb202344a2f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Jan 2 13:22:26 2008 +0200

    Use size_t, not int for header format padding args

commit addac6a86be1a6c7a668f97c9d7a0d85277f6779
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Dec 19 12:05:56 2007 +0200

    Add rpm_data_t (and _constdata_t) for header data, use everywhere
    - consistent, easy to grep for and change...
    - bogus consts removed where spotted

commit 44c90ce3f0c5f4686338b4b746b9bece8510d757
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Dec 19 14:13:34 2007 +0200

    Const-pedantry here and there...

commit 5bd20cf9db51053ff8d72a306e6296c8c744856f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Dec 19 13:19:55 2007 +0200

    Const pedantry

commit 8e00018c88f1467eb1c1e2760a4de96bcdbae68a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Dec 19 12:49:52 2007 +0200

    Fix some const issues in macros

commit 123beef21660b446c124acb7a9ed769c95844eed
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 31 11:56:06 2007 +0200

    tgi doesnt need fts private include

commit 85b0dd95318af5d252f4e1064df2eacdddd627fe
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 22 14:26:56 2007 +0200

    Fix multi-line macro expansion in last line of spec (mdvbz#27417)
    
    Patch from Pascal Rigaux:
    
    1. creating functions restoreFirstChar(), copyNextLineFromOFI() and
       copyNextLineFinish() out of copyNextLine()
    2. creating function readLineFromOFI() out of readLine()
    3. inline copyNextLine
    4. if we can restoreFirstChar, then use it, otherwise read from file
       (ie don't read from file when we have a remaining multiline macro in
       spec->nextline)

commit d4676384db65b0480af9dbbd488d43bd37b2080c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Dec 19 18:53:14 2007 +0200

    Hum, NSS needs to be shut down too
    - some ~80 bytes are still leaked apparently from somewhere within nss/nspr

commit 19f943cc3673830a5b41daa8e295a94e5570af20
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Dec 19 09:38:32 2007 +0200

    Add dumb maintainer snapshot generator make rule

commit 4d2151d160c34e7bfca0bae0f44c5eab21e6d401
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Dec 19 09:12:59 2007 +0200

    Ordering fix (don't use RPMSENSE_PREREQ) from Pascal Rigaux

commit 20bfd34975b7c4af9cce95167a227e9f9fe58316
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Dec 19 09:01:39 2007 +0200

    Drop duplicate locale.h check

commit 23b25e10c65216b6cb271e9ba43ad7bf8d53192e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 18 21:33:40 2007 +0200

    Avoid exiting too early due to nested rpmdbCheckSignals()

commit 38bb0b6c9b355915a69db476974bea1503fce147
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 18 13:42:42 2007 +0200

    Update translations

commit fd53dcf94d1836e5ba30b390f5289c6e29e7ff8c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 18 13:41:29 2007 +0200

    Todo-update...

commit 4f9b7814fa510b5197d5e423b48ad8aa304ce133
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 18 12:53:57 2007 +0200

    Drop bogus const from rpmReadPackageManifest argvptr
    - it's malloced and needs freeing by the caller..

commit 22d169c102c531aa2eca7d232818bfe38ca73f45
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 18 12:53:16 2007 +0200

    Cast to silence gcc..

commit 6965fbd5c8c3d6228989fa89cc8e2c107965aba9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 18 12:29:46 2007 +0200

    One leftover reloc constfree..

commit 640b2e3e797f8b13c9fe3360b5c78300214e88b4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 18 12:08:13 2007 +0200

    Remove bunch of bogus cli-related consts...

commit a631616ce168756f10e58863b053d630dd06a9b4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 18 12:07:48 2007 +0200

    Use constfree on relocations for now
    - they're totally exposed all they way to rpmqv.c.. eww

commit bb61fa21e8d93f74a568aec7426e086330d276b9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 18 11:56:54 2007 +0200

    Remove bogus const from rpmInstallSource*() parameters
    - spec and cookie are malloced and need to be freed by caller
    - unconst various other rpmQVK arguments & friends, no api exists to
      free them so caller needs to handle anyway

commit bed447552a88665c566a4b2760f89db03812c12d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 18 10:48:10 2007 +0200

    Use constfree for rpmts suggestions
    - goes down to rpmal returning fnpykey which is declared as const..
      check this later

commit f88fe168dde009f542f20fc151a595885801364f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 18 10:46:39 2007 +0200

    Stick constfree to extension cache cleanup
    - deal with the header mess later on

commit 083a82f7f159f8763b9904b32f08bfc58d0c7e40
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 18 10:45:36 2007 +0200

    Drop bogus const from headerFreeTag() data pointer
    - no effect as it's hft is unused
    - mark headerFreeData(), the real problem case, with _constfree() for
      easy finding (inlined in header.h so can't use _constfree())

commit 75df2ea9ad1a44cc836e1c4f43a54ca679b03879
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 18 10:30:35 2007 +0200

    Drop bogus const from temp variable

commit 78ad426f040a85965ad33ab64f040012f1b2af39
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 18 10:17:18 2007 +0200

    Remove bogus const from rpmGlob argv return type
    - it's malloced and must be freed by caller

commit 8bdd462b50b7930c94cd7c59ac068934f60cffcf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 18 10:10:03 2007 +0200

    Make fsm failedFile non-const like it really is
    - callers need to free it so const is wrong
    - cast away the hardlink-hackery in fsm instead

commit 28723cae06d49e4ba2401199da47763b1fc28ef8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 18 09:54:46 2007 +0200

    Drop bogus const from temp variable

commit 2b61b2199fa14b8e48eecda9731ab7ae30f7ee37
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 18 09:49:03 2007 +0200

    Drop bogus const from zapRelation return
    - the retval is malloced from rpmdsNewDNEVR() and must be freed by the caller

commit addf724f1afa39d7ab0a781d012812cf02890e3d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 18 09:40:33 2007 +0200

    Drop bogus const from rpmts rootDir + curDir

commit a56a86a79247f18e79f1ecab90879c05e6df5f45
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 18 09:30:40 2007 +0200

    Remove bogus const from rpmdb match iterator contents

commit fbd12ed8018a146ca339f684348302d1f85c277d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 18 09:27:42 2007 +0200

    Use constfree for rpmdb_s contents for now
    - rpmdb_internal is included from several places over the tree

commit e4c1bb5ac21983de746f5446b3ed564a210d55d4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 18 09:19:22 2007 +0200

    Use constfree for rpmhash key + data for now
    - sometimes they point to real const data, sometimes not, eww

commit bf723ddc66d4eeef1e3d3d6d62381ef597df7d78
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 18 09:08:04 2007 +0200

    Use constfree for fsm path, opath for now
    - neither should really be const but they sometimes point to real
      const variables, sometimes not.. needs careful review

commit d0a289256745887dd77c39411aaa9fd8d0b896e0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 18 07:51:45 2007 +0200

    Drop const from fs mountpoint, contained within fs

commit 5c9ab2a393e1cb2e253015a46433c536fa11fad0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 17 21:45:56 2007 +0200

    Drop const from temporary variable

commit 132017fb8dc2f120885f16c84639527466e211ee
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 17 19:15:14 2007 +0200

    Use constfree for rpmte internals for now
    - the internals are exposed all over the tree...

commit cddbd0db316bfc0886609ac1a1e10fe6c10d11e9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 17 19:11:33 2007 +0200

    Use constfree for rpmfi internals for now
    - the internals are exposed all over the tree...

commit aee47fd11270a5d5b3db4925e84864a728ae9475
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 17 19:05:19 2007 +0200

    Remove const from rpmal dirInfo dirname
    - add comments wrt needle dirName - those are actual const pointers to
      within rpmfi, must not free

commit 04978c4eaec7c86be705177b53fadd80a459ba00
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 17 16:10:32 2007 +0200

    Use constfree for psm failedFile for now
    - failedFile is malloced by fsm, but removing the const from fsm
      gets a bit hairy as failed it's not always malloced
      (see changeset 4062:02b0c237b675) so we'd have const-disabling casts instead
      of freeing consts...

commit 3ea6e70658c4554f7c1ed2bb2ccf5cdaef67b12a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 17 15:32:25 2007 +0200

    Include toplevel Makefile from Makefile.maint
    - allow "make dist" in one step with .maint..

commit 7460a8db5d22c736e99ff95d123d9518432637d2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 17 15:30:01 2007 +0200

    Drop db3 from dist subdirs
    - db3/configure is covered by EXTRA_DIST already

commit cd760eaa96af8854a89fb61a7a859a45ee7654ea
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 17 15:07:41 2007 +0200

    Unconst psm pkgURL and rpmio_flags members
    - fully contained within psm

commit f1df7a59b86b7e40568885d0d51a9bd72453c964
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 17 14:58:02 2007 +0200

    Make psm opaque, add minimal methods to cover internal needs

commit bad888114da54dd7b7eaf36faa34f0dfd4139432
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 17 14:03:31 2007 +0200

    Drop bogus const from dnlFreeIterator() param

commit b9f2b0788a19f76e89d782a76bc3d784fe4307d9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 17 14:01:23 2007 +0200

    Drop bogus const from fsmFsPath() return type
    - path is malloced and caller needs to free
    - additionally use size_t, not int for string size

commit 898d2cc867aef7a8637a2afcded0d24baf6d110e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 17 13:50:11 2007 +0200

    transaction.c doesn't need fsm.h, rpmfi_internal does

commit 6966c5d8ace09789a22d294e4204e918d91c43d9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 17 13:39:14 2007 +0200

    Stick cpioMapFlags to fsm.h, they're only used by fsm.c
    - allows FSM* typedefs to move to where they belong

commit 8268570f48aab6bb2126af2754d1007c959f3da3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 17 13:20:05 2007 +0200

    Hide fsmIterator_s struct details, not needed outside fsm.c

commit c3353c2af418398fe92c0f4febca555ee035474e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 17 13:15:53 2007 +0200

    Stuff hardLink_s inside fsm.c, add pointer typedef to fsm.h
    - nothing outside fsm.c needs except for the declaration in fsm_s

commit 05a16276ba5c4337f7bd1e1d64957c3ce364c106
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 17 13:01:50 2007 +0200

    badDeps is encapsulated within depends, doesn't need const
    - remove ancient reference code

commit 834a79e6158d9950f37cf66922875f2a10af4d21
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 17 12:37:01 2007 +0200

    Use constfree for cleaning up pgp digest contents
    - the contents are exposed within rpm (although not public interfaces)
      but pgpCleanDig() should be used to free the struct
    - pgpDig and pgpDigParams need encapsulation + api...

commit bef37587d3a0eb9a59661bd02c5e8b46a6c24e7a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 17 11:38:12 2007 +0200

    Don't bother configuring db3 directory if using external db

commit 2b50ad6fa4d936ef555235cb032cf7ed9e14ac67
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 17 11:19:55 2007 +0200

    Make ChangeLog creation depend on .hg directory
    - .hg dir timestamp changes on commits, otherwise updating ChangeLog
      requires removing it

commit 68d4b87e8a71335f11ebab703effd972c36631a6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 17 10:32:16 2007 +0200

    Unbreak internal dep generator
    - more signed -> unsigned count type breakage

commit 4671a35bbcac3097259657a1425ece1d3fe22ba1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 17 10:05:37 2007 +0200

    Unbreak array query formats
    - signed -> unsigned count type change broke it..

commit acdbac335e7a8e287afa68a1127501029a5a3bec
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Dec 16 23:05:21 2007 +0200

    pgpReadPkts returns malloced memory, unconstify

commit 7be0a763d282475763c11d7b05d1479d7f0a617e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Dec 16 22:50:30 2007 +0200

    Lone module-private const .. remove

commit c41f61369ea55d0c998dfb733b2e74921bca63bf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Dec 16 21:28:37 2007 +0200

    Unconstify macro contents, fully contained within macro module

commit 506a8f5365c9cc044754b4365e75481cf26b1b5e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Dec 16 21:24:44 2007 +0200

    Bogus const in temporary variables

commit 4a72c2a3f6b3fd712e8e1733be6761c89ec4e25c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Dec 16 21:21:31 2007 +0200

    More const removals
    - both username cache and log records are fully contained within their
      modules...

commit 1f443958a9b0ce3b29db5faa77532a437c77e663
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Dec 16 21:20:05 2007 +0200

    More constfree for spec exposed stuff for now
    - openfileinfo seems easily made opaque, revisit later

commit 598d6efc865d3c1ca03d52cc1987966d581de288
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Dec 16 20:58:44 2007 +0200

    Much of spec internals is exposed, let them be const for now

commit baa659b24dc7098009a7351a4ee1ab921b4b05ee
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Dec 16 20:46:56 2007 +0200

    Remove couple of bogus const uses in temp variables

commit 35d4663a0dae9abfc6a533bb00d1ee100f477beb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Dec 16 20:39:46 2007 +0200

    Unconstify FileListRec and AttrRec contents
    - not exposed outside files.c

commit 4ab5affca9c78a8e7f3bfc4af738722db1b93bfa
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Dec 16 20:04:30 2007 +0200

    Unconstify rpmrc internal tables
    - not exposed outside rpmrc, "protection" buys nothing

commit 98fdd1c0617be4182bc23d655ac7778c6f26a24e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Dec 16 19:00:07 2007 +0200

    Fix default queryformat setting when _query_all_fmt not set
    - \n in the rpmExpand() causes the result to always be non-null...
    - also avoid unnecessary const free

commit 14284725e6ddca4a8cd8ce5b0a4ba1e1ee7cef73
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Dec 16 16:25:09 2007 +0200

    Remove const from rpmMkTempFile file name parameter
    - the temp name is malloced and needs to be freed by caller

commit 3ec48632b894c299c4c4455d4e12bbce0bd37f32
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Dec 16 15:58:40 2007 +0200

    Refactor rpmReSign() and associated file "management" a bit
    - remove unused + useless "rc" parameter from manageFile()
    - remove tempfile handling from manageFile(), create + open a tempfile
      explicitly in rpmReSign()
    - both copyFile() and manageFile() now require file names to be passed always
    - split file closing into separate closeFile() function

commit 3184b09c1bc87f495e45ea25664589ef805c433f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Dec 16 15:51:09 2007 +0200

    Stupid reversed-logic thinko..

commit ca3ac216585253ddccaadcbcc65119cc4babed46
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Dec 16 13:02:27 2007 +0200

    Remove const from bunch of error message return types
    - headerCheck(), rpmReadHeader() and rpmReadSignature() return malloced
      error messages the caller needs to free, shouldn't be const
    - adjust rpmdb hdrchk callbacks accordingly

commit 3c002a4c8b0fe71a4ae4685e0fa7f9ddad17fda5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Dec 16 12:20:45 2007 +0200

    IDT* structures are exposed, let them use constfree for now..

commit fcf5efe3482a79281d4d4f2ea5f4146c8715f0e8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 15 21:17:59 2007 +0200

    And yet more bogus const removals...

commit f013449361e92ea2a6ffaa1bf3e9d1939f18d8b4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 15 21:14:44 2007 +0200

    Couple of bogus const removals from rpmgi and rpmds structs
    - in both cases the data is already directly inaccessible from outside,
      no further protection necessary

commit aa61b715e6639d6900ca66facb62d5f099dd11d4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 15 18:33:28 2007 +0200

    Make bunch of build internal code take const parameters

commit e8ffcb26b5cb60d4ad6130b0010c40ffd8d83b81
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 15 17:15:39 2007 +0200

    Further bogus const removals...

commit 3b6a6c903092409256b81e840b9b0dd67be0f6f3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 15 17:08:18 2007 +0200

    Fix incorrect rpmCleanPath() retval comment

commit dc9c504ae6de92c84e272f4a6e4f68c350236e23
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 15 16:03:18 2007 +0200

    Another bunch of bogus const removals

commit bc93e6bc874a8a07e33f7e77a8f09bde5588eeaf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 15 15:59:07 2007 +0200

    Avoid mixing local malloc with const parameter pointer

commit 23777730c25e440a20b6a7b0ad7d3e348992e976
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 15 13:43:40 2007 +0200

    Use _constfree() for freeing urlinfo contents

commit 20f6d481aabb9dcd4f38e486c80677f5a0d23f67
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 15 13:41:58 2007 +0200

    Expose compiler warnings from freeing data declared as const
    - _free() is just a wrapper to free() which additionally returns NULL
    - add _constfree() for use in the cases where const is used to protect
      malloced "read-only" data in long-lived otherwise exposed structures etc

commit b4588a1202783dd9bbee594aa5aea022ccdd3fc9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 15 11:47:47 2007 +0200

    Remove bogus const from rpmProblemString() return type
    - its malloced so it needs to be freed
    - fix all users to actually free

commit 1a0a12cf159a05471be62016a318d0653ae30ff5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 15 11:03:46 2007 +0200

    Remove bogus const from rpmGenPath() temporary targets

commit b6a663ec0a421d822578d831153bc0fbc2d6f410
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 15 10:39:15 2007 +0200

    Remove bogus const from rpmGetPath() temporary targets

commit dce66192f2f01817b3027c8c61c770abbfffd131
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 15 09:39:32 2007 +0200

    Remove bogus const from rpmGenPath and rpmGetPath return type
    - both return malloced strings you're supposed to free

commit 55cb2ff12836f124d5f8114a3c0d4246ec8fa7a0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 14 19:52:11 2007 +0200

    Bogus const removal continued...

commit 61e35657a6e3ce33a44067e82bb382b8b3133c4a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 14 17:55:31 2007 +0200

    Another bunch of bogus const removals

commit f5a712db5be29cf130a545af0645432713f6fa26
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 14 15:50:17 2007 +0200

    Remove bogus consts..

commit 8d66d5fe4a9d7b7b015153edb8f204ac3b89c461
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 14 15:17:59 2007 +0200

    Bunch of const char* corrections
    - functions returning string constants as char * etc...

commit 7d817f06b14c221bf0fbf74610e4d85093e77fcc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 14 13:39:52 2007 +0200

    Bunch of silly ssize_t casts
    - cuts down the amount of spurious warnings from inlined code...

commit faa7e730f1ff0e2345d0413fa3bdf6d0fec2d8af
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 14 13:38:58 2007 +0200

    Oops, major + minor are not dev_t...

commit 0c8401abff5eef9d94ae9f124a7ccf639b12acbd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 14 12:54:35 2007 +0200

    Yet more type-pedantry...
    - use rpm_tag_t for signature tag types too
    - size_t, not rpm_count_t in signature generation
    - add couple of explicit casts to rpmpgp.h to avoid spurious noise from
      inlined functions

commit 8946859700432de2f8462b672a7d3d3790b09a76
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 14 11:38:20 2007 +0200

    More size_t type pedantry

commit 00eea2ffa02db546dac707f4f32595f915062a1d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 14 11:36:48 2007 +0200

    Oops, rpmpsi index needs to be signed type

commit 4c91a3b6ad0eca18c732117ca4adf09914b6a64c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 14 10:30:46 2007 +0200

    More misc size_t issues

commit 421da7c94a9a81a00de847de2ff7249c9909594a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 14 10:30:01 2007 +0200

    Use dev_t instead of ints..

commit 4714ced24c99460f3af63c140b73b2125eaafd31
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 14 09:35:06 2007 +0200

    Couple of more size_t uses + related format string fixes

commit c76674a257dbdc5d4bae11f443d319f6f5ad042c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 13 21:18:37 2007 +0200

    Big bunch of int -> size_t corrections for various signature items

commit 1e96ead46670088e6b867cc989eadbef092d5864
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 13 20:35:33 2007 +0200

    Use rpm_tagtype_t everywhere for rpm (header) tagtype type
    - typedef'ed as uint32_t, doesn't matter much but negative types dont exist
    - easy to grep, easy to change...
    - define RPM_FORCEFREE_TYPE instead of -1 "magic" for forcing
      headerFreeData (ugh), easier to grep for and change than "-1"

commit 7e56c6355bea552d89e8d5d0a317763a8cee4a4f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 13 19:32:37 2007 +0200

    Use rpm_tag_t everywhere for rpm (header) tag type
    - typedef'ed as int32_t for now, negative values used in some places for
      error cases
    - easy to grep, easy to change...
    - add RPMTAG_NOT_FOUND define, used in place of -1 "magic",

commit cbd7a684a4cd88bfa3383658790be3c37230707a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 13 18:16:39 2007 +0200

    Use rpm_count_t everywhere for header data count
    - typedef'ed as uint32_t as that's the key size limit imposed by BDB,
      relevant for RPM_BIN_TYPE
    - easy to change to whatever later on as it's now consistent everywhere
    - explicit casts where needed to avoid new warnings from signedness

commit c8f2927227471758abbdb36781d885b01798f0f8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Dec 13 09:25:10 2007 +0200

    Use rpm_count_t everywhere for header data count
    - typedef'ed as uint32_t as that's the key size limit imposed by BDB,
      relevant for RPM_BIN_TYPE
    - easy to change to whatever later on as it's now consistent everywhere

commit 50f608b42f55e1d2c89dc22a01c58b7621b620a5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Dec 12 14:33:56 2007 +0200

    Install the new fancy perldep extractor

commit 4fbc35614992a1e4a62a926aff1f72d8af10729d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 11 14:18:50 2007 +0200

    Register the new SCRIPT_ERROR callback value to python

commit 6d1cb6a9da1bf92bac9b630a12df7b9aaafaa2cc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 11 13:50:20 2007 +0200

    Silly thinko/typo causing source rpms to provide their nvr..

commit f407b391c0df60186c301a255448eae89404d05e
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Dec 11 09:12:12 2007 +0100

    test == is a bashism.

commit c68860f056ce6bb122b8fac469d07414e8caefdf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 11 08:51:54 2007 +0200

    OS X Leopard fixes from Giulio Eulisse

commit 6a779c394b9ed70f7d7b45716b207ebae170f732
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 10 16:06:22 2007 +0200

    Add python method for for retrieving (without setting) ts vsflags

commit 0c4fe4a0b092e2fb32d517f45513c1a44fb6ca55
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 8 16:12:10 2007 +0200

    Drop non-existing librpmmisc from rpm.pc

commit 9ce13e09ef42ea4efca3c624054b0b719f06a43b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 8 14:02:32 2007 +0200

    Switch to <rpm/foo.h> style for public headers
    - adjust include paths accordingly

commit 179ebc1745b51eb9821dd783c9df475ac539532d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 7 16:19:21 2007 +0200

    Add RPMCALLBACK_SCRIPT_ERROR callback type
    - TS callback has a better chance at doing something semi-intelligent
      on scriptlet failure than rpmlog callback
    - Hijack "amount" for script tag that failed, "total" for exit code
      which might be pretty much anything: waitpid child, actual exit status or
      lua error... not probably very useful in the callback but...
    - Notify before logging so users can catch the actual error message
      on next rpmlog callback.. or something

commit b67d678e2af33ef9b2b43b7692322fd2898df1aa
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 7 15:32:11 2007 +0200

    Scriptlet (name) handling tweaks
    - look up scriptlet names just once inside runScript()
    - map trigger types to their real names too

commit 80f76a471c1b91feb453dead5053bfcc60591ca1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 7 14:00:34 2007 +0200

    Add getter methods for rpmlogRec items (message + level)

commit 037ceff668aca5985870e4c6cf5250adbddc4dae
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 7 13:31:31 2007 +0200

    Rename RPMLOG_CONT to RPMLOG_DEFAULT to make purpose clearer (hopefully ;)

commit a5f65886f897223fbb476c6ddb8c9f4b7d81d6a3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 7 13:16:45 2007 +0200

    Logging enhancements
    - add parameters to rpmlogCallback: current log record pointer +
      optional user data
    - callback return flags to enable/disable default logging behavior and
      to perform exit() after cleaning up
    - add method for retrieving prefix string for a given message priority
    - move default logging behavior out of rpmlog() proper

commit ca9ff17a4d21ec6ba78a56a0f1ddae7e8e16682e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 7 11:43:06 2007 +0200

    Make rpmLogRec opaque

commit e98903dbb10d860f9eb2b5b315aab4b1fa9b4152
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 7 11:34:33 2007 +0200

    Fix misleading indentation

commit 8769e540f1d3c25ba05c5d24b69a8b71cfd8dc6b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 7 11:33:18 2007 +0200

    Don't do tilde expansion if HOME not set
    - If you dont have a home, glob() will go out of its way to give you one,
      even a wrong one.  This isn't good when daemons etc have purposefully
      unset HOME...

commit 2cce0653a15f6a0bdcb4de8b7a5133e810744d32
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 7 10:57:39 2007 +0200

    More rpmerr.h removal fallout
    - use rpmRC types where appropriate

commit d4622b51354ac454f67d30baf9d3b7276672d3b1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 7 10:43:53 2007 +0200

    Yet more rpmbuild fixups from rpmerr.h removal (partly from rpm5.org)
    - additionally use rpmRC return types instead of numbers where appropriate

commit 23bba68b56c590f2fdbb955c60e5dff0ceae5bb8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 7 09:53:53 2007 +0200

    rpm5.org: fixup spec parsing from rpmerr.h removal
    - renumber rpmParseState enums to leave room for error codes

commit 2a9c146f31f9695da5af489ff9d14540b10fcd13
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 7 09:28:39 2007 +0200

    rpm5.org: fixup various places in rpmbuild from rpmerr.h elimination

commit 79fed162acf572d7c88fd6b08fae3ea088cf5fd3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Dec 7 09:01:25 2007 +0200

    rpm5.org: use pid_t, not int in build doScript

commit 9f34c3244316ee239739229bd3d1c89053d08179
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Dec 5 15:56:18 2007 +0200

    Fix query return codes (rhbz#244236)
    - count + skip over errors in rpmgiNext() instead of stopping iteration
    - add rpmgiNumErrors() for retrieving errors, use it in query

commit c54001150fa2c1408759c81998f8a00e2a96c5c4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Dec 5 12:57:07 2007 +0200

    Only platform.in is to be included in dist, dont use wildcard

commit d4cc1de207c0cefa0331734aa412f0d65fdaf204
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Dec 5 11:56:33 2007 +0200

    Rename apidocs to hackingdocs, only build on request

commit 882f3905859e85086dc8e95f36600fe1f0bab541
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Dec 5 11:27:50 2007 +0200

    base64.h missing from librpmio sources

commit 8ddd81d0d4815cffa70a4cb0ba284bfec977fc7d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Dec 5 10:37:38 2007 +0200

    Update potfiles wrt recent changes, regenerate translations

commit 131cadaec4c78be4e462003fb49681c22f1dc392
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 4 11:28:16 2007 +0200

    Match filestage strings with the fact that we dont use "rpc" anymore

commit dbdbe8010cd944f026a5a4e5d071eb31d29d81c4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Dec 4 11:25:46 2007 +0200

    Eliminate debug junk from the API
    - add some helper macros to accomplish the same, easy to enable / disable
      as needed

commit 422fa0dbeab879d3510c879554c33e359e0ab7ab
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 3 18:47:40 2007 +0200

    Document rpmtsScoreGetEntry() parameter to please doxygen

commit 2db7896a3489f270a58adc06ae6854c1210f7338
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 3 18:44:48 2007 +0200

    Remove double documentation from header methods to make doxygen shut up

commit 6d37b9782dde11b25d22d5922b698b4ec8841eb0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 3 18:39:56 2007 +0200

    Fix couple of doxygen file paths, remove obsolete comment

commit ee9cd5176f308313fc0eae3c59119c454d58c9fa
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 3 16:58:16 2007 +0200

    Rename _rpmdbMatchIterator -> rpmdbMatchIterator_s
    - consistency with other similar constructs

commit a96f817d50953f0fd48bd2a7bb983c82256bbc78
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 3 10:47:25 2007 +0200

    Split RPMTAG_* and related functions out of rpmlib.h to rpmtag.h

commit 6bdbbefbecf06b75581b8dfbd4ea626c59ebe312
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 3 10:15:46 2007 +0200

    Split rpmte internals to separate header
    - avoid unnecessary include-order issues
    - stop define-games for internals access

commit 1929c68ea3e61007fb6388fd923e01a0e2dd8378
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 3 10:09:37 2007 +0200

    Duh, remember to add rpmfi_internal.h to sources..

commit 7b98c02efb519c6634ab27f0a774b01e99a7b40d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Dec 2 23:02:33 2007 +0200

    Move rpmprobFilterFlag enums to rpmps where they logically belong

commit 249113b17a378e55532a1bd4b01c6954d56e59e1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Dec 2 22:59:38 2007 +0200

    Move rpmsenseFlag enums to rpmds where they logically belong

commit b2a11e03510dde4e805c388b24dcb41e5d1278ab
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Dec 2 22:54:14 2007 +0200

    Move rpmtransFlags enums to rpmts.h where it logically belongs

commit 361a8e697b934789a8d5560ba42266c9a22a3a0f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Dec 2 22:50:55 2007 +0200

    Move file action enums to rpmfi where they logically belong

commit bb7e7862e5256874189a008a00cd3d420d8119c9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Dec 2 22:00:18 2007 +0200

    Move RPMFILE_* file attributes to rpmfi.h

commit 2c3ce0c8b04938b7b35483d27a4e87eda9afa3e7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 3 09:46:51 2007 +0200

    Split rpmfi internals to separate header

commit 9ae6424a40cc8982bf9e46a0bb50dee6b49df782
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Dec 2 12:19:24 2007 +0200

    Move file state defs to rpmfi where they logically belong

commit 432467a793333c88321fa1b1429673c29a92a25c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Dec 2 12:11:46 2007 +0200

    Move file type enums to rpmfi where it logically belongs

commit c36bef96bbe016a68007ef715f521a1bfb8f9de0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 3 16:35:33 2007 +0200

    Drop RPC from apidocs..

commit 9cbf0349b84fb19c6dddbe4f7a3246d4c949ad09
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Dec 3 16:33:18 2007 +0200

    Use regular POSIX functions instead of "RPC" clones
    - we only care about local files

commit fb90b8b4eb46c4008326e07b4866f4bf0875df83
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 1 21:06:00 2007 +0200

    More rpmlead exorcizing
    - remove unnecessary type parameter to writeRPM(), header knows what it is
    - remove unnecessary lead argument to readRPM(), it's unused anyway
    - drop lead from cpioSourceArchive struct

commit c4f116eec34eb283d37dca84b6a38c9fb38537e5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 1 21:01:51 2007 +0200

    Erm, stuff headerIsSource() implementation to hdrNVR with other cruft
    - avoid dragging rpmlib for RPMTAG_* into header.h

commit 5219448fa1c27988e40450063db5fd49139faace
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 1 19:29:57 2007 +0200

    Use headerIsSource() everywhere for determining package type

commit 935df9b81e698d60181f7ca97730e22ac6b8b34e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 1 19:25:10 2007 +0200

    Silly helper method for determining if a header is source or binary

commit 3ec04bd5add043955a94d735330bf374fbe8a269
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 1 18:41:14 2007 +0200

    Lose leftover davRead() related hack

commit 92698df7236d0fc6c1d9c683210ddc12984764b3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 1 18:31:09 2007 +0200

    Make rpmlead opaque, add methods to deal with it
    - rename methods to rpmLead*() for easy grepping
    - populate lead from header where necessary
    - add rpmLeadCheck() method for minimal compatibility checking to get
      rid of duplicate code in several places
    - conditionalize lead version on dirtokens

commit 89ecfb96d0ea48419ceecd0cd1527fd6a4d465d9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 1 12:47:30 2007 +0200

    Include sanity
    - don't pull in rpmio into rpmpgp.h
    - make rpmpgp.h standalone (needs rpmints and stdio)
    - rpmfileutil needs rpmio (makes sense, huh...)

commit 656732848a942fd41cdb75b52820676cb15a6077
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 1 12:32:30 2007 +0200

    Make rpmDoDigest() honor algorithm parameter

commit 50b21a8a07934344178018c080d0e27f739538d3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 1 10:42:22 2007 +0200

    Don't include idtx.h in rpmcli.h
    - nothing needs it there, include directly where needed instead

commit d7e484237a3dfeaa49b5366a9210a68492f43598
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 1 10:31:20 2007 +0200

    Move headerN(E)VR(A) declarations from rpmlib.h to header.h

commit 82a9632d4f71987cf106cbc04b2005015ba69988
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 1 10:29:44 2007 +0200

    Move hGet*() funcs from rpmts, rename to headerGet*()

commit 9169947ebd5d22bfb25e6dfa8cb0abea8390ea64
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 1 10:16:20 2007 +0200

    Add bunch of missing rpmts doxygen groupings

commit 2f7c913723baa8135201fa930678d9e3d4138f2c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 1 10:09:54 2007 +0200

    (no)dirtokens is just a build-time option now

commit fc53d307ad1a9c01ea622d53d93befdea472240b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 1 10:08:13 2007 +0200

    Add separate public header for callback declarations
    - rpmmessages is empty, lose it
    - adjust includes

commit 7fc7ac2161afd379ca8dda58de7eca98299e9ecf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sat Dec 1 09:38:33 2007 +0200

    Lose url callback mechanism
    - not used and cant support it with external helper anyhow

commit 89527785bbf5e6410cbb5303238fc4f45c8ab02e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 30 23:08:38 2007 +0200

    Move logging related macros rpmlog, adjust includes

commit 95953fe6101829efc8895193205b991c5faf7eb9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 30 22:13:15 2007 +0200

    rpmurl and argv doxygen groupings

commit e01888a40376f47ac76a3c7060716a16d601ac62
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 30 21:58:18 2007 +0200

    rpmpgp doxygen tagging

commit 976f0f891c69268e8ddf7c04de6efdca307ecf30
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 30 21:47:27 2007 +0200

    rpmrpc doxygen groupings..

commit fe086189b7bb836807dbc89ef707da678575f1f1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 29 10:38:58 2007 +0200

    Export FTS_* options through rpmgi.h as RPMGI_*
    - permit full rpmgi use without exporting all of fts

commit bcbc8126093dc200fbdcbc37451ac5b59d559de9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 29 10:26:53 2007 +0200

    Lose python fts
    - nothing uses it
    - hardly "rpm" functionality

commit 47b6892500671b87aa3cb189b4fe4162771cce74
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 29 10:09:43 2007 +0200

    Avoid building rpmdebug-py

commit 41fc048c9a1938dffae9660db1cc874125c878af
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 29 10:06:57 2007 +0200

    Lose the useless rpmrc python stuff, group macro handling to rpmmacro-py

commit 44e1c051fcecc8db6033ed8a4d21c0d5cd1c8702
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 28 17:51:39 2007 +0200

    Lose rpmgi_internal.h, no longer needed

commit 6d2214791328708254279b872f53db5aca1d7095
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 28 17:47:58 2007 +0200

    Add rpmgiGetFlags() method to avoid rpmgi_internal.h in query

commit 9e098f531234cfe776aa63b86e8afafb11865c30
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 29 09:42:56 2007 +0200

    Bunch of compiler warnings silenced
    - bunch of silly debug message casts to shut up whining
    - cpuid() arg signedness

commit 0728a2d5df68d2f11e44c2d7b05c49a036fdefda
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 28 10:21:26 2007 +0200

    Remove code duplication

commit 1eb8b83dcaafdf81514ead25a605e8a4a7cd4350
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 27 22:47:19 2007 +0200

    Revert the buggy dbiFindByLabel() epoch handling for now
    - ':' is a valid character in version (ick), need to account for that too

commit f82d0d88a956b15cdac424c87b86bcfd36de2b6e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 27 22:44:54 2007 +0200

    Back out the epoch-in-queryformat changes, too much breaks

commit ca4497b51bb0e91f31218dc3b76bb03451cc887f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 27 13:38:18 2007 +0200

    Make epoch present in generated filenames too

commit 9699454ebc6c69251cc30cc8667ed3756e162bda
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 27 13:27:15 2007 +0200

    Comment out hRET_s and HE_s for now...

commit d3fc506c2bc2b38b306162c9beefcc45c3a7063b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 27 13:00:47 2007 +0200

    Make default queryformat show epoch if present

commit 6cfe6e2f10d5c68a2ac037d5ae63be81abb9d09e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 27 12:42:12 2007 +0200

    Remove unneeded includes

commit 5c6f56af8f41a1ddeaf066d45a9559d143ac04d9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 27 11:58:17 2007 +0200

    One missed legacy include

commit b97705e0182aee1e05b2b546530f0d3daf0f1586
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 27 11:56:05 2007 +0200

    Drop extra include paths from python too

commit 5fa08529cb805bb36a8b9a1d7300cade13d05e4b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 27 11:52:55 2007 +0200

    Move rpmfiBuildFNames() to rpmfi now that it's possible

commit fbf0cb4441816b70a6a8e5ecef37d54b152b9db0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 27 11:34:23 2007 +0200

    Move legacy%#¤%#¤ to librpm now that it can move..

commit 30bb2ba4c8e481dd1c18e6ce8cd6c563fb42e23b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 27 11:24:45 2007 +0200

    Drop support for old style file tags in rpmdb

commit eacfabfe70595aa54e23b3c0b0803b4ccc97c805
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 27 10:13:16 2007 +0200

    Remove long since dead fields from python header objects

commit 2380dc146e7b8e08fd92f1f870a62536d62307bd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 27 09:55:40 2007 +0200

    Remove debug junk

commit 2a228d767cb903889c96507aadd919ef610b33ec
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 27 09:52:44 2007 +0200

    Teach dbiFindByLabel() to grok epochs

commit 333f10c56ecc693c272b5ba74b190549d76b8c12
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Nov 26 15:41:24 2007 +0100

    Adjust comment.

commit 8f82c7309eb62e42a2e53d56af591374668afa49
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Nov 26 15:38:20 2007 +0100

    s/build_os/host_os/

commit 0169e936799d0ccd6061ae949359faed5dd75673
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Nov 26 15:37:30 2007 +0100

    Remove #include "lib/signature.h" (unused)

commit f33505085a9893847a2eb299024db35556d9f44d
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Nov 26 15:18:02 2007 +0100

    Regenerate.

commit 3f4a332f955ffd5193b4a7f4507125a201b19e5f
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Nov 26 15:17:52 2007 +0100

    Fix quoting bug in preinstall.am generation.

commit 7ab750975b08680f1c003f9c66031464f72eded0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 26 14:48:29 2007 +0200

    One more pgpKeyID_t use

commit a798f725eb7f54bccde6fce47ad2b41eaeb65041
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 26 14:42:14 2007 +0200

    Define and use pgpTime_t for pgp creation time(s)

commit 399dca5c05440350c39008b9e682b95945bc35ff
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 26 14:19:38 2007 +0200

    Define and use pgpKeyID_t for pgp key id's

commit 7ac39db1f4117442b49e4717c79f04295886d9be
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 26 13:51:44 2007 +0200

    Fix "byte" usage in test programs too

commit 95c55a1af9fa4be56737eefa0e709ce5abe47deb
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Nov 26 10:53:47 2007 +0100

    Type pedantery.

commit 96e7f513cde29bbcec1d21f62d23b74cd1b26ed7
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Nov 26 10:42:39 2007 +0100

    Eliminate type "byte"

commit ffd534e4b8893c13b4368b7c80bea873ffdadc67
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Nov 25 05:43:52 2007 +0100

    Remove unnecessary deps on NSS.

commit 0fc2acbe28328d1a999993ff1da2c7694df70f50
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 15:58:17 2007 +0200

    Add a bunch of missing groupings to rpmlib.h items

commit 266cea562f99d7ec1b96eae1f829111088b3a163
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 15:57:51 2007 +0200

    Add a few missing rpmcli doxygen groupings

commit ba1915ecb9a68033e68d9e78666049d9c8ed05e8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 15:47:34 2007 +0200

    Doxygen cosmetics..

commit 3053a2d8d013070d3ff6a2c0c44591ba84bbe8fc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 15:44:32 2007 +0200

    Add rpmtag + tag api doxygen group to librpm + hacking docs

commit 59ac81ac386601cee4f3b68e97d05f4c5b0befa5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 15:26:58 2007 +0200

    Remove non-public stuff from librpm doxygen groups

commit e6b0dfd1555cf423c649ddd295f22357bbfbebaa
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 15:20:46 2007 +0200

    Add doxygen group for rpmstring, include in librpm + hacking docs

commit 31ad46a64b50caae0f023393c6182b8ded1fba25
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 15:11:21 2007 +0200

    Add doxygen group for rpmfileutil, include in librpm + hacking docs

commit 1a0968f31f18c627982b857fe717b54d691f0c78
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 14:57:18 2007 +0200

    Add doxygen group for rpmfc, include in librpm + hacking docs

commit 43d1baef0f8e95e9340802faebc7ffffe3421b2d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 14:51:30 2007 +0200

    doxygen groupings for rpmsw, add to librpm + hacking dox

commit bdce039f8567890296ef80cc80574dd9947ab571
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 14:48:13 2007 +0200

    Bunch of rpmio doxygen groupings

commit bd01557f273fec707a5b626838defc848be856e5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 14:45:16 2007 +0200

    Add doxygen group for rpmsq, include in librpm + hacking docs

commit 7bba4e7cb11244e75b95526f3c0619836ed55584
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 14:43:01 2007 +0200

    Add doxygen group for rpmlog, include in librpm + hacking docs

commit a1cab56e703bb1dcab70384631dd081070194db4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 14:39:51 2007 +0200

    Add doxygen group for macros, include in librpm + hacking docs

commit 975aa86c3709c61dd452b3c2930500b839f55f17
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 14:33:34 2007 +0200

    Add rpmps to doxygen modules page

commit 2d396ce001ff48c1f644bcd8e2bb6d5a3cc4d297
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 14:32:45 2007 +0200

    Add doxygen grouping to rpmps methods

commit 31cbde7d3cf9156aefb65bc642eb2fa195606f09
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 14:19:15 2007 +0200

    Add doxygen grouping to rpmgi methods

commit b2a9283a18519397bee5ce7ac51680169ef46da0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 14:17:33 2007 +0200

    Add doxygen grouping to rpmfi methods

commit 9938bb48f5f4b1202d25bee842e402d4c257717d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 14:15:20 2007 +0200

    Add doxygen grouping to rpmds methods

commit 73bcaf56359cf5d106288f2a4e066730858adde8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 14:12:43 2007 +0200

    Add doxygen grouping to rpmte methods

commit c78e4767c217c20ac2b9895cad4aa918160179cf
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Nov 23 11:48:18 2007 +0100

    Expand private include file names to be relative to $(top_srcdir)

commit 38560be92b1a8b5782d5637cede4a74ddec35dc6
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Nov 23 11:47:48 2007 +0100

    Expand private include file names to be relative to $(top_srcdir)

commit 71aabbde880f6c0895e4a45c2650839bd4973a71
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Nov 23 11:44:04 2007 +0100

    Reflect changes to include files.

commit f103adbb269ff7b69b2bedca0e7ec6476ab42d82
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Nov 23 11:43:51 2007 +0100

    Expand private include file names to be relative to $(top_srcdir)

commit b1c0f900aa6879b6f337bd26b2fe425741f3548c
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Nov 23 11:41:38 2007 +0100

    Reflect changes to include files.

commit 8ef0ca213fa5e668508f6cbb601b2c2a73e89997
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Nov 23 11:41:29 2007 +0100

    Expand private include file names to be relative to $(top_srcdir)

commit 79599bafb42c673cc7b9be9a5d3589b5891df0d5
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Nov 23 11:39:40 2007 +0100

    Reflect changes to include files.

commit 60939d6985e007d55a81b12f98173b5bbe4d31e8
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Nov 23 11:39:29 2007 +0100

    Expand private include file names to be relative to $(top_srcdir)

commit e8aa138e987e5989d6b22696ebd7d37e2d76c63a
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Nov 23 11:38:17 2007 +0100

    Reflect changes to include files.

commit 4d49dfcf414c1d886f5dac9b746cdc80bd41c192
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Nov 23 11:37:54 2007 +0100

    Expand private include file names to be relative to $(top_srcdir)

commit dae44a0f38913fc34f55bb345c84d7cc3f6be722
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 12:00:22 2007 +0200

    Export FD statistics enums + fdOp() method

commit c10e5da8a0a724cb3cc9ed23124d83f44d384ac8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 10:32:10 2007 +0200

    librpmio needs elf include+lib now, librpmdb doesn't

commit 7b2a9b3c24ccb445a332ba57f42d526a32567644
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 10:27:29 2007 +0200

    Move rpmGlob() from macros to rpmfileutil.h
    - macros is all just macros now
    - internal copy of popt stuff needs to go...

commit 2c6905dd9f10589e80eb9292992510a1e070cfe4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 10:20:19 2007 +0200

    Move rpmGetPath and rpmGenPath to rpmfileutil.h

commit 178e32a0baf9759035e6f4b633b0396623eba80b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 10:10:19 2007 +0200

    Move rpmCleanPath from macros to rpmfileutil.h

commit ff77bfa2a1951473edab1db8e3bd29be5d69443b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 10:05:49 2007 +0200

    Move file compression stuff from rpmmacro.h to rpmfileutil.h

commit c1bcd633010cce22919dbc5c347fa401971d39a8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 09:58:02 2007 +0200

    Move rpmioMkpath from rpmio.h to rpmfileutil.h

commit 2068f4611220bbe90c98b65c58dd1b6c62858152
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 09:13:18 2007 +0200

    Remove unnecessary paths from python includepath
    - no private includes from lib and build are needed
    - document rpmdb and rpmio private needs

commit eb72616113a82daf76fecc8ed1da6a0d16f65cb5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 09:06:45 2007 +0200

    Regenerate preinstall.am

commit d10acc7c2a280bd8e0aea3610d9a8767521a5f40
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 22 16:28:30 2007 +0200

    Move makeTempFile() from misc.h to rpmfileutil.h
    - probably not very useful outside rpm but used all over the tree,
    - rename to rpmMkTempFile() for namespacing since we're exporting it now

commit d8cd2b6b1139a81b819e66817e22d814fe25a9e7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 08:45:59 2007 +0200

    librpmdb doesn't need rpmio private headers anymore

commit e063854e118b92af3072e888e0887b2943187895
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 22 16:06:11 2007 +0200

    Calculating file checksums aint legacy activity
    - move the checksumming into librpmio
    - rename domd5() to rpmDoDigest() with future expansion for specifying
      digest algorithm type (ignored for now)
    - rearrange parameters to leave return types last
    - one less rpmio_internal outside rpmio/, whee...
    - used all over the rpm tree, and it knows about handling prelinked files
      so potentially useful for outside users, might as well export it
    - new public header rpmfilutil.h for collecting this sort of stuff

commit b32cca34808e46bb5ce6972c99d95b3e53f555ab
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 22 14:48:07 2007 +0200

    Move pgpDig typedefs to a more obvious place..

commit 4ad615daf5a650359d4be4d7b317d0678040cd37
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 08:32:08 2007 +0200

    Move the xstr*() string variants from rpmio.h to rpmstring.h

commit 115d698252464f907a4932b4bfeb63d72a3098a2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 08:21:23 2007 +0200

    Move string helpers from lib/misc.h to rpmio/rpmstring.h
    - splitString, freeSplitString, stripTrailingChar

commit d6990c3ab1095e0637a5113fe062b9eb6e1a2238
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 23 08:11:42 2007 +0200

    Move stringbuf into librpmio, rename header
    - rename header to more generic rpmstring.h for collecting misc string
      helper functions

commit abeea80a38682069fe97e69c0c8ca71475497ba0
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Nov 23 06:46:19 2007 +0100

    Use #include <x.h> syntax to include public headers.

commit 7fb2899b84ec829cb72a059e7ae32b23a7c80669
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Nov 23 06:28:09 2007 +0100

    Fix typo in preinstall.am generation.

commit fb370262070657ea317309885b84defe73b520da
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Nov 22 18:13:53 2007 +0100

    Remove AM_CPPFLAGS += -I$(top_srcdir)/build.
    Remove AM_CPPFLAGS += -I$(top_srcdir)/lib.
    Let librpmdb only depend on librpmio.

commit 32f29c4cbf44fb16e8a915ecff30289fb4065e21
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Nov 22 17:22:22 2007 +0100

    Regenerate.

commit 575a851fb79d693a96d36e6091051ab34a4f781a
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Nov 22 17:19:38 2007 +0100

    Fix preinstall.am handling for VPATH builds.

commit d9c9a980699dfe792f6f4de947ccdc600ce52886
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Nov 22 16:45:25 2007 +0100

    Add -I$(top_builddir)/include/rpm to AM_CPPFLAGS.

commit 94d2343931fba7efbffdbd598625fe48b525de58
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Nov 22 16:17:20 2007 +0100

    Add -I$(top_builddir)/include/rpm to AM_CPPFLAGS.

commit 583140460100ea99553d883174065ca22a3099b2
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Nov 22 15:54:42 2007 +0100

    Add header preinstallation.

commit 3553b86aff3e77ac57a3103d9e84f21ac43f5c14
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Nov 22 13:41:06 2007 +0100

    Include system.h

commit 5787b3f7c12fd7d3a67663a8b1c85789a968ccf3
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Nov 22 09:43:40 2007 +0100

    Eliminate egrep (Abandoned by POSIX many years ago).

commit d58b2666e3f80aac47130e4157ab0be16dc847dc
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Nov 21 16:51:57 2007 +0100

    Add AC_USE_SYSTEM_EXTENSIONS.
    Remove AC_AIX, AC_MINIX.
    Do not add -D_GNU_SOURCE to CFLAGS.

commit 4990a44984437320a0e39ce19e4f045235756b6b
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Nov 21 11:34:30 2007 +0100

    s/build_cpu/host_cpu/

commit 17ef8a3082a7c9e2c2defd9f516923afca67e027
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 19 16:32:15 2007 +0200

    Lose rpmerr.h for good

commit 43c5de66fb90cbc59dc23eb0774d66c66d34b37d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Nov 19 16:25:24 2007 +0200

    Replace all RPMERR_* etc uses with corresponding RPMLOG_* levels

commit db4e34966ed9d52a920ab1b135309e6e21f12473
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 16 17:37:54 2007 +0200

    Split digest-stuff out of rpmio_internal to separate header

commit 3587db8c4e5f72cf00d2722d06f9d7face2d072a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 16 17:18:55 2007 +0200

    Eliminate copy-paste fd digest stealing, stuff into rpmio_internal

commit 157759299d796bad84a58b5a50586883e086b1ec
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 16 14:16:56 2007 +0200

    Eliminate need for rpmio_internal.h in rpmdb.c
    - adds new pgpExtractPubkeyFingerprint() API function
    
    Ported from rpm5.org work of Jeff Johnson

commit f38eeadf415acff83a04e723db0f0b3575f8369e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 15 15:16:10 2007 +0200

    Debugedit needs crypto initialization too..

commit 3a168d6927ec8c647b06b3f437c1503305141eb3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 15 14:03:19 2007 +0200

    Remember to initialize crypto in various test programs..

commit 589b54af909b5ee8e98818291b955b0951b1c26c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 15 11:08:53 2007 +0200

    Initialize NSS early (rhbz#382091)
    - add new rpmInitCrypto() API for directly initializing any crypto
    - call it from rpmReadConfig() to ensure it's always initialized early on

commit 905ea76db4153b3e82eaac3c0291b4c7e4c597c5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 14 21:52:42 2007 +0200

    Fix base64 decoder related crash (rhbz#380911)
    
    The base64 decoder code incorrectly assumed that char is a signed type.
    Patch from Tomas Mraz

commit 9ee49db5a4a63f3ba400ca431083159ea4c275ac
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 13 13:11:54 2007 +0200

    Fix rpm -K segfaulting on corrupted header
    (http://qa.mandriva.com/show_bug.cgi?id=#33735)
    
    Patch from Pascal Rigaux

commit b6a1ee9bb75b3f11bbab035137f37ccfd0db1cc3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 13 13:09:01 2007 +0200

    Make "rpmbuild -bb --quiet" quiet as should be
    (without this patch, the option is simply ignored in rpmcliAllPoptTable)
    
    Also rhbz#124300
    
    Patch from Pascal Rigaux

commit b3591654458361becb2db7c36594469b89a3bf6a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 13 13:03:38 2007 +0200

    Fix build without O2
    
    HAVE_LOCALE_H is used by system.h, ensure it is defined properly
    (the issue only occurs when compiling without __OPTIMIZE__ (ie -O2)
     otherwise libintl.h do include locale.h)
    
    Patch from Pascal Rigaux

commit ec8761f231c345d7a6f07ad8bafeff2cc8c69e83
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 13 13:00:52 2007 +0200

    Align rpm -V output
    Patch from Pascal Rigaux

commit 142b5e5301940ebb3be8e31da5cf11fe64ef1f7b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 13 12:57:49 2007 +0200

    Use dgettext() instead of just gettext()
    Patch from Pascal Rigaux

commit 827399cda09b7271b735a6c502e9a58597117bd7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Nov 13 12:47:22 2007 +0200

    Fix free on invalid pointer after displaying "Unable to open temp file"
    Patch from Pascal Rigaux

commit 7d032c35059210b81ac8054a81ea1270837fadc3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 8 15:59:45 2007 +0200

    Minimal protection for rpmsqElem internals

commit 262f2d63b1789cd832db37037e7988306db524a6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 8 15:56:11 2007 +0200

    Remove unnecessary psm.h include

commit 751bba091b495f081a8966992b64c86023a82c30
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 8 14:17:17 2007 +0200

    Dead code removal

commit 49cd2645747b393cfacaa5ca2308e676f6e0879e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 8 14:09:31 2007 +0200

    Use _target_* macros for build restriction checking
    - avoid using deprecated rpmrc rpmGet*Info()
    - plug a small memory leak (remember to free the values too)
    - potential linux vs Linux as the os name implications but as isMemberInEntry()
      ignores case anyway, shouldn't matter

commit df84cabff6804650532763563f996127fef6bac3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 8 12:57:02 2007 +0200

    Disable -Wstrict-prototypes for now due to noise from NSPR headers

commit af0fe8255c46f3826f4a69b0006b02d62f640b88
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 8 12:54:05 2007 +0200

    Kludge around mktemp() uses to shut up build warnings

commit 5320dcba034996397a7e9aa5efe04671109ce8fc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 8 12:34:53 2007 +0200

    Remove dead & unused RPMTAG_ARCHIVESIZE addition code
    - rpm >= 4.1 uses RPMTAG_PAYLOADSIZE and translates RPMTAG_ARCHIVESIZE to
      it automatically

commit 77d3ce054980b0bc88207c4468dc11ca8620fc72
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 8 12:26:46 2007 +0200

    Kill off deprecated and unused rpmGetMachine() entirely

commit de7a247133fd618aa9438b70646cc30d1577d0f1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 8 12:22:14 2007 +0200

    Remove rpmSetMachine() from API

commit f81c916aeff8015221de568998649d1555d890c2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 8 12:16:48 2007 +0200

    Remove dead rpmrc-related code from spec parsing

commit e6f7c3c84846e1aa4f772694f7bbf29ce6a2c5ed
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 8 10:33:47 2007 +0200

    Remove useless "var" parameter from setup helpers
    Additionally rename setVarDefault to addMacroDefault which it really is

commit cc465dd6233990630a064b6a1634be61c9f845a7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 8 10:32:10 2007 +0200

    Kill off unused rpmSetVar() and freeRpmVar()

commit f7604b1a0eb6ec1bc436a2c137ae4fbc4cb1e165
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 8 10:24:58 2007 +0200

    Remove rpmGetVar, rpmSetVar from API

commit 2967cb5ffa616566e7daf0487b80c335e03b2250
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 7 13:47:42 2007 +0200

    Eliminate dead code

commit 1688cc62bf51ccb611e4c3403c24d93e72fb270f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 7 13:21:06 2007 +0200

    Eliminate rpmlead use from legacy retrofitting.

commit b497a4f05fe04e86a120851830f578211fd5ddc1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 7 13:12:20 2007 +0200

    Only do legacy retrofitting on packages that need it.
    
    Packages not having RPMTAG_DIRNAMES (ie compressed filenames) is
    used for checking, packages not having it are either built with rpm <= 3.x
    or with --nodirtokens, retrofitting in the latter case doesn't really
    hurt that much for the obscure case...

commit 8f97ae9d265bcdc542055320b1a2bd346eb63475
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 7 11:57:12 2007 +0200

    Document rpmio_internal.h uses outside rpmio/

commit 4471d9c371400141896fc9d838408b086af232ca
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 7 11:41:06 2007 +0200

    Move tufdio under rpmio/ where it belongs..

commit c5588096d37f6747ab6da075a2cd8e5e8d909430
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 7 11:02:41 2007 +0200

    Silence bogus gcc warning

commit 85a1226e8ca7324e070977fd2c32f897c1418931
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 7 11:01:40 2007 +0200

    Remove unused variable

commit d8741ebc13c4981b0df788c7a481396b79213a01
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 7 11:00:51 2007 +0200

    Silence bunch of compiler warnings from test programs

commit 8239d0b74b6363fa38576d40a3ab4531a047f028
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 7 10:17:10 2007 +0200

    Remove bunch of unnecessary misc.h includes, document others

commit d8babb5a863011d4a384c8f77ca4aefa07805bbb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Nov 7 09:01:38 2007 +0200

    Remove .splintrc's in the tree

commit 51f5b6caa8c432d04cb1d33d525495f2d6820b6e
Author: Peter Jones <pjones@redhat.com>
Date:   Fri Nov 2 09:23:05 2007 -0400

    - Fix segfault in %{lua:...}'s rpm_print
    - Don't automatically print a newline in rpm_print

commit 43b1952dc66f264cadb02cc09abce5d7b0b8a8ba
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 2 12:11:33 2007 +0200

    Eliminate rpmfcToken struct internals from the API

commit 0781adb1cb4fbe63b83303527ce1aa4dc35bb319
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 2 10:40:45 2007 +0200

    Eliminate macro struct internals from the API

commit cefa324b3af8512b9d993ae45f8fbf56eceaf45b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 2 10:24:42 2007 +0200

    Meh, actually add the base64 files..

commit 1a32f9959ea4daf8e3b2ae1e099a73f871900ff9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 2 10:18:23 2007 +0200

    Remove tinv

commit 1a2a82d38972e7f24f014ea9533c162d7de98ebc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 2 10:16:16 2007 +0200

    Document NSS requirement, revise obsolete comment in macros.in

commit 524262f6f83cc98407597b35924aff7eba3a754c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 2 10:11:54 2007 +0200

    Resurrect tax from NSS switchover

commit 8afe81c21445f96e4f5430b05304242c958a9ad0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Nov 2 10:02:40 2007 +0200

    Use NSS instead of beecrypt for encryption (Tomas Mraz)

commit e7fd0807d3b0a024634bc62a3bf2571a1bd6098d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 1 12:26:05 2007 +0200

    Include "legacy.h" not <rpmdb/legacy.h>

commit 02fb3b030d61c187a9563cd3003c5609559e2310
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Nov 1 12:25:41 2007 +0200

    Remove two unnecessary legacy.h includes

commit e1cc5871a0cd351276035d501a0e8b617b0716b7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 31 12:48:43 2007 +0200

    Document RPMFI_INTERNAL uses

commit bb976c2f57310cb293df642ab4b8c36e2c022c44
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 31 12:39:23 2007 +0200

    Document _RPMTE_INTERNAL needs

commit 05fcdb49f1d366b65a72b1a22f299fa5be41f4e3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 31 12:08:01 2007 +0200

    Split out rpmts_internal.h from rpmts.h

commit a3bbfc98eefa6cf272ed8f391c8efc6af995bd53
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 31 11:57:04 2007 +0200

    Disable python available package mechanism for now at least.
    - need saner available mechanism to resurrect, OTOH nothing appears to
      use it anymore
    - last remaining RPMTS_INTERNAL offender...

commit c13ab3401614249d8321b9e920a479602dc8291d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 31 10:37:18 2007 +0200

    Add rpmtsPrintSuggests() method for now, use instead of ts internals mucking
    - temporary measure to get rid of silly RPMTS_INTERNAL uses, the whole
      solve/suggestion thingie needs love or ripping...
    - clean up the suggestions in rpmtsClean() instead of while printing

commit 68dcdd5dad79481bd5561ba6b0556bdb84b66c15
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 31 10:15:05 2007 +0200

    Kill ts->goal, instead set NOSUGGEST flag on erase
    - it aint quite the same thing but will suffice for now...

commit a100133e561488ac02c0286aa34c36fcae5ac5af
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Oct 30 15:51:53 2007 +0100

    Replace AC_CHECK_FUNCS by AC_REPLACE_FUNCS for basename, getcwd, getwd, putenv, realpath, setenv, stpcpy, stpncpy, strcspn, strdup, strtol, strtoul, strspn, strstr.

commit 3f5d1e332e56b7011c47df48c1c5a333b2f35912
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 30 16:48:38 2007 +0200

    Add rpmtsCleanProblems() method for rpmgi needs

commit e898dde44d34ce7e937acbefb530287e718d36bf
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Oct 30 15:24:10 2007 +0100

    Remove memcmp.c.

commit bb19662ebdd821047b7c3c61356198c0040e6175
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 30 15:53:44 2007 +0200

    Erm... rpmdsFlags are int32_t type, not char*

commit 41cd82aa7c6a0b9115d776a387b3177664075d44
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 30 15:50:35 2007 +0200

    Split transaction score stuff to separate private header

commit 6abf077d7096e2f48bf1316e37ee6651c78b536e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 30 15:15:19 2007 +0200

    No callback, no repackaging info - tough...

commit 51c91957ae7378547604f1e008fabab28145d7b8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 30 14:31:42 2007 +0200

    Use rpmtsRootDir() instead of ts->rootDir + url guess-games in psm

commit 41d0a9fd3e8615efbb333746dfd2ea1ad9e285e3
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Oct 30 12:16:21 2007 +0100

    Don't build stubs.c.
    Link against ../misc/libmisc.la.

commit d57d7105502758114110f0c2f7877c129274a776
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Oct 30 12:15:34 2007 +0100

    Build libmisc.la.
    Use @LTLIBOBJS@.

commit fa1b02934ae53f2924398da78b69ab3b8e603f8f
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Oct 30 12:14:23 2007 +0100

    Add -I$(top_builddir)/rpmdb to AM_CPPFLAGS.

commit 591cd9852badfcbed43737deeb60ea8f4ae9c10e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 30 13:11:13 2007 +0200

    rpmds is now fully opaque, eliminate _RPMDS_INTERNAL games

commit 798323c39f3a6e67e145a95f46b13ec40b2f330b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 30 13:07:59 2007 +0200

    Whole lotta trouble to avoid rpmds internals access in rpmfc

commit 80ab4ab3f6f6853736f47ee7214a9fbd24f9510b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 30 12:25:15 2007 +0200

    Avoid rpmte internals access in rpmtsNotify()

commit 9482ab214144679f2736598e784f9fc59e3dbf4b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 30 12:12:54 2007 +0200

    Use rpmteKey() instead of te internals

commit be83036b2a9231b6ee496f7fea78124f1b53fea5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 30 11:54:22 2007 +0200

    Remove unnecessary rpmShowProgress duplicate prototype

commit ae0e274e3511f6fe821adbf3f9769a311c5b3c89
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Oct 30 04:06:24 2007 +0100

    Remove rpmdb/tsql.c.

commit 002b3ca2f933fe33ca41e76e29c264345c77db09
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Oct 30 03:27:07 2007 +0100

    Include <assert.h>.

commit 88b0c2bbb5b82efcbe557176077c5edbd7b1c5be
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Oct 30 02:57:14 2007 +0100

    Remove URLMAGIC, URLSANE.
    Remove include <assert.h>.

commit d2384456f71b47793d8f7fd67211f74a4b5b4269
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Oct 30 02:56:34 2007 +0100

    Add URLMAGIC, URLSANE.
    Remove URLDBG (unused).
    Include <assert.h>.

commit 7208dc82beb8e9668bd980d5104eb0a7aa11fdbc
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 29 13:37:50 2007 +0100

    Fix typo.

commit 69f402c32073126b16a994d7536a76cb537801c7
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 29 13:34:53 2007 +0100

    Split out rpmgi_internal.h from rpmgi.h

commit 9a628446ce33c31895c5f58cdfb5c288139be253
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 29 12:24:15 2007 +0100

    Add SKIPSPACE, SKIPNONSPACE.

commit f7a062a74c8d507acec1bfbdbcb2bcf1f4939c66
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 29 12:24:00 2007 +0100

    Add SKIPSPACE.

commit 7d75c393a5a1d5328d6a5376fec685a1b08f72e2
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 29 12:22:59 2007 +0100

    Remove SKIPSPACE, SKIPNONSPACE.

commit bdd3cd3ed6806466b4891db49098970fe335fb84
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 29 12:55:43 2007 +0200

    Macros don't need rpmio_internal

commit 6713c1e8078facbdc97a3e0073fa39a79dafa841
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 29 11:45:20 2007 +0100

    Don't include <ctype.h>.
    Include "rpmio.h".

commit 3d74c511ff497ce11ec22d0b1a5296eac5733f58
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 29 11:43:51 2007 +0100

    Move typedef into extern "C".

commit b8b986dd20a4ebbd8f2d297f3b02b424734abf71
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 29 11:40:51 2007 +0200

    Remove unnecessary includes

commit bf106534a4b56f4a54d4e5bdb9e3e8212a5fd266
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 29 11:39:40 2007 +0200

    Remove unused variable

commit 00c6f2540654037067d52cbc0932b64b121ce965
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 29 11:38:52 2007 +0200

    Hum, files.c needs rpmio_internal afterall for rpmioSlurp

commit ebd3578be8c68673c66bcc039d74815762fba06a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 29 11:32:51 2007 +0200

    Include rpmlib.h instead of lib/rpmlib.h

commit a82d639d7288cd088f539371b5e710311451c3cc
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 29 10:21:32 2007 +0100

    Use size_t for sizes.

commit 546867c2b5d2ccade4ade4f399eab8a2c180cdec
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 29 10:20:41 2007 +0100

    Move typedef into extern "C" {.

commit acb80574504f39011875012c5253edda1169c7ea
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 29 09:58:15 2007 +0100

    Don't include <assert.h>.

commit b818f44da21cbe7660dafbd9938ca9e34834bf32
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 29 09:55:53 2007 +0100

    Don't include <assert.h>.

commit 3b9fd01a9aa03dc8d6dc1b9b6cbed6c2b4104628
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 29 10:32:38 2007 +0200

    Remove couple of useless rpmio_internal includes

commit 35bcc3dc3061bf0c15a9ce16ff84b35e7f7076f6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 29 10:32:20 2007 +0200

    Include rpmpgp.h instead of rpmio_internal.h

commit cebb292fd362475929fe35b74933140389c7eefb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 29 10:10:17 2007 +0200

    Avoid need for header_internal.h
    - headerCopyLoad() achieves the same thing..

commit 1fe38d71cc4049175b1a0ba7c493e20665880692
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 29 09:42:21 2007 +0200

    Use fdGetFILE() instead of rpmio internal fdGetFp() everywhere

commit 46e669f06e9a66f1833d19481c8f2d1adc041645
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 29 09:30:45 2007 +0200

    Remove bunch of long-dead macro debug stuff

commit 57a5a2b3db28faf7c7d54e955593caf88beaba74
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 29 08:22:15 2007 +0100

    Remove local def of uint32_t, uint16_t.
    Include <stdint.h>.
    Use size_t for sizes.

commit 3357b1fc76b4fbf118017fbb6f6e69207612690a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 29 09:20:59 2007 +0200

    Export rpmio fdGetFILE()

commit da6a613eb321917045d61cf9b59deada4bd67d6b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 29 09:11:37 2007 +0200

    Remove leftover comment

commit d255841988e7ef4bd02f83a9079d3a1c31d9f3a0
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 29 07:28:15 2007 +0100

    Remove AM_C_PROTOTYPES.

commit 7134487e2887a31cb0cdcb3711050e8bbbf17657
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 09:53:01 2007 +0100

    Eliminate BSD-types.

commit 6a10fc88147ad7f55dab51565abe216852f8bd5c
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 07:30:58 2007 +0100

    Eliminate u_short, USHRT_MAX.

commit a02ad3f220decc6a8e54fbf8e436f64ee0c07243
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 07:20:59 2007 +0100

    Remove unnecessary includes.

commit bb3c8db9a5428f1cab562ea205e6b1305822fe5d
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:57:08 2007 +0100

    Include "config.h" instead of <config.h>.

commit a241ba959f4b47a9631995cb3991e61ee70fb6a4
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:49:57 2007 +0100

    Include "rpmfc.h" instead of <rpmfc.h>.

commit 14d434925b894dff35f5a53e25ba651218a86634
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:49:54 2007 +0100

    Include "rpmbuild.h" instead of <rpmbuild.h>.

commit 64b30f9a12a18e30ca52512e41e977e009e8e14e
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:47:46 2007 +0100

    Include "rpmts.h" instead of <rpmts.h>.

commit a59c95205cc4ae3406f4f994fa615b11a1efb0e8
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:47:42 2007 +0100

    Include "rpmte.h" instead of <rpmte.h>.

commit f25c04c1e45c3c579fa182b5345409bde87a3da1
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:47:31 2007 +0100

    Include "rpmlib.h" instead of <rpmlib.h>.

commit 226e626dae6e2f55351b4616620ac616eaaea698
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:47:23 2007 +0100

    Include "rpmgi.h" instead of <rpmgi.h>.

commit 8f3334f5f73c80ee6ea9dd9d272ac757abeebd6d
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:47:20 2007 +0100

    Include "rpmfi.h" instead of <rpmfi.h>.

commit f15b940aa03ec7c1b4054991e1764bbf7ea1177c
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:47:17 2007 +0100

    Include "rpmds.h" instead of <rpmds.h>.

commit cff3ba0882e30dc6accaeba2b811c18afa78c203
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:47:14 2007 +0100

    Include "rpmcli.h" instead of <rpmcli.h>.

commit 9735f83012e8b5652284c62cf0c376e1e46a32fe
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:43:19 2007 +0100

    Include "rpmio_internal.h" instead of <rpmio_internal.h>.

commit 227633a3e0e54212568b6809a833f9e8c71461ba
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:43:16 2007 +0100

    Include "rpmio.h" instead of <rpmio.h>.

commit 2e1f554c2a2fc8910e8c17a43098c3e8b6ef4669
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:43:13 2007 +0100

    Include "rpmhook.h" instead of <rpmhook.h>.

commit d7c7203182670ed977ed7dd0b4644ae3d74e89b0
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:43:10 2007 +0100

    Include "rpmerr.h" instead of <rpmerr.h>.

commit 12ef6e8d3a139de1509a41c79c443815e5a8d35b
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:43:07 2007 +0100

    Include "fts.h" instead of <fts.h>.

commit f507baed1fb7d0129c21000b396a56e9bb776857
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:43:04 2007 +0100

    Include "argv.h" instead of <argv.h>.

commit 3e194cf5fe7aa53b6cec5d911e6eabfa6bc8b643
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:42:58 2007 +0100

    Include "rpmdb_internal.h" instead of <rpmdb_internal.h>.

commit 759e91bf4a2ce1770b17d9e176869f8998f45a3c
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:42:55 2007 +0100

    Include "rpmdb.h" instead of <rpmdb.h>.

commit 979907ae8bd2f9fd5878ed057d16837bd6f064dc
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:42:46 2007 +0100

    Include "header_internal.h" instead of <header_internal.h>.

commit 2c4e4c0ccbeac09dbfb6cbaa3ae1ef2c6e4a149d
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:42:44 2007 +0100

    Include "header.h" instead of <header.h>.

commit b1f8da9252c68b57389cf4ae990385060e8f23ac
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:36:11 2007 +0100

    Include "rpmlua.h" instead of <rpmlua.h>.

commit e36a2ac8da29f324bb93e2e99898ef0f0b9e1e16
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:32:05 2007 +0100

    Include "rpmmacro.h" instead of <rpmmacro.h>.

commit 942da0293731a75e3f8c40467f079b863d36ea30
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:17:49 2007 +0100

    Include "rpmmessages.h" instead of <rpmmessages.h>.

commit 454f5ddff8cc114c0ff0b0dc13d6570276ce715d
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:14:27 2007 +0100

    Include "rpmpgp.h" instead of <rpmpgp.h>.

commit 7075f640502b781b4428e412556231029550c087
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:11:57 2007 +0100

    Include "rpmsq.h" instead of <rpmsq.h>.

commit 562c47ce8b65a1eb1c56c1d246beac5fc24f5fa8
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:08:20 2007 +0100

    Include "rpmsw.h" instead of <rpmsw.h>.

commit 26a9e836085475ac57f22828864f846c5e03e1bd
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 06:05:12 2007 +0100

    Include "rpmurl.h" instead of <rpmurl.h>.

commit 2259b9e2d2685401cd19b27e6918ae60312f5faa
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 28 05:02:00 2007 +0100

    Add @cond NODOXYGEN to hide away __RPM_USES_STDINT_H__ from docs.

commit b85e0d00ca79dfa50677b05964708d2f2b5cd32d
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sat Oct 27 06:38:35 2007 +0200

    Abandon addlib().
    Stop messing around with /usr/ucblib.

commit bfb5e17e77d329b629281a6a909206131489b1a8
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sat Oct 27 06:28:06 2007 +0200

    Eliminate AC_TRY_COMPILE (Obsolete).

commit 519977112e83b1265bd277a5f3835671ab26921a
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sat Oct 27 06:19:26 2007 +0200

    Stop messing around with triple-arg'ed main.

commit 6765e53464a432a2931cbdd4bf72cff0e610223d
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sat Oct 27 06:13:03 2007 +0200

    Remove AC_PROG_GCC_TRADITIONAL (Obsolete).

commit d4988efe1b41bef82810dbfb1c7aa52ec68678a3
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sat Oct 27 06:10:13 2007 +0200

    Remove AC_ISC_POSIX (Obsolete).

commit 2d78c0a51a2b8f069b6e68ba1fc14e5bc76f9348
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sat Oct 27 05:50:32 2007 +0200

    Add deprecation notices.
    Depend upon C99-fixed-size types.

commit a468d3004d15e7b14268b64b5b168e04130f97d8
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sat Oct 27 05:48:53 2007 +0200

    Use AC_CONFIG_HEADERS instead of AM_CONFIG_HEADER.

commit 4275322229d82f8137627c26c5080fdc40aaa372
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sat Oct 27 04:13:07 2007 +0200

    Remove HAVE_VA_LIST_AS_ARRAY.

commit 6d13cf177dbf989d9201ff58315682535e071982
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sat Oct 27 04:12:19 2007 +0200

    Check for stdarg.h.
    Check for va_copy.

commit bebeb89735f4927d33233d2c49e91919187f1c91
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Oct 26 13:24:14 2007 +0200

    Eliminate [u|]int_[8|16|32]. Use c99 stdint.h types instead.

commit bbce1581029dddbdcdd76a0b969fd8a8fe5614bf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 26 12:08:34 2007 +0300

    Fix invalid ifdef-outed code breaking syntax highlighting :)

commit 7964312e1a52de4dfb10bea47cd6ff2a295f6b02
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 26 11:21:31 2007 +0300

    Remove dead rpmal code

commit eabf92f842039af35d1da0e0271e84483959a7e8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 26 10:21:12 2007 +0300

    Rename rpmcliImportPubkey() -> rpmtsImportPubkey()
    - it's not a cli function in any way
    - move to rpmts group

commit 4b185fc48d9e1206624002618ff661bc1770356c
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Oct 24 18:38:18 2007 +0200

    Add rpmdb/rpmints.h.in.

commit 7445f99b38d134ac5965ffab52ff55e611b2c28e
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Oct 24 18:36:19 2007 +0200

    Pass HEADERS to doxygen instead of pkginclude_HEADERS.

commit d46dd7352f673643396c060d49af055af7592a0a
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Oct 24 18:26:07 2007 +0200

    Minor fixes to generation of hacking

commit 462b0dbca01836d9a6f7112a014f0e16c54e7ad0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 24 15:59:28 2007 +0300

    Regenerate translations

commit 44b1137f099406028999c34064e1a82679c24b85
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 24 15:58:19 2007 +0300

    Fix a format specification error in the Catalan translation

commit c00f5828e552d87c73f3c780e3ee8c8bcc93418a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 24 15:56:48 2007 +0300

    Add Catalan translation (rhbz#350401)
    From Xavier Conde & others

commit b2e90be34639db2fd39b00305ae6e60aac8b6329
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 24 10:31:42 2007 +0300

    Use NEVRA, not NEVR everywhere for rpmProblems

commit 19330086994ef02b8dcc18e71087b5c34110801d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 24 09:48:02 2007 +0300

    Remove dead code

commit 62cc76e25cdfad78ac30bb28f626b474efdecddc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 24 09:47:22 2007 +0300

    Don't mess up problem pkgNEVR in python ts.check() (rhbz#349091)

commit dd8bbe42a3e357cccf6c0d3e7d63cfd76ae4bfa6
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Oct 24 03:11:19 2007 +0200

    Remove AC_CHECK_FUNC(fchmod).

commit b990c6737a92af63d537b4b071f44721401802a3
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Oct 24 03:10:57 2007 +0200

    Remove dead and unused, ifdef'ed HAVE_FCHMOD block of code.

commit 0034893b789590e49aebc09127ae90c6170cc148
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Oct 24 03:04:01 2007 +0200

    doScript: Change 'what' to rpmBuildFlags.

commit f5a2728aa1d31185949cf31cb5baa4de4da68804
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Oct 23 13:21:52 2007 +0200

    Pass -1 to setVarDefault for optflags.

commit 4c647570dcc4d56a84b8adf12bd0ad61dc4c974c
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Oct 23 13:20:19 2007 +0200

    Eliminate dead code in setPathDefault, setVarDefault.

commit 2f40e205403bf39549b3406836b0bde981096904
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 23 11:04:31 2007 +0300

    Remove unused python header stubs

commit 9832bc53308ed84b7cbcc4100ed13c345b0846ff
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 23 11:03:52 2007 +0300

    One more useless rpmio_internal include removal

commit fda6c48f004dace80d8a36c514f5d4bc04b9a74d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 23 09:50:23 2007 +0300

    Remove bunch of unnecessary rpmio_internal & related includes from python

commit 3b7a6a5a95e010ff1c9699342ce63520f35b4cb3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 22 16:51:41 2007 +0300

    Remove bunch of unnecessary rpmio_internal includes

commit c0e198bd624d3ae899194b18fec43cf065a484e9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 22 16:32:58 2007 +0300

    Remove ancient test-stuff

commit 2c499d40dc0598fc726923215697db9aefac35f8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 22 15:51:04 2007 +0300

    Fix rpmts.h internal includes instead of working around in rpminstall.c

commit 549fb4393ed4e95fa8fba16dfc1f00c8df64fbcf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 22 13:37:02 2007 +0300

    Use rpmtsGet/SetDBMode() to eliminate needs for _RPMTS_INTERNAL

commit 8d70f186814c8dce2362a803d15bf5fc3149888c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 22 13:33:53 2007 +0300

    Add set + get methods from ts dbmode

commit b96a582dd46c6fe3e24998ad7e4af1384397dc6c
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Oct 19 17:22:08 2007 +0200

    Reflect changes to doxygen docs.

commit 25ffe1cef94c9753297567407174402f98a6ddef
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Oct 19 17:21:36 2007 +0200

    Add Doxyheader.

commit ef6ea9b508a4309d069f25fe0871660cedae532e
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Oct 19 17:21:08 2007 +0200

    New (copied from Doxyheader).

commit 4a465aff3b1273f086b1b22639cb5d9a085ceb72
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Oct 19 17:20:09 2007 +0200

    New (renamed from Doxyfile.in).

commit 7c517216e789c9c9e7b6d319eb3746eb290a1c20
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Oct 19 17:19:20 2007 +0200

    Remove.

commit afaba3f831dfe1e6e082a105d8f69011688648a3
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Oct 19 17:19:05 2007 +0200

    Remove.

commit 8e59724a585d8049c6a4569197e836ccaf4d717f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 19 13:25:08 2007 +0300

    Rename rpmpsProblem() -> rpmpsGetProblem() to make purpose clearer

commit 92919c2f73518a14e45c7586b5d2ef5f278255d4
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 19 13:01:58 2007 +0300

    Make rpmProblem standalone "object"
    - problem set contains array of pointers to problems, not problems themselves
    - add functions for creating and destroying problems
    - use iterators for freeing the problems in rpmpsFree()
    - new rpmpsAppendProblem() API for appending existing problem item to
      problem set
    - implement rpmpsAppend using the new methods

commit 63f94fb4a8c3ac52941b60cdde1332fb3db67716
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 19 12:30:26 2007 +0300

    Eliminate rpmps and rpmProblem structure details from the API

commit 1e74fccef78dd4f5ad80d853646af678d7f6390b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 19 12:28:30 2007 +0300

    Use rpmProblem access methods instead of direct access to the struct

commit 494c56e94ae9bbdcec46b746dc6837de42c890b7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 19 12:27:41 2007 +0300

    Add methods for accessing rpmProblem data

commit dffede23507ece9740de2523171f021b6772fa4f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 19 12:15:35 2007 +0300

    Only permit appending to a problem set from python
    - remove subscript assignment (which mucks about rpmps internals way
      too deeply)
    - add append() method instead

commit 7f9678705457cb5c884fd1bc901ead9558a3a6eb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 18 16:09:05 2007 +0300

    Use iterators for rpmpsPrint()

commit b882a428bde280957844faebfc7bb90b0a847753
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 19 12:36:11 2007 +0300

    Implement python rpmps iteration with rpmlib level iterator

commit 36f1ac2c7b3a9030e3dfd59510fe285eac3dce88
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Oct 19 12:35:20 2007 +0300

    Convert rpmts-py to use rpmps iterator

commit b1c64e37fd170d08896d4cb9209f5241be138b09
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 18 13:46:59 2007 +0300

    Make verifyDependencies() use rpmps iterator interface
    - use stock problem string for problem descriptions instead of handcrafted
      messages, changes output format quite a bit but OTOH reports conflicts
      and requires separately and more clearly
    - should perhaps filter out duplicates, otoh the old code didn't either...
    - return number of problems encountered instead of 0 vs 1

commit 96ad2a1fee7528e182f3923440f7b3038e996c8d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 18 13:45:25 2007 +0300

    Add API for iterating over rpmps problems

commit 2240b4f8d67ab98de0f090ad146103f7a1984cfb
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Oct 19 07:04:15 2007 +0200

    Ship Doxyfile.in, doc/librpm.doxy.in.

commit 8059dfcdf000f234d93097cebb24ab0c8fa11e72
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Oct 19 07:03:23 2007 +0200

    Regenerate.

commit 39454472fc4ca6f776b44edc742f0a307a28ab53
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Oct 18 18:13:02 2007 +0200

    Reflect changes to apidocs generation.

commit 6f74ccbb75005643ee54216e449c8e2f7509270b
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Oct 18 18:12:29 2007 +0200

    Eliminate @WITH_APIDOCS_TARGET@.
    Add EXCLUDE.

commit b9d56b6dc93f301babdd3af113c8374636af25e9
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Oct 18 16:04:00 2007 +0200

    Remove WITH_APIDOCS, WITH_APIDOCS_TARGET.
    Add AM_CONDITIONALS DOXYGEN and APIDOCS. Don't generate Doxyfile.

commit 5c923b9b772a1c10021613f9a9cc80764edd9a79
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Oct 18 16:02:14 2007 +0200

    Add stub doxygen @param to silence doxygen.

commit 4d02494caa4e02a7f0c7f6b4160c9dc076969b5b
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Oct 18 09:51:32 2007 +0200

    Remove doc/librpm during maintainer-clean.

commit 6e3017d0406a5490a802f7ebac7abfb3f1afa019
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Oct 18 09:04:17 2007 +0200

    Use DOXYGEN instead of __DOXYGEN.
    Rework processing pkginclude_HEADERS for doc/librpm.doxy.

commit f110fe4b50376bfc530d47b33fe64479e4a56726
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Oct 18 08:41:31 2007 +0200

    Use DOXYGEN instead of __DOXYGEN.

commit 0174c534875868fe95c06dd58704a532157aace7
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Oct 18 08:36:00 2007 +0200

    Use DOXYGEN instead of __DOXYGEN.
    Rework processing pkginclude_HEADERS for doc/librpm.doxy.

commit 8bcd871811ccea8b704264ce273843e7b1c508d1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 18 09:34:54 2007 +0300

    debugedit mishandles empty DW_AT_comp_dir (Roland McGrath)
    - fpc seems to produce DW_AT_comp_dir with value "" (empty string).
      This is dubious, but debugedit makes it even worse by getting confused by it.

commit 0a7a4877245d5de1d74aec09245cb94b8181a642
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Oct 18 08:34:37 2007 +0200

    Use DOXYGEN instead of __DOXYGEN.

commit abdc47d51cbd910f47d8415cadc93a5b94fcecf6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 18 09:33:56 2007 +0300

    debugedit mishandles DW_FORM_ref_addr (Roland McGrath)

commit a6c8f8086e263352afc16a76800966d9c3696f59
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Oct 17 16:42:21 2007 +0200

    Add automake option nostdinc.

commit ea2968b6b3b2a79e40f986f148c77a1d15be3365
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Oct 17 16:41:59 2007 +0200

    Cleanup CPPFLAGS.

commit cc113a1aad048c06d71a64516f83741902c65a45
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Oct 17 15:38:33 2007 +0200

    Add find-provides, find-requires, db.h to DISTCLEANFILES.

commit 6e2f4d1351ce1f46c35acb5754f5c072744c59ff
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Oct 17 15:23:15 2007 +0200

    Move db.h handling to toplevel Makefile.am

commit 8a4f77043520785a93a08f3df077ded30ea7cc25
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Oct 17 15:08:01 2007 +0200

    Expand AM_CPPFLAGS.

commit ca1f17b9ff26b46ffbcbcdc66a90b2f3083fe315
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Oct 17 15:07:13 2007 +0200

    More reversions.

commit 83b24d5f005b9cbb8403ef1b016454cf81ccc63d
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Oct 17 14:36:47 2007 +0200

    Add preliminary support for doc/librpm.

commit 9e826374baad5b040d9bd70932be3908a8402f91
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Oct 17 14:35:54 2007 +0200

    #include <db.h>

commit 0f681e6625b13a75f5ea69bfa787f8e98f1c21d8
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Oct 17 14:31:06 2007 +0200

    Add tput.

commit 38fe2fd9731522bf4485c9734fcf645b8bf82329
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 17 14:55:41 2007 +0300

    Use HAVE_DB_H for both internal and external BDB
    - rpmdb only cares whether we have BDB or no, internal vs external
      doesn't matter
    - makes external BDB actually work

commit 44768215d390b5e4f792c2a0c2e69c4476d26f9c
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Oct 17 12:13:14 2007 +0200

    Revert yesterday's patch (we're not ready for it).

commit 0e43d3b5705d8cb33cb74817d2808aa9996c0241
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Oct 17 11:33:52 2007 +0200

    Set GENERATE_MAN=NO.

commit 0c1a781c29a83f977e93724c698fb1a5fb97210d
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Oct 17 11:30:59 2007 +0200

    New.

commit 2d0f3b069d3b22c7593379ec8c1b01eed79fe767
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Oct 17 08:41:53 2007 +0200

    Remove refs to emacs.

commit 2200b96fe318d92398c57c5ccced7c16e30397e3
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Oct 16 18:32:24 2007 +0200

    Move tdbi, tjfn, tsql to toplevel Makefile.am.

commit ba6a27ce9bd86f34d5fc1199943d89f484da4764
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Oct 16 18:22:23 2007 +0200

    Move ACLOCAL_AMFLAGS to the top.

commit 8ea05eb1dff52e93ea33433cc1efd274518ce167
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Oct 16 18:21:27 2007 +0200

    Use \$(SED) instead of hard-coded sed.

commit 1a2234fe5429388b60e217b1608a6d5028a7d208
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Oct 16 14:25:04 2007 +0200

    Fix broken doxygen comment.

commit a75b7cc5e912f1af4d63a90b483770059d60d61a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 16 13:40:45 2007 +0300

    Remove unimplemented, unused RPMPROB_BADPRETRANS
    - leftover from pretrans syscall hack before %pretrans script days

commit e4aeb6fac52b02068c15f3175a8818b11caf60d9
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Oct 16 12:28:09 2007 +0200

    s/dbiflags/flags/ in doxygen comment.

commit d1e24fa5867df035020eba8ce4ab59014b98a11a
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Oct 16 12:26:42 2007 +0200

    Don't include stdio.h.

commit 6f0dabbcdda98b95a30cc7073d2446b99ebd5840
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 16 11:51:57 2007 +0300

    Rename norwegian translation no -> nb (rhbz#332271)

commit b97951097805632470725ef162e6eb8dde250de2
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 15 10:40:52 2007 +0200

    Add usrsrcdir.
    Reflect using $(usrsrcdir)/packages instead of $(usrsrcdir)/$(RPMCANONICALVENDOR).

commit 77c2edcb76cea4aecb7cdc3b2dbed6a8ad470b22
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 15 10:03:16 2007 +0200

    Regenerate.

commit e044dc0d5a1adf366650878e66dad2e0d9f0d491
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Oct 14 12:49:18 2007 +0300

    match *.omf, not *omf in find-lang (Ville Skyttä)

commit c7b8053f463a6a6fdcc039908ff399f4c1f591d1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Oct 14 12:48:32 2007 +0300

    find-lang Qt translation support (Ville Skyttä)

commit 5656af9cf31c053ad97815692a0c0736c7e4bc42
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Oct 14 12:42:47 2007 +0300

    find-lang localized man page support (Ville Skyttä)

commit 52de2160183593ae515f08eeed9d62d0b05a37d2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Oct 14 12:31:24 2007 +0300

    find-lang POSIX-compliant find(1) usage (Ville Skyttä)

commit aeeee26cbc5fad78d0b7bb902eb01c75d0c353b2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Sun Oct 14 12:30:31 2007 +0300

    find-lang spelling fixes (Ville Skyttä)

commit ce13d1b3180ca338ae6d48bfb54c9c716089d329
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 11 14:14:00 2007 +0300

    Clarify rpm manpage wrt rpmbuild binary vs package (rhbz#233232)

commit 9d614eba54a83adc07113a8e9c75ab3ce55af2f1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 11 13:50:05 2007 +0300

    Clarify rpmtsCheck() documentation (rhbz#159638)

commit 37fd4820e587194375fbab9e8fd1ed642dc236db
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Oct 11 11:17:48 2007 +0300

    Make default _topdir vendor agnostic
    - follow Suse behavior by using /usr/src/packages instead of /usr/src/redhat
    - the default should probably be %(echo $HOME)/rpmbuild instead for
      sane non-root behavior, revisit this later...

commit 9a555bf696eed5601bbead738a8ef366114c1670
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 10 14:37:18 2007 +0300

    header_method.h is now private

commit fc0463782edce3f3ed416b57435eb996fb8bb174
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 10 14:36:13 2007 +0300

    Remove now unneeded hdrinline

commit 9204c957881d960c179c6e96a21a8ea56f6cf0a2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 10 14:35:14 2007 +0300

    Un-inline common header access functions
    - implementations in new header_common.c
    - definitions in main header.h

commit 8f9eea953f6e0d13cf622ce5ef5d958a55977282
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 10 14:25:29 2007 +0300

    Move per-header method typedefs etc out of main header.h
    - new header_method.h include

commit 67b68cbeb97831844dd06bc6e231a3d78f348bde
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 10 11:30:37 2007 +0300

    pkgconfig dependency script fixes
    - don't make assumptions about package names, requires pkg-config executable
      path instead
    - always create pkg-config dependency if *.pc files are included, not just
      in case the *.pc file requires other pkg-config modules

commit 5dd1a5baa571b994559579fb6ee6289f51257ee7
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Oct 10 10:08:26 2007 +0200

    Move pkginc_HEADERS to toplevel Makefile.am

commit ad40930a801e8b3e42bcdbbb33f759eb6f19e6f7
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Oct 10 09:54:39 2007 +0200

    Recursively pick up files.

commit 272d591973ce001d50c745f1796873e8ad1e8d77
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Oct 10 08:39:01 2007 +0200

    Add support for lzma compressed "Sources".

commit 9f7a8dab5928a60ab641e480f921251a93ffa0be
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Oct 10 08:37:35 2007 +0200

    Kick out polish comment.

commit fca4ed8d97cd98d1cb4a1d0e26339f3ff4171d3e
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Oct 10 07:14:47 2007 +0200

    Install libtooldeps.sh and pkgconfigdeps.sh (Used by lib/rpm/macros).

commit 1567d8baf485e7501c79d96f8231084dd0a8ecb7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 9 15:47:15 2007 +0300

    Permit src.rpm installation to be interrupted

commit 01d5c13a4299e6fb7357d14fee7c5a368ec3bbbd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 9 15:33:49 2007 +0300

    Make rpmerr.h private, include directly where needed.

commit 9fba2d43d162bc12e84a4e5120429609ff17fc72
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 9 15:07:08 2007 +0300

    Let tufdio use rpmio_internal.. its for internal testing only anyway

commit 8b4b6eeb47c00d072b159e9dc84cd2a2a8f72868
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 9 15:06:39 2007 +0300

    Oops, one missed rpmError() user..

commit 275835643bf84cace7a8e506f6d3f76fd618c73c
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 9 15:02:57 2007 +0300

    Remove rpmError() and rpmMessage() retrofitting

commit 6bdd2384c47637b0a054b98b2ce8aeec6d7930a7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 9 15:01:52 2007 +0300

    More RPMMESS -> RPMLOG verbosity fixes

commit 81c4b4ba27784047e78c83398f9ff925b26a632b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 9 14:58:13 2007 +0300

    Use RPMLOG_ levels instead of RPMMESS for verbosity setting

commit b21bf6018e915de2348e10d52c3576d1e4b5eae9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 9 14:54:28 2007 +0300

    Replace all uses of RPMMESS_FATALERROR with RPMLOG_CRIT

commit 832580670ac00f1de425960cf15261b784388077
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 9 14:53:09 2007 +0300

    Replace all uses of RPMMESS_VERBOSE with RPMLOG_INFO

commit 0eb7deabf2364e0c2671090a1f774c06a9217c1e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 9 14:52:01 2007 +0300

    Replace all uses of RPMMESS_WARNING with RPMLOG_WARNING

commit 56366e3890a913c43642132340e2b7c6292e1227
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 9 14:50:42 2007 +0300

    Replace all uses of RPMMESS_NORMAL with RPMLOG_NOTICE

commit c52905d33460c020f8f1b33da47c06a8a2cb513f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 9 14:49:02 2007 +0300

    Replace all uses of RPMMESS_ERROR with RPMLOG_ERR

commit 9d9aa852ca724e4c7671cec249189f7dfef5ed8e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 9 14:48:04 2007 +0300

    Replace all uses of RPMMESS_DEBUG with RPMLOG_DEBUG

commit 4ce9fc1a1725b883190de2c415842a342ed3dbe3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 9 14:14:30 2007 +0300

    Replace all rpmMessage() uses with rpmlog()

commit d709195c3a51522629443cddea042513f47c5ad3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 9 14:06:06 2007 +0300

    Replace all rpmError() uses with rpmlog()

commit 46921b2334ddde2cad8f3546c6f9b6124adbb3a2
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 9 13:41:59 2007 +0300

    Fix tufdio build after rpmio cleanups

commit 210e1ac82638d4ad9b802819ac615cbfed471813
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 9 13:33:39 2007 +0300

    Hide rpmio vectors from API, only internal needs

commit adb4ca9ecdca6b5e9542ea611fc3834b1dd530c6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 9 13:27:47 2007 +0300

    Remove timedRead() define

commit 72d2b41cb6aaca54b1d641fc1d890248ac7ba184
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 9 12:52:09 2007 +0300

    Remember to remove fdFgets() definition too..

commit 2ad02a60b9ea9ba95bd5c2d5c9b947a7c30b4034
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 9 12:48:04 2007 +0300

    Rename __fd*() to fd*() for consistency now that the names are free

commit 0e8db956e39a0d95da7be7c9286d77dd1d9eed82
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 9 12:42:27 2007 +0300

    Remove unused legacy rpmio fd*() interfaces
    - fdFileno(), fdOpen(), fdRead(), fdWrite(), fdClose() gone...

commit 351c1f35f79da9f8c0cccf7c569a5b6ba2f896f5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 9 12:28:59 2007 +0300

    Remove unused rpmio fdFdopen() function

commit c961e7a5f8f388a2e8608fecac575e2adecbd762
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 9 12:28:08 2007 +0300

    Remove unused rpmio fdFgets() function

commit cabd0b9ffd35e0a1f288f03f65a9b9360a6944cd
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Oct 9 11:55:00 2007 +0300

    Remove bunch of deprecated and unused rpmlog functions

commit 6ea5c0023f6995cfd7104f8657c4fa35e77367c4
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Oct 9 10:02:49 2007 +0200

    Use enums instead of ints for processing isCompressed ret-val.

commit 0f22ea8830574d936e38e84a3806af97c33e0530
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 8 17:12:10 2007 +0200

    Add __LZMA.

commit 58daebe1075f21e12fecdc1cf6ea878006cb8d6c
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 8 16:53:52 2007 +0200

    Make hdrinline.h a private header.

commit 1788472d6223618ec3a16dc8ad2ef61115eb745d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 8 15:40:31 2007 +0300

    Fix up the download path in README

commit 722e03c1a85bd9ae5df0ef9fb4b94650b90f6917
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 8 15:37:38 2007 +0300

    Dust off INSTALL docs a bit

commit 36217918fc131c4f63f18793c6eb471f46e97945
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 8 13:14:10 2007 +0200

    Make headers self-contained.

commit 847e431706d1d1a5e7cfde12b2a027b77870aadd
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 8 12:40:55 2007 +0200

    Make headers self-contained.

commit 619aa0e650a380f227a520549d367dd05dcc3df4
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 8 12:32:41 2007 +0200

    Make headers self-contained.

commit 069a92e2cd2c2bb213e204e4a5c120c8d827876a
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 8 12:29:02 2007 +0200

    Make headers self-contained.

commit 9e1a6a395f4785007e9e1e8935dfaae20f3afa8f
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 8 11:03:57 2007 +0200

    Make headers self-contained.

commit 98dc9ce7bb6c9371bd5713b64d838dd2a1b10083
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 8 10:35:28 2007 +0200

    Make header files self-contained.

commit d6ffa9891fcd1e5e2b66d3dc543165fb475716c0
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 8 10:05:06 2007 +0200

    'Adjust extern C {} blocks.'

commit c29845f5848cfdbab3e06dbcb27dccfc7b5a3ae0
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 8 09:55:55 2007 +0200

    'Include <magic.h> instead of magic.h'

commit b97dbc7fd3a499c6d9d83439a0bd33edfab4f43a
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 8 07:38:19 2007 +0200

    Use 'file' instead of '<file>' to work around include path issues.

commit 5b654b0086db80c884db4658a80621a21f5e7102
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 8 06:33:04 2007 +0200

    Do not include "system.h".

commit c68b8bcd8fc9f317f811e5e528d359b427874e21
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 8 06:32:24 2007 +0200

    Include "system.h"

commit b27dcbe5460c931ad174c6112d3266a425633a4f
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sun Oct 7 18:15:20 2007 +0200

    Remove generation of ChangeLog to Makefile.maint.

commit 49382a79ca7e0bb2912c7653f423a3069e654863
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sat Oct 6 08:54:44 2007 +0200

    const-ify second arg of strtokWithQuotes.

commit 144038e68277b58980453f1e2a5fe44a1b7e508d
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sat Oct 6 08:53:11 2007 +0200

    s,\\subsection,\\section,

commit 3df8f54ef5b95ca7b133404e8e003d65e900f81e
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sat Oct 6 05:56:57 2007 +0200

    Add description stub to strtokWithQuotes to silence doxygen.

commit fae3f7eb92436b838b720c0888df8927a49bf16b
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sat Oct 6 05:52:15 2007 +0200

    Rename static headerXXX functions into _headerXXX to avoid clashes with static headerXXX fintions in hdrinline.h.

commit acedc1b611ebb7fa287a0ea2eecb50273996e124
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sat Oct 6 04:45:37 2007 +0200

    Add quotes in doxygen blocks.

commit 4770cb41c2f49fe86cef24dbabe27772f253fac5
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sat Oct 6 04:43:37 2007 +0200

    Add quotes to \# in doxygen blocks.

commit 391ed7bbad93f0500debb0019dbb45bdd3d311ab
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sat Oct 6 04:42:47 2007 +0200

    Remove POPT.

commit 72e97bd3da9c926106cdf96e9a85dff96805cab1
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sat Oct 6 04:42:32 2007 +0200

    Add description-stubs to doxygen blocks to silence doxygen warnings.

commit 0d595e407a7588870f0e79da18eecb6600e7fc2a
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sat Oct 6 04:41:49 2007 +0200

    Add description-stubs to doxygen blocks to silence doxygen warnings.

commit 1cbede4fd20b87e5b97ee41b4750b52c4cf7b342
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sat Oct 6 04:41:21 2007 +0200

    Add description-stubs to doxygen blocks to silence doxygen warnings.

commit be342ad3c5966a851d98cc8d40b6436f8c8cab36
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sat Oct 6 04:39:58 2007 +0200

    Add quotes to \# in doxygen blocks.

commit f109c7b830edcf44b87553840a7e5b7cce83715a
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Oct 5 17:45:33 2007 +0200

    Remove stray splint comment.

commit d9a6863e11932fff42e02d79df899e45ab4f70f7
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Oct 5 10:40:58 2007 +0200

    Support DOT.

commit cac07870f502a21ff539e9a9d255b92e75687bd0
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Oct 5 10:40:07 2007 +0200

    Use size_t for sizes.

commit 5a909969fd24315832487a5386e074e88031afcb
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Oct 5 10:38:36 2007 +0200

    Fix doxygen comments.

commit 4f7db64d7c1a2c0bfbe07a40925e9b76b3e9851c
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Oct 5 10:37:55 2007 +0200

    Add HAVE_DOT.

commit d14cb3bd16755d9e2545290aa87440c51c4cd7b4
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Oct 5 09:36:12 2007 +0200

    Fix doxygen commands.

commit cc4754e7641578c9bd970615695f9d15200aa1cb
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Oct 5 08:13:33 2007 +0200

    Adjust broken \file.

commit a3cf72167f9a59a09b0b5e6ee9ba81b1173017ba
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Oct 5 08:03:37 2007 +0200

    Remove refs to tools/rpmxp.c, tools/rpmxml.c, tools/rpmtool.c.

commit 55ba0d5b84cf8c22df69e4d7b24addf1d423af4c
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Oct 5 08:02:53 2007 +0200

    Remove refs to lib/rpmsx.*

commit f7c95206fd6a051dd210af7a3a359df2e897f1f5
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Oct 5 07:48:56 2007 +0200

    Remove refs to "popt/".

commit 7d3b29301165ed57a85ce5a6e74d301328222f49
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Oct 5 07:48:20 2007 +0200

    Remove refs to "file/src".

commit 5d45e150766c4f6103451392d0fed5e9d0e31e21
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Oct 5 07:46:51 2007 +0200

    doxygen -u.

commit 32c98f766e3c4d98245679c94473a1cafed01f89
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Oct 5 07:21:20 2007 +0200

    Fix minor typo.

commit 4d5be35b6a907c2c2523075165fb7e3eb96b53cd
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Oct 5 07:11:33 2007 +0200

    Add __fdFileno, fdFileno.

commit dfc6fae6be0ed51c49d694887e1acda8662307be
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Oct 5 07:11:02 2007 +0200

    Remove fdFileno.

commit d4bea657c402a67a3ab2104c9ed1ae93314f3441
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Oct 5 07:10:36 2007 +0200

    Make fdFileno an external function.

commit 4fa0c255494a535780b98d7877657aa187080fde
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Oct 4 09:23:50 2007 +0200

    Change strntoul to using size_t for sizes.

commit 6d77a00e54d32d3543f71df7d46408cd553cb4f5
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Oct 4 09:18:16 2007 +0200

    Include rpmts.h instead of depends.c.

commit 9cf8f9c31ad81d45cd206bad5df98fc8b75a5333
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Oct 4 08:55:07 2007 +0200

    Stop checking for machine/types.h (Private header of some OSes not supposed to be used).

commit 70ed33417d1d122bfc434d106824a9035dea9463
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Oct 3 08:57:54 2007 +0300

    Small cronjob fixes (rhbz#315271)
    - create cronjob temp file in target directory to ensure corrent
      selinux context (rhbz#315271)
    - dont leave empty files around in case of failure

commit 249dad1475e08ba86f911f722a0421771ceb30ce
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Oct 3 05:07:12 2007 +0200

    Remove $dirs (unused).

commit 67e8f946fef1be13dbfdf6040a9604baa6e8dd4f
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 1 18:44:29 2007 +0200

    Fix minor typo.

commit 8c7b929a14cc7118b7eb046bc2b5bc184b7c2f06
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Oct 1 14:29:10 2007 +0300

    Never glob non-local paths in rpmGlob()

commit 343effc65f1b388abe4f3a68af10a27f34b4feb5
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 1 13:17:15 2007 +0200

    Remove (unused).

commit 6e4e9c751635cac3f79f1bf873cd922d1f10d82a
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Oct 1 12:58:32 2007 +0200

    Remove AC_CHECK_HEADERS(thread.h).
    Remove duplicate check for pthread.h.

commit ea2a8068cd8f974b10e8db56bdfd93b36dd199c7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 27 14:11:00 2007 +0300

    Add %alpha arch macro for supported alpha cpus (rhbz#295941)

commit 4bef9fd8b3826551f06911c99abb0dfe5061806b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 27 12:28:30 2007 +0300

    Own omf directories found by find-lang (rhbz#296731)

commit c8e421400c007442646af2979a8d622319c5d30d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 27 10:37:19 2007 +0300

    Fix disk space checking when zero blocks / inodes are available (rhbz#308171)
    
    When installing a rpm that requires space on a filesystem with 0 blocks free,
    rpm will not detect that there is insufficient space on the filesystem.
    
    Patch from Patrick Bergamin

commit 7c1a960aaea9fcdf5346844069eee9daae66d58b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 26 11:29:53 2007 +0300

    Remove the deprecated rpmlead definitions from public headers

commit 39072f5ee4e2f8d1314fc7d9e81b01c88a263bf6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 26 10:40:02 2007 +0300

    Avoid unnecessary rpmps internals access

commit 610306fb528b48ce2077b85f0fe90512ef79cc55
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 26 09:57:26 2007 +0300

    Rename whatis() to rpmfiWhatis() for namespacing

commit 0ddf824afb647b620400dba8172fa4ca7ee1690b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 26 09:52:55 2007 +0300

    Eliminate some unnecessary rpmte-internal accesses in psm

commit d16bdb1550c3c3c13a40ad9cb6869564a6abbc38
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 25 22:46:02 2007 +0300

    Hide rpmfc structure internals from the API

commit b813710a4ca26743291cbdbbb3831d5a3e103356
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sat Sep 22 07:58:56 2007 +0200

    Take out __FILE__, __LINE__ from DBGREFS.

commit cb076959008f3b50abac3747b1b1bad2be2c8e6b
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sat Sep 22 04:14:07 2007 +0200

    Add missing extern "C".

commit 07e383ce0d9a8beeec36e9aed66a2017e95dfcb6
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Sep 21 17:02:45 2007 +0200

    Add fdLink, fdFree, fdNew functions.

commit 0e3ce6c9a8f5d658ae5ef57e9aca1f3c0fc55b9f
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Sep 21 17:02:33 2007 +0200

    Remove fdLink, fdFree, fdNew defines.

commit 273ddaf0a95f01f68dd2d0df61ec1f466884d3d3
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Sep 21 16:51:15 2007 +0200

    Remove fdClose, fdOpen, fdRead, fdWrite defines.

commit 868ff87bc0ac8adbff6629e91111e79c49553c91
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Sep 21 16:50:59 2007 +0200

    Add fdClose, fdOpen, fdRead, fdWrite functions.

commit e152864a768c7cd81479b43e15f4f2cb340e0910
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Sep 21 16:22:23 2007 +0200

    Eliminate fdClose/__fdClose preprocessor games.

commit 14a28ee82db5ce0293519ed3a13627c560812aad
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Sep 21 16:20:52 2007 +0200

    Eliminate fdWrite/__fdWrite preprocessor games.

commit 9b929ba2941d1f4849f9ec92f8acb877adf788ea
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Sep 21 16:10:32 2007 +0200

    Eliminate fdRead/__fdRead preprocessor games.

commit 9ed70bbbc0315bb103c2c2fed6fa88a1d40dce3b
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Sep 21 16:08:39 2007 +0200

    Eliminate fdOpen/__fdOpen preprocessor games.

commit 9456c927a4fb4dca29fcbea84cb6727bf7b0bbb9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Sep 21 16:11:48 2007 +0300

    Hide rpmtag index details from the api
    - move rpmTagGet*() out of line from rpmlib.h to tagnames.c
    - move headerTagIndices_s from header.h into tagnames.c

commit 3a26a31576ea482c6c6e06b6cc685c2e57d47177
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Sep 21 15:58:50 2007 +0300

    Fixup after Spec -> rpmSpec

commit 2c2538d7260baf0eae36d5ee07ddbfffc20044d9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Sep 21 15:23:02 2007 +0300

    Rename Spec -> rpmSpec for namespacing

commit 0a16dadffc398617de3f4d0b6d45e4f017e19b9f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Sep 21 15:19:47 2007 +0300

    Oops, remove, not ifdef out..

commit 7e5099e52fe90a054a6b1bdf45265e7185233a1e
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Sep 21 15:17:08 2007 +0300

    Rename alKey & alNum to rpmal* for namespacing
    Also move it out of rpmlib.h, nothing needs it there...

commit cd5956633eeaf6b27f026726991a9a20d28ad7ee
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Sep 21 13:09:18 2007 +0300

    Silence compiler whining from bad popt argv definition

commit 1517b852b4ea8baca9d64f3d956f0fee5d0d93dc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Sep 21 13:02:55 2007 +0300

    Unexport FSM(I) typedefs, only usable within rpm

commit bb5d450d229146026ebd7789d67d6ac8448201e9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Sep 21 12:47:45 2007 +0300

    Unexport rpmpms typedef, nothing outside rpm can use it anyway

commit 91a2f856c8c428993d80e1fe24ae9656dc5596a1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Fri Sep 21 09:58:46 2007 +0300

    Remove unused proof-of-concept arch-filtering of dependencies

commit 97f6e790bbea6154fda27fcd733f1e2c8812f0c5
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Sep 21 05:59:26 2007 +0200

    Add another HAVE_ZLIB_H conditional.

commit 81745883980a8089655e0b3f8503e82c94847d23
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 20 17:37:39 2007 +0300

    Rip out unused anaconda-related hacks

commit c0149602a60e21d82b2564daee06a6b43b1f671a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 20 16:19:25 2007 +0300

    Rename public fileAction* type to rpmFileAction for namespacing

commit a5aa5ffce8ab5a5eb27f2f021b5affd60a2e518b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 20 16:12:46 2007 +0300

    TODO reminder about compat headers

commit e1b4af366ac5f71d1696615502013d028604755e
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Sep 20 14:57:04 2007 +0200

    Remove checks for synch.h, semaphore.h, aio.h (unused).

commit 83f09475869d7f86568b897719521be3a3b47365
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 20 15:52:03 2007 +0300

    Rename MacroContext and -Entry to rpmMacro* for namespacing

commit 48f4ff6583f09a12d539072e1337c40c92161ee3
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Sep 20 14:45:40 2007 +0200

    Make zlib conditional.

commit 154a7f7603dca24ee56d66f3ca6b2b2757f29cd9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 20 15:37:48 2007 +0300

    Move _free() into system.h for now
    - get rid of dozen _free() definitions
    - get _free() out of public headers

commit e53ceb6706125bead8630ba52f232c792ca1fdae
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 20 15:27:07 2007 +0300

    Rename fileTypes(_e) to rpmFileTypes for namespacing

commit a6e75c7595b64736d2a449c6d0191439c13abc85
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 20 15:22:59 2007 +0300

    Rename freeFilesystems() to rpmFreeFilesystems() for namespacing

commit 3b8da4ed118c6da6dab1cb7c0ecd0933c6d14f14
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 20 15:09:52 2007 +0300

    Rename tagName/Value/Type() to rpmTagGet*()
    - namespacing
    - avoid collisions with tagtype definitions

commit bbf7574d7f368e8d11600b54dfd24a0dafd555fb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 20 14:39:24 2007 +0300

    Match hardcoded _query_all_fmt default with current macro default

commit 55b7c431c44ee968b502d80c20af0f9a602b8225
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 20 14:33:35 2007 +0300

    Move rpmInstall/EraseInterfaceFlags to rpmcli.h

commit f65d670f023616f472f4451c08f5e3ee298a9ddc
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 20 13:23:58 2007 +0300

    Remove the "dressed" rpmHeaderGetEntry() from API
    - only used by the python bindings, move it there...

commit 4df06c5df7dadab2baa85d80522044da9dbe5733
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 20 12:19:34 2007 +0300

    Conditionalize rpmlib(PayloadIsBzip2) provide on HAVE_BZLIB_H

commit ae40a9001426e6f8fd17a0e00d1d63550dc7fca9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 20 12:18:22 2007 +0300

    Remove unnecessary include

commit 7cd1cf99412e76e10d072eae82bd608cc1e76c24
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Sep 20 10:37:48 2007 +0200

    Conditionalize zlib conditional pieces of code on HAVE_ZLIB_H.

commit 8fa3c5becf6f417c1a01fe878f5548d7d3bf2292
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Sep 20 07:36:41 2007 +0200

    Remove splint remnants.

commit 7871bb4547564297c26bd44a7a14759843d881c4
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Sep 20 06:53:59 2007 +0200

    Add dependency on libbz2.

commit 2d157c07f032b2ec97b26c919f1b182bd3123ec7
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Sep 20 06:53:19 2007 +0200

    Tighten check for libbz2.

commit f04137f591d7dae368c9a1280613aa941a66fbcf
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 19 17:38:33 2007 +0300

    Dead code removal
    All packages have provide their own name = ver-rel, checking deps against
    provides is sufficient, name doesn't need separate handling.

commit d76bef815fce0d29024b9356ec3b615d2caae2f5
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 19 17:32:29 2007 +0300

    Remove long since dead rpmrc provides code

commit 472167d7a89d32b576f434914e91ed6712708bae
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 19 16:51:32 2007 +0300

    Remove ancient RH-comment

commit a9d59cc2adb5acc9b70b103f9a228a1a94560c9d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 19 16:49:41 2007 +0300

    Remove --redhatprovides / --redhatrequires popt aliases
    - highly vendor specific
    - not even RH uses them anymore..

commit 19f8c80d56e2fc3c999e4437294faba7a3cefd38
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Sep 19 15:03:39 2007 +0200

    Adjust pktlen to pgpReadPkt's expectations.

commit 88b0e2c09334f32f869d86c5f8698668064f518a
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Sep 19 15:02:40 2007 +0200

    Cosmetics.

commit 4f222fef8fe09aacf8e87aaa36e50eb2becd94c8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 19 13:59:05 2007 +0300

    Remove db-internals from rpmts
    - api for getting db ops out of rpmdb
    - api for setting chrootdone of rpmdb
    - replace db internal accesses with api uses

commit fa1f65cef6f20ebe544306e8663dcb4a886168c9
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Sep 19 05:32:03 2007 +0200

    Rename --with-internal-db into --with-external-db.
    Add AC_ARG_WITH(external_db).

commit 211a30a615f4b3a85a70e30be401d0e60f61b94e
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Sep 19 05:15:57 2007 +0200

    Remove NPTL games.

commit 18e62fe8eb332b6febdca2296f7cff806b157e6e
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Sep 19 05:13:13 2007 +0200

    Move EXTRA_DIST out of PYTHON block.
    Move system.h into *_SOURCES.

commit 6503abadc9a50522a220d5ce340e3d523d356131
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Sep 18 19:18:09 2007 +0200

    Add check for db_threadid_t in db.h.

commit 618333608f427980507db0e3eeb534e89754a964
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Sep 18 19:04:34 2007 +0200

    Build db3isalive only if it is being used.

commit 5e1b759552716f6afb7819c7fd011d697f42f650
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Sep 18 14:59:22 2007 +0200

    Rename BZIP2BIN->__BZIP2.
    Rename GZIPBIN->__GZIP.
    Rename UNZIPBIN->__UNZIP.

commit eacf1841db81c1e64577f66e3619daf0c4baadf7
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Sep 18 14:55:58 2007 +0200

    Rename BZIP2BIN->__BZIP2.
    Rename GZIPBIN->__GZIP.
    Rename UNZIPBIN->__UNZIP.

commit 4f250b5c36233fc3e598fa1e419f983a08d58de2
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Sep 18 12:44:29 2007 +0200

    Remove WITH_POPT (Unused).

commit c89c1481eb6f2fd045f98e09010635698225abcc
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Sep 18 12:41:47 2007 +0200

    s/libary/library/

commit a0558abcf2b07c0b46a9ab52bef047dd30209b69
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Sep 18 10:49:20 2007 +0200

    Fix ancient typo.

commit b46e405e40d0dc110d670ba76aac63238fc656e0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Tue Sep 18 11:36:17 2007 +0300

    Use rpmquery instead of rpm for --querytags check to avoid popt aliasing issues

commit 2607d282e7cc90c2f0b1932cb3f940921ab13b5f
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Sep 18 08:09:43 2007 +0200

    Invert logic to setup WITH_INTERNAL_DB.

commit af4ba121d8457ac5282508ddec195ff3297fbeae
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Sep 18 07:38:50 2007 +0200

    Fix typo in previous patch.

commit 1cf3cb513951105c5a4259c788f54644327c66da
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Sep 18 06:40:16 2007 +0200

    Append @WITH_DB_LIB@ to Libs.private.

commit 01899614bdc055903e8c2a8cb07e2fdb94f9c9a0
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Sep 18 06:39:31 2007 +0200

    Reorganize *_DATA creation.

commit b9bd1303f6a5b3e3f6f6c8d2e5c0cf94b9625c2e
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Sep 18 06:19:57 2007 +0200

    Include fts.h.

commit 79abda764e23685c3e9b47643d3eafd3dc561aa3
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Sep 18 06:04:07 2007 +0200

    Remove.

commit 71f0cce9f9ffe70a849d9c0739c9632f8a90cd19
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Sep 18 06:03:33 2007 +0200

    Remove.

commit 90688278db31b05a56ca1e7c29cc71d38ce28ff3
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Sep 18 05:53:37 2007 +0200

    Change main() to using char *argv[].

commit f2a250702a4d4f2735cacffef033ad8b6e3062d1
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Sep 18 05:01:45 2007 +0200

    Add preliminary --with/without-internal-db processing.

commit de4fc157baf037c091e7ff5a35ccd3339013c870
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Sep 18 04:23:21 2007 +0200

    Conditionally build db3.
    Add DIST_SUBDIRS.

commit b5e01099e723f75f937605ccc928f1957301ba0d
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Sep 18 03:22:16 2007 +0200

    Regenerate.

commit 7bf56f58295b6ee19f3896b726f6a7e3435af446
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Sep 17 17:58:29 2007 +0200

    Eliminate WITH_DB_SUBDIR.

commit 6394b2a2e9314af0909e49323fe0012b549be29d
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Sep 17 17:53:47 2007 +0200

    Add db.h to nodist_librpmdb_la_SOURCES.
    Eliminate WITH_DB_SUBDIR.

commit 5bb02c33a92956f695d8f4e5bd8271e825d2a531
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 17 17:54:51 2007 +0300

    Include rpmts.h, not depends.c (!)

commit 006f6c710bf084ad1080cbaac94c217d307d1a20
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Sep 17 16:52:34 2007 +0200

    Don't install db.h.
    Move db.h to librpmdb_la_SOURCES.

commit a002ceebac6cd2642de6daf9d986efc179cb7f5a
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 17 17:20:11 2007 +0300

    Make fts.h a private header

commit ed55d4e033f62157275daafb1ec1a7b6a39d5dfe
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Sep 17 15:02:07 2007 +0200

    Remove db_emu.h.
    > Remove stray EXTRA_DIST.

commit 8e4f309c310aef69f128349a2d01749dc83c1f4d
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Sep 17 15:01:27 2007 +0200

    Remove (Unused).

commit 37e6caf2dc1023abcdfc781c2c72c4683e1512f3
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Sep 17 14:27:43 2007 +0200

    Add support for extern BDB.

commit 47257ec9e84106fb7132945aef614f6b4f54e1a6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 17 15:15:03 2007 +0300

    misc.h is a private header
    - include it from build/*.c where needed
    - dont install

commit 8a055b3cb36f7aa682fe7c80f7b32c530dcf03ef
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Sep 17 14:14:05 2007 +0200

    Reorganize.

commit adbab387377cdfba261229650cb6189fa5a257c6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 17 14:50:25 2007 +0300

    ugid.h is private header really

commit d7838b77735fefb17a626f5b7f46c000122992a3
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 17 14:43:32 2007 +0300

    More typos, doh...

commit 08a47f7472df4492f72b6a5a9d7b85b5e2beb868
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 17 14:42:18 2007 +0300

    Typo..

commit f35edd40321c2072ae5e4bcf81e6c63f1338e151
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 17 14:41:22 2007 +0300

    Use rpmtsGetRdb() instead of poking ts internals

commit f4296e723b14c6ed1831efddbf93204160dd80e7
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 17 13:45:39 2007 +0300

    rpmhash is an internal header really

commit 174c83b4881ee5cea8862a549df5588b87abdea9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 17 13:31:11 2007 +0300

    Mergesort is internal details..

commit b390d800193f87c246475cfe5c427a142ac10532
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 17 13:26:22 2007 +0300

    Make sqlite backend build again.

commit 66dcee4a67caea7c1d144ffab1717efe90613c73
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 17 13:17:07 2007 +0300

    Hide db index handling and rpmdb details from the API to internal header

commit ae3127c0f2cc74f5b662ec8414c231662f0df592
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Sep 17 11:33:18 2007 +0200

    Add with_internal_db.

commit 2975d9b63b650103f8994a5bd2e5ae00b6980314
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Sep 17 11:29:52 2007 +0200

    Apply poptReadDefaultConfig iff RPM_USES_POPTREADDEFAULTCONFIG is set.

commit 7484efff8e9a338e6b662c4867a5093969c75892
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Sep 17 11:28:27 2007 +0200

    Apply poptReadDefaultConfig iff RPM_USES_POPTREADDEFAULTCONFIG is set.

commit 53d8adadd34f94ff94121531ab8d720aac7aaa17
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 17 11:59:45 2007 +0300

    Oops, unbreak file opens..

commit b2517d99673d574ea75264b910571e9f7e6ce7c1
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 17 10:45:42 2007 +0300

    Remove unneeded socket/inet includes

commit cc68353128ea704bae581811583fe98969e1a0f8
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 17 10:31:18 2007 +0300

    Remove url caching, don't need it...

commit c07bda47cff8e854ca220d1d49d9ae28f83006a0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 17 09:48:08 2007 +0300

    Remove url/transport items from rpmio file handle structs and uses

commit 94e953382dc4b470dbdfe8a3e847924a10d6b5ac
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Mon Sep 17 09:32:25 2007 +0300

    Remove transport-related items from urlinfo and uses

commit be53bc654c0bec15da8f253ad3664285753807a4
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Sep 17 05:52:29 2007 +0200

    Remove (Dead, unused).

commit 68c30501ea89bdd7d015db85b9532c59e3818e8f
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Sep 17 05:51:34 2007 +0200

    Remove (Dead, unused).

commit 826e49f5d9a5095a53fc5216048983483b4be1e7
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Sep 17 04:32:23 2007 +0200

    Rename options into optionsTable for source-code consistency.

commit c72ad4dca58421486e1462b3b079dd55293dab6a
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Sep 17 04:31:28 2007 +0200

    Reformat call to poptGetContext for better source-code readability.

commit 0cde26ecde1e384d3c8c8c3ae3c7f764b52747e0
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Sep 17 04:11:59 2007 +0200

    Add if defined(ENABLE_NLS) guards around i18n-initialization.

commit 05088ee6db16aba98ca305eb034bb90be7348fca
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Mon Sep 17 04:11:39 2007 +0200

    Add if defined(ENABLE_NLS) guards around i18n-initialization.

commit fcbe2c86ef008ca58dbc0149f880b1e4a2eb7213
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Sat Sep 15 08:07:15 2007 +0200

    Remove unused var "selected"

commit af121321501f917e1f2ae8c57fd95a9628859be8
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Sep 14 13:14:43 2007 +0200

    Kick unused defines.

commit ff599b55162d39c78cabac070f0f4cb6aaaad3c5
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Sep 14 13:14:09 2007 +0200

    Fix typo in doxy-comment.

commit 91a33fbdc56bca5330fc232a9bbda485f244746e
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Sep 14 05:21:39 2007 +0200

    Move generating rpmpopt-@VERSION@ from configure into Makefile.

commit ede9e902729b0ef56e853c63efa0c63c8c2f3458
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Sep 14 05:09:17 2007 +0200

    Move generating rpm.pc from configure into Makefile.

commit 48fbe1b4c148bac646841de164a0d1c51f44dd0a
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Fri Sep 14 04:55:19 2007 +0200

    Remove @WITH_NEON_LIB@.

commit df1b72d3af290d3d15d0501f2a116f685aee3c0c
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Sep 13 14:51:46 2007 +0200

    Process platform in Makefile instead of configure.

commit a019b8bf9670eb1d5b374defa646a32b47a0e097
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Thu Sep 13 14:49:35 2007 +0200

    Move handing scripts/gendiff from scripts/Makefile.am to Makefile.am

commit 0138937d1d06dba083ed47382d8ca1223a9d825b
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 13 15:09:47 2007 +0300

    Remove support for BDB < 4.1

commit 305632b9574a6a4a56906cca4b588cc6d82cd5cb
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 13 14:50:45 2007 +0300

    Remove dead code

commit 6e14a001938d7245b1f288d183f51cefe18fdf9f
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 13 14:48:50 2007 +0300

    Stale lock removal on BDB >= 4.5.x

commit 26aa538d2600085680012538025ca6387964fde0
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Thu Sep 13 12:47:43 2007 +0300

    Remove db3 "annotate me" comments

commit 36ec63896eba02fbed5e6882b7f3f592456b800b
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Sep 12 14:54:48 2007 +0200

    For the moment, allow the testsuite to fail.

commit c98e35c9767d32685484def02ae21c9d5fbdf0c9
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Sep 12 14:54:04 2007 +0200

    Rebuild with updated spec.

commit 8a81b4a9af27747cf8687dd5d336f0446fc33f53
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Sep 12 14:52:40 2007 +0200

    Adjust BuildRoot.

commit d974ed0136e52ef4c4f7e1ed7fc312f8a2a2345d
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Sep 12 12:42:59 2007 +0200

    Add __CURL.

commit 0eb731f01b5e9dd9e5be42bac9d9c32ab255f1e2
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Sep 12 12:12:59 2007 +0200

    Update.

commit e1c33ad52fe25dc6b4a560ad84da1fd4accb9965
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Sep 12 12:06:38 2007 +0200

    Remove rpmio/rpmdav.c

commit 07c9cac938a6b5e8c41bf8afa92180c9ffd5b90c
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Sep 12 11:31:17 2007 +0200

    Fix indentation of --enable-python.

commit f2541835c5343582ca7b5e163c367e529a5ae944
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Sep 12 11:30:39 2007 +0200

    Adjust several *_LDADDs

commit 9f325855d74db499159773427a0965c93c822b72
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 12 12:23:55 2007 +0300

    Rip out internal http and ftp transports.

commit fb7433b3648a82a4ef8229f6254dac3d5b95cce0
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Sep 12 10:56:37 2007 +0200

    Merge tools/Makefile.am into Makefile.am

commit 65012c42534f9dca99655f56c26c5ba34e0f32a6
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 12 11:51:52 2007 +0300

    Use configurable helper application to deal with urls.
    - hijack ufdOpen() operation to call an external helper (if configured)
      to fetch any remote url's, return fd to local temporary file

commit d36716c25cde44c3aa9d96e75424b1ae46e6608d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 12 10:55:55 2007 +0300

    Remove NEON and related hacks

commit d2d4dae1add3e5e48810baf2524f8bd01cb6e02c
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Sep 12 09:34:31 2007 +0200

    Formating cleanup.

commit 56fc5ee3406b110bdf9e041b8df37afc7716562d
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 12 10:11:06 2007 +0300

    Remove now unused DAV RPC code.

commit 142cbc590662d3a90ab3ffc2a20cbc8f1e16dcf3
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Sep 12 09:10:06 2007 +0200

    Cleanup. Regenerate.

commit d70d21b878fc31a59c80ba2a3be4ed3fadda2ad9
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Sep 12 09:05:35 2007 +0200

    Remove splint annotation from getdate.c.

commit d7c9ba442581f3ee703928b018180c4b078b77a9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 12 09:45:11 2007 +0300

    Remove now unused FTP RPC code.

commit cf9d90aedbdb1325aaa030f9fa62180f5d5f64a9
Author: Panu Matilainen <pmatilai@redhat.com>
Date:   Wed Sep 12 09:38:04 2007 +0300

    Only deal with local files in "rpc"

commit cfe8875a0f9f15b8402eeaf71c54152d2766a075
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Sep 12 08:23:51 2007 +0200

    Reduce minimum python version to 2.3.

commit 2ff6e07f502f44f46e506616145d9d975e3d2075
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Sep 12 01:21:13 2007 +0200

    Reduce minimum python version to 2.4.

commit 37b193419c54124db9fb07d5b6efa5aa7938e914
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Wed Sep 12 01:18:29 2007 +0200

    locally define Py_ssize_t and lenfunc for python < 2.5

commit 65cbae7709bad0021182e2ba2b35c6c356e15ab5
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Sep 11 23:26:22 2007 +0200

    Remove splint tags.

commit 9f914c252453628999e72018354e95310f10f4a9
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Sep 11 23:22:45 2007 +0200

    Remove (Unused).

commit a1eb1f41d91572fc0ad4f8805d946c8bdbdb8d2a
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Sep 11 23:20:16 2007 +0200

    Remove splint tags.

commit 32419bbba54b1e3847d5890d990edf7012a7dada
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Sep 11 23:03:27 2007 +0200

    Remove splint tags.

commit e9f3a5c89df5b5509ac520d03d516b9f33b270fd
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue Sep 11 19:23:32 2007 +0200

    Remove splint tags.

commit ebef4f211c1f3b4392530eb1a5889aa13345893e
Author: Ralf Corsépius <corsepiu@fedoraproject.org>
Date:   Tue