{"thread":{"id":"19658","subject":"[ANNOUNCE] GIT 1.6.3.2","startedAt":"2009-06-04T07:20:17Z","lastAt":"2009-06-04T07:20:17Z","messageCount":1,"participants":["Junio C Hamano"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"115415","messageId":"7veiu0zcr2.fsf@alter.siamese.dyndns.org","threadId":"19658","inReplyTo":null,"subject":"[ANNOUNCE] GIT 1.6.3.2","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2009-06-04T07:20:17Z","receivedAt":"2009-06-04T07:20:17Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"The latest maintenance release GIT 1.6.3.2 is available at the\nusual places:\n\n  http://www.kernel.org/pub/software/scm/git/\n\n  git-1.6.3.2.tar.{gz,bz2}\t\t\t(source tarball)\n  git-htmldocs-1.6.3.2.tar.{gz,bz2}\t\t(preformatted docs)\n  git-manpages-1.6.3.2.tar.{gz,bz2}\t\t(preformatted docs)\n\nThe RPM binary packages for a few architectures are found in:\n\n  RPMS/$arch/git-*-1.6.3.2-1.fc9.$arch.rpm\t(RPM)\n\nGIT v1.6.3.2 Release Notes\n==========================\n\nFixes since v1.6.3.1\n--------------------\n\n * A few codepaths picked up the first few bytes from an sha1[] by\n   casting the (char *) pointer to (int *); GCC 4.4 did not like this,\n   and aborted compilation.\n\n * Some unlink(2) failures went undiagnosed.\n\n * The \"recursive\" merge strategy misbehaved when faced rename/delete\n   conflicts while coming up with an intermediate merge base.\n\n * The low-level merge algorithm did not handle a degenerate case of\n   merging a file with itself using itself as the common ancestor\n   gracefully.  It should produce the file itself, but instead\n   produced an empty result.\n\n * GIT_TRACE mechanism segfaulted when tracing a shell-quoted aliases.\n\n * OpenBSD also uses st_ctimspec in \"struct stat\", instead of \"st_ctim\".\n\n * With NO_CROSS_DIRECTORY_HARDLINKS, \"make install\" can be told not to\n   create hardlinks between $(gitexecdir)/git-$builtin_commands and\n   $(bindir)/git.\n\n * command completion code in bash did not reliably detect that we are\n   in a bare repository.\n\n * \"git add .\" in an empty directory complained that pathspec \".\" did not\n   match anything, which may be technically correct, but not useful.  We\n   silently make it a no-op now.\n\n * \"git add -p\" (and \"patch\" action in \"git add -i\") was broken when\n   the first hunk that adds a line at the top was split into two and\n   both halves are marked to be used.\n\n * \"git blame path\" misbehaved at the commit where path became file\n   from a directory with some files in it.\n\n * \"git for-each-ref\" had a segfaulting bug when dealing with a tag object\n   created by an ancient git.\n\n * \"git format-patch -k\" still added patch numbers if format.numbered\n   configuration was set.\n\n * \"git grep --color ''\" did not terminate.  The command also had\n   subtle bugs with its -w option.\n\n * http-push had a small use-after-free bug.\n\n * \"git push\" was converting OFS_DELTA pack representation into less\n   efficient REF_DELTA representation unconditionally upon transfer,\n   making the transferred data unnecessarily larger.\n\n * \"git remote show origin\" segfaulted when origin was still empty.\n\nMany other general usability updates around help text, diagnostic messages\nand documentation are included as well.\n\n----------------------------------------------------------------\n\nChanges since v1.6.3.1 are as follows:\n\nAlex Riesen (7):\n      Introduce an unlink(2) wrapper which gives warning if unlink failed\n      replace direct calls to unlink(2) with unlink_or_warn\n      print unlink(2) errno in copy_or_link_directory\n      Clarify kind of conflict in merge-one-file helper\n      ls-tree manpage: use \"unless\" instead of \"when ... is not\"\n      ls-tree manpage: output of ls-tree is compatible with update-index\n      http-push.c::remove_locks(): fix use after free\n\nCharles Bailey (2):\n      t6023: merge-file fails to output anything for a degenerate merge\n      Change xdl_merge to generate output even for null merges\n\nClemens Buchacher (1):\n      fix segfault showing an empty remote\n\nDan McGee (1):\n      Fix type-punning issues\n\nDaniel Cordero (1):\n      builtin-checkout: Don't tell user that HEAD has moved before it has\n\nDave Olszewski (1):\n      merge-recursive: never leave index unmerged while recursing\n\nGiuseppe Bilotta (1):\n      completion: use git rev-parse to detect bare repos\n\nJeff King (6):\n      fix GIT_TRACE segfault with shell-quoted aliases\n      add: don't complain when adding empty project root\n      for-each-ref: fix segfault in copy_email\n      lock_ref: inform callers of unavailable ref\n      fetch: report ref storage DF errors more accurately\n      fix cat-file usage message and documentation\n\nJim Meyering (2):\n      format-patch let -k override a config-specified format.numbered\n      pre-commit.sample: don't print incidental SHA1\n\nJohannes Schindelin (2):\n      Add NO_CROSS_DIRECTORY_HARDLINKS support to the Makefile\n      Cope better with a _lot_ of packs\n\nJohannes Sixt (1):\n      t3701: ensure correctly set up repository after skipped tests\n\nJunio C Hamano (8):\n      Revert \"git-add--interactive: remove hunk coalescing\"\n      completion: fix PS1 display during a merge on detached HEAD\n      completion: simplify \"current branch\" in __git_ps1()\n      completion: enhance \"current branch\" display\n      Prepare for 1.6.3.2\n      Update draft release notes for 1.6.3.2\n      blame: correctly handle a path that used to be a directory\n      GIT 1.6.3.2\n\nLinus Torvalds (3):\n      Avoid unnecessary 'lstat()' calls in 'get_stat_data()'\n      Teach 'git checkout' to preload the index contents\n      dir.c: clean up handling of 'path' parameter in read_directory_recursive()\n\nMatt Graham (1):\n      Splitting a hunk that adds a line at the top fails in \"add -p\"\n\nMichael J Gruber (1):\n      merge-options.txt: Clarify merge --squash\n\nNanako Shiraishi (1):\n      test: checkout shouldn't say that HEAD has moved if it didn't\n\nNicolas Pitre (1):\n      allow OFS_DELTA objects during a push\n\nRené Scharfe (4):\n      grep: fix word-regexp colouring\n      grep: fix word-regexp at the beginning of lines\n      grep: fix colouring of matches with zero length\n      grep: fix empty word-regexp matches\n\nStephen Boyd (6):\n      api-parse-options.txt: use 'func' instead of 'funct'\n      tests: Add tests for missing format-patch long options\n      completion: add missing options to show-branch and show\n      show-branch: Fix die message in parse_reflog_param()\n      commit: -F overrides -t\n      apply, fmt-merge-msg: use relative filenames\n\nThomas Rast (2):\n      Documentation: teach stash/pop workflow instead of stash/apply\n      add -i: do not dump patch during application\n\nTony Kemp (1):\n      Turn on USE_ST_TIMESPEC for OpenBSD\n"}]}