{"thread":{"id":"12895","subject":"[ANNOUNCE] GIT 1.5.4.5","startedAt":"2008-03-28T06:26:41Z","lastAt":"2008-03-28T06:26:41Z","messageCount":1,"participants":["Junio C Hamano"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"73240","messageId":"7vtzir2wke.fsf@gitster.siamese.dyndns.org","threadId":"12895","inReplyTo":null,"subject":"[ANNOUNCE] GIT 1.5.4.5","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2008-03-28T06:26:41Z","receivedAt":"2008-03-28T06:26:41Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"The latest maintenance release GIT 1.5.4.5 is available at the\nusual places:\n\n  http://www.kernel.org/pub/software/scm/git/\n\n  git-1.5.4.5.tar.{gz,bz2}\t\t\t(tarball)\n  git-htmldocs-1.5.4.5.tar.{gz,bz2}\t\t(preformatted docs)\n  git-manpages-1.5.4.5.tar.{gz,bz2}\t\t(preformatted docs)\n  RPMS/$arch/git-*-1.5.4.5-1.$arch.rpm\t(RPM)\n\nAmong many fixes, a notable one is a regression we introduced in 1.5.4\nthat changed the behaviour of \"git fetch there\" when you have the URL\ninformation for \"there\" in .git/branches/there.  Such a fetch should have\nupdated your local branch \"there\", but 1.5.4 and later didn't.  This\nshould fix the breakage.\n\n----------------------------------------------------------------\n\nGIT v1.5.4.5 Release Notes\n==========================\n\nFixes since v1.5.4.4\n--------------------\n\n * \"git fetch there\" when the URL information came from the Cogito style\n   branches/there file did not update refs/heads/there (regression in\n   1.5.4).\n\n * Bogus refspec configuration such as \"remote.there.fetch = =\" were not\n   detected as errors (regressionin 1.5.4).\n\n * You couldn't specify a custom editor whose path contains a whitespace\n   via GIT_EDITOR (and core.editor).\n\n * The subdirectory filter to \"git filter-branch\" mishandled a history\n   where the subdirectory becomes empty and then later becomes non-empty.\n\n * \"git shortlog\" gave an empty line if the original commit message was\n   malformed (e.g. a botched import from foreign SCM).  Now it finds the\n   first non-empty line and uses it for better information.\n\n * When the user fails to give a revision parameter to \"git svn\", an error\n   from the Perl interpreter was issued because the script lacked proper\n   error checking.\n\n * After \"git rebase\" stopped due to conflicts, if the user played with\n   \"git reset\" and friends, \"git rebase --abort\" failed to go back to the\n   correct commit.\n\n * Additional work trees prepared with git-new-workdir (in contrib/) did\n   not share git-svn metadata directory .git/svn with the original.\n\n * \"git-merge-recursive\" did not mark addition of the same path with\n   different filemodes correctly as a conflict.\n\n * \"gitweb\" gave malformed URL when pathinfo stype paths are in use.\n\n * \"-n\" stands for \"--no-tags\" again for \"git fetch\".\n\n * \"git format-patch\" did not detect the need to add 8-bit MIME header\n   when the user used format.header configuration.\n\n * \"rev~\" revision specifier used to mean \"rev\", which was inconsistent\n   with how \"rev^\" worked.  Now \"rev~\" is the same as \"rev~1\" (hence it\n   also is the same as \"rev^1\"), and \"rev~0\" is the same as \"rev^0\"\n   (i.e. it has to be a commit).\n\n * \"git quiltimport\" did not grok empty lines, lines in \"file -pNNN\"\n   format to specify the prefix levels and lines with trailing comments.\n\n * \"git rebase -m\" triggered pre-commit verification, which made\n   \"rebase --continue\" impossible.\n\nAs usual, it also comes with many documentation fixes and clarifications.\n\n----------------------------------------------------------------\n\nChanges since v1.5.4.4 are as follows:\n\nAndy Whitcroft (1):\n      shortlog: take the first populated line of the description\n\nBernt Hansen (1):\n      git-new-workdir: Share SVN meta data between work dirs and the repository\n\nBrandon Casey (1):\n      t/t3800-mktag.sh: use test_must_fail rather than '!'\n\nClemens Buchacher (1):\n      merge-recursive: handle file mode changes\n\nDaniel Barkalow (2):\n      Tighten refspec processing\n      Fix branches file configuration\n\nEric Wong (1):\n      git-svn: don't blindly append '*' to branch/tags config\n\nGuanqun Lu (1):\n      Fix the wrong output of `git-show v1.3.0~155^2~4` in documentation.\n\nJakub Narebski (1):\n      gitweb: Fix bug in href(..., -replay=>1) when using 'pathinfo' form\n\nJeff King (2):\n      t0021: tr portability fix for Solaris\n      Documentation: clarify use of .git{ignore,attributes} versus .git/info/*\n\nJohannes Schindelin (3):\n      launch_editor(): allow spaces in the filename\n      git fetch: Take '-n' to mean '--no-tags'\n      merge-file: handle empty files gracefully\n\nJonas Fonseca (1):\n      Make man page building quiet when DOCBOOK_XSL_172 is defined\n\nJunio C Hamano (9):\n      filter-branch: handle \"disappearing tree\" case correctly in subdir filter\n      git-pull documentation: warn about the option order\n      quiltimport: fix misquoting of parsed -p<num> parameter\n      format-patch: generate MIME header as needed even when there is format.header\n      rebase -m: do not trigger pre-commit verification\n      Start draft ReleaseNotes for 1.5.4.5\n      git-fetch test: test tracking fetch results, not just FETCH_HEAD\n      Update draft release notes for 1.5.4.5\n      GIT 1.5.4.5\n\nLinus Torvalds (1):\n      rev-parse: fix meaning of rev~ vs rev~0.\n\nMarc-Andre Lureau (2):\n      git-svn: fix find-rev error message when missing arg\n      git-cvsimport: fix merging with remote parent branch\n\nMike Hommey (1):\n      git rebase --abort: always restore the right commit\n\nPierre Habouzit (1):\n      git-quiltimport: better parser to grok \"enhanced\" series files.\n\nVineet Kumar (1):\n      Minor wording changes in the keyboard descriptions in git-add --interactive.\n"}]}