{"thread":{"id":"12274","subject":"[ANNOUNCE] GIT 1.5.4.3","startedAt":"2008-02-23T21:07:24Z","lastAt":"2008-03-11T19:11:41Z","messageCount":7,"participants":["Junio C Hamano","Jeff Garzik"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"69708","messageId":"7vr6f31iwj.fsf@gitster.siamese.dyndns.org","threadId":"12274","inReplyTo":null,"subject":"[ANNOUNCE] GIT 1.5.4.3","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2008-02-23T21:07:24Z","receivedAt":"2008-02-23T21:07:24Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"The latest maintenance release GIT 1.5.4.3 is available at the\nusual places:\n\n  http://www.kernel.org/pub/software/scm/git/\n\n  git-1.5.4.3.tar.{gz,bz2}\t\t\t(tarball)\n  git-htmldocs-1.5.4.3.tar.{gz,bz2}\t\t(preformatted docs)\n  git-manpages-1.5.4.3.tar.{gz,bz2}\t\t(preformatted docs)\n  RPMS/$arch/git-*-1.5.4.3-1.$arch.rpm\t(RPM)\n\nLargest user visible change in this is RPM packaging updates by\nKristian Høgsberg.  'git-core' will only be pure git without\npulling foreign SCM packages in as its dependencies anymore when\nyou do \"yum install git-core\".\n\n----------------------------------------------------------------\n\nGIT v1.5.4.3 Release Notes\n==========================\n\nFixes since v1.5.4.2\n--------------------\n\n * RPM spec used to pull in everything with 'git'.  This has been\n   changed so that 'git' package contains just the core parts,\n   and we now supply 'git-all' metapackage to slurp in everything.\n   This should match end user's expectation better.\n\n * When some refs failed to update, git-push reported \"failure\"\n   which was unclear if some other refs were updated or all of\n   them failed atomically (the answer is the former).  Reworded\n   the message to clarify this.\n\n * \"git clone\" from a repository whose HEAD was misconfigured\n   did not set up the remote properly.  Now it tries to do\n   better.\n\n * Updated git-push documentation to clarify what \"matching\"\n   means, in order to reduce user confusion.\n\n * Updated git-add documentation to clarify \"add -u\" operates in\n   the current subdirectory you are in, just like other commands.\n\n * git-gui updates to work on OSX and Windows better.\n\n----------------------------------------------------------------\n\nChanges since v1.5.4.2 are as follows:\n\nGerrit Pape (1):\n      git-clone.sh: properly configure remote even if remote's head is dangling\n\nJay Soffian (2):\n      git-gui: support Git Gui.app under OS X 10.5\n      send-email: squelch warning due to comparing undefined $_ to \"\"\n\nJeff King (4):\n      push: indicate partialness of error message\n      Documentation/push: clarify matching refspec behavior\n      push: document the status output\n      hash: fix lookup_hash semantics\n\nJunio C Hamano (1):\n      GIT 1.5.4.3\n\nKristian H淡gsberg (1):\n      Rename git-core rpm to just git and rename the meta-pacakge to git-all.\n\nMiklos Vajna (1):\n      Documentation/git-stash: document options for git stash list\n\nPekka Kaitaniemi (1):\n      Clarified the meaning of git-add -u in the documentation\n\nShawn O. Pearce (5):\n      git-gui: Ensure error dialogs always appear over all other windows\n      git-gui: Paper bag fix error dialogs opening over the main window\n      git-gui: Default TCL_PATH to same location as TCLTK_PATH\n      git-gui: Avoid hardcoded Windows paths in Cygwin package files\n      git-gui: Focus insertion point at end of strings in repository chooser\n\nWincent Colaiuta (1):\n      git-gui: relax \"dirty\" version detection\n"},{"id":"71505","messageId":"7vr6eknpjf.fsf@gitster.siamese.dyndns.org","threadId":"12274","inReplyTo":"7vr6f31iwj.fsf@gitster.siamese.dyndns.org","subject":"[ANNOUNCE] GIT 1.5.4.4","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2008-03-09T10:46:12Z","receivedAt":"2008-03-09T10:46:12Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"The latest maintenance release GIT 1.5.4.4 is available at the\nusual places:\n\n  http://www.kernel.org/pub/software/scm/git/\n\n  git-1.5.4.4.tar.{gz,bz2}\t\t\t(tarball)\n  git-htmldocs-1.5.4.4.tar.{gz,bz2}\t\t(preformatted docs)\n  git-manpages-1.5.4.4.tar.{gz,bz2}\t\t(preformatted docs)\n  RPMS/$arch/git-*-1.5.4.4-1.$arch.rpm\t(RPM)\n\n----------------------------------------------------------------\n\nGIT v1.5.4.4 Release Notes\n==========================\n\nFixes since v1.5.4.3\n--------------------\n\n * Building and installing with an overtight umask such as 077 made\n   installed templates unreadable by others, while the rest of the install\n   are done in a way that is friendly to umask 022.\n\n * \"git cvsexportcommit -w $cvsdir\" misbehaved when GIT_DIR is set to a\n   relative directory.\n\n * \"git http-push\" had an invalid memory access that could lead it to\n   segfault.\n\n * When \"git rebase -i\" gave control back to the user for a commit that is\n   marked to be edited, it just said \"modify it with commit --amend\",\n   without saying what to do to continue after modifying it.  Give an\n   explicit instruction to run \"rebase --continue\" to be more helpful.\n\n * \"git send-email\" in 1.5.4.3 issued a bogus empty In-Reply-To: header.\n\n * \"git bisect\" showed mysterious \"won't bisect on seeked tree\" error message.\n   This was leftover from Cogito days to prevent \"bisect\" starting from a\n   cg-seeked state.  We still keep the Cogito safety, but running \"git bisect\n   start\" when another bisect was in effect will clean up and start over.\n\n * \"git push\" with an explicit PATH to receive-pack did not quite work if\n   receive-pack was not on usual PATH.  We earlier fixed the same issue\n   with \"git fetch\" and upload-pack, but somehow forgot to do so in the\n   other direction.\n\n * git-gui's info dialog was not displayed correctly when the user tries\n   to commit nothing (i.e. without staging anything).\n\n * \"git revert\" did not properly fail when attempting to run with a\n   dirty index.\n\n * \"git merge --no-commit --no-ff <other>\" incorrectly made commits.\n\n * \"git merge --squash --no-ff <other>\", which is a nonsense combination\n   of options, was not rejected.\n\n * \"git ls-remote\" and \"git remote show\" against an empty repository\n   failed, instead of just giving an empty result (regression).\n\n * \"git fast-import\" did not handle a renamed path whose name needs to be\n   quoted, due to a bug in unquote_c_style() function.\n\n * \"git cvsexportcommit\" was confused when multiple files with the same\n   basename needed to be pushed out in the same commit.\n\n * \"git daemon\" did not send early errors to syslog.\n\n * \"git log --merge\" did not work well with --left-right option.\n\n * \"git svn\" promprted for client cert password every time it accessed the\n   server.\n\n * The reset command in \"git fast-import\" data stream was documented to\n   end with an optional LF, but it actually required one.\n\n * \"git svn dcommit/rebase\" did not honor --rewrite-root option.\n\nAlso included are a handful documentation updates.\n\n----------------------------------------------------------------\n\nChanges since v1.5.4.3 are as follows:\n\nAdeodato Simó (1):\n      Really make the LF after reset in fast-import optional\n\nBjörn Steinbrink (1):\n      receive-pack: Initialize PATH to include exec-dir.\n\nBrandon Casey (1):\n      builtin-reflog.c: don't install new reflog on write failure\n\nBryan Donlan (1):\n      Documentation/git-am.txt: Pass -r in the example invocation of rm -f .dotest\n\nCaio Marcelo de Oliveira Filho (1):\n      filter-branch documentation: non-zero exit status in command abort the filter\n\nCarl Worth (1):\n      Eliminate confusing \"won't bisect on seeked tree\" failure\n\nDaniel Barkalow (3):\n      Use a single implementation and API for copy_file()\n      Don't use GIT_CONFIG in t5505-remote\n      Correct name of diff_flush() in API documentation\n\nGerrit Pape (2):\n      templates/Makefile: don't depend on local umask setting\n      git-merge.sh: better handling of combined --squash,--no-ff,--no-commit options\n\nJay Soffian (2):\n      rev-parse: fix potential bus error with --parseopt option spec handling\n      send-email: fix In-Reply-To regression\n\nJeff King (1):\n      revert: actually check for a dirty index\n\nJohan Herland (2):\n      Add testcase for 'git cvsexportcommit -w $cvsdir ...' with relative $GIT_DIR\n      Fix 'git cvsexportcommit -w $cvsdir ...' when used with relative $GIT_DIR\n\nJohannes Schindelin (4):\n      http-push: avoid invalid memory accesses\n      http-push: do not get confused by submodules\n      http-push: avoid a needless goto\n      cvsexportcommit: be graceful when \"cvs status\" reorders the arguments\n\nJohannes Sixt (2):\n      daemon: send more error messages to the syslog\n      daemon: ensure that base-path is an existing directory\n\nJohn Goerzen (1):\n      Fix dcommit, rebase when rewriteRoot is in use\n\nJonathan del Strother (1):\n      Prompt to continue when editing during rebase --interactive\n\nJunio C Hamano (6):\n      Fix \"git log --merge --left-right\"\n      Start preparing for 1.5.4.4\n      tests: introduce test_must_fail\n      Update draft release notes for 1.5.4.4\n      test-lib: fix TERM to dumb for test repeatability\n      GIT 1.5.4.4\n\nMatthieu Moy (1):\n      Fix incorrect wording in git-merge.txt.\n\nMike Hommey (2):\n      Set proxy override with http_init()\n      Fix random crashes in http_cleanup()\n\nMike Ralphson (1):\n      Documentation cherry-pick: Fix cut-and-paste error\n\nMiklos Vajna (2):\n      Documentation/git-filter-branch: add a new msg-filter example\n      Documentation/git svn log: add a note about timezones.\n\nPierre Habouzit (1):\n      unquote_c_style: fix off-by-one.\n\nPing Yin (1):\n      git-submodule: Fix typo 'url' which should be '$url'\n\nRémi Vanicat (1):\n      git.el: find the git-status buffer whatever its name is\n\nSanti Béjar (1):\n      ident.c: reword error message when the user name cannot be determined\n\nSebastian Noack (1):\n      git-svn: Don't prompt for client cert password everytime.\n\nShawn O. Pearce (6):\n      Ensure 'make dist' compiles git-archive.exe on Cygwin\n      Protect peel_ref fallback case from NULL parse_object result\n      Correct fast-export file mode strings to match fast-import standard\n      git-gui: Paper bag fix info dialog when no files are staged at commit\n      Fix 'git remote show' regression on empty repository in 1.5.4\n      git-gui: Gracefully fall back to po2msg.sh if msgfmt --tcl fails\n\nSteven Drake (1):\n      timezone_names[]: fixed the tz offset for New Zealand.\n\nUwe Kleine-König (1):\n      config.txt: refer to --upload-pack and --receive-pack instead of --exec\n\n\n"},{"id":"71518","messageId":"47D416C8.1020903@garzik.org","threadId":"12274","inReplyTo":"7vr6eknpjf.fsf@gitster.siamese.dyndns.org","subject":"Re: [ANNOUNCE] GIT 1.5.4.4","fromName":"Jeff Garzik","fromEmail":"jeff@garzik.org","sentAt":"2008-03-09T16:56:40Z","receivedAt":"2008-03-09T16:56:40Z","isPatch":false,"sender":{"key":"jeff@garzik.org","avatar":null},"body":"Junio C Hamano wrote:\n> The latest maintenance release GIT 1.5.4.4 is available at the\n> usual places:\n> \n>   http://www.kernel.org/pub/software/scm/git/\n> \n>   git-1.5.4.4.tar.{gz,bz2}\t\t\t(tarball)\n>   git-htmldocs-1.5.4.4.tar.{gz,bz2}\t\t(preformatted docs)\n>   git-manpages-1.5.4.4.tar.{gz,bz2}\t\t(preformatted docs)\n>   RPMS/$arch/git-*-1.5.4.4-1.$arch.rpm\t(RPM)\n\nDoes it address the following issue, present in git-core-1.5.4.1-1.fc8?\n\nhttp://marc.info/?l=git&m=120423022832530&w=2\n\nThanks,\n\n\tJeff\n"},{"id":"71521","messageId":"7v1w6jod54.fsf@gitster.siamese.dyndns.org","threadId":"12274","inReplyTo":"47D416C8.1020903@garzik.org","subject":"Re: [ANNOUNCE] GIT 1.5.4.4","fromName":"Junio C Hamano","fromEmail":"junio@pobox.com","sentAt":"2008-03-09T20:28:39Z","receivedAt":"2008-03-09T20:28:39Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Jeff Garzik <jeff@garzik.org> writes:\n\n> Junio C Hamano wrote:\n>> The latest maintenance release GIT 1.5.4.4 is available at the\n>> usual places:\n>>\n>>   http://www.kernel.org/pub/software/scm/git/\n>>\n>>   git-1.5.4.4.tar.{gz,bz2}\t\t\t(tarball)\n>>   git-htmldocs-1.5.4.4.tar.{gz,bz2}\t\t(preformatted docs)\n>>   git-manpages-1.5.4.4.tar.{gz,bz2}\t\t(preformatted docs)\n>>   RPMS/$arch/git-*-1.5.4.4-1.$arch.rpm\t(RPM)\n>\n> Does it address the following issue, present in git-core-1.5.4.1-1.fc8?\n>\n> http://marc.info/?l=git&m=120423022832530&w=2\n\nI do not think so.\n\nIs it really an issue, or is it just a warning message unread/unfollowed?\n\nI am comparing the last line you quoted from the command output in that\nmessage, which suggests the user to run 'git prune', and your comment on\nthe next line in that message that says \"I regularly run 'git gc'\", and\nscratching my head.  I cannot tell if you regularly run 'git prune' or not\nfrom it...\n"},{"id":"71529","messageId":"47D459D5.2020101@garzik.org","threadId":"12274","inReplyTo":"7v1w6jod54.fsf@gitster.siamese.dyndns.org","subject":"Re: [ANNOUNCE] GIT 1.5.4.4","fromName":"Jeff Garzik","fromEmail":"jeff@garzik.org","sentAt":"2008-03-09T21:42:45Z","receivedAt":"2008-03-09T21:42:45Z","isPatch":false,"sender":{"key":"jeff@garzik.org","avatar":null},"body":"Junio C Hamano wrote:\n> Jeff Garzik <jeff@garzik.org> writes:\n> \n>> Junio C Hamano wrote:\n>>> The latest maintenance release GIT 1.5.4.4 is available at the\n>>> usual places:\n>>>\n>>>   http://www.kernel.org/pub/software/scm/git/\n>>>\n>>>   git-1.5.4.4.tar.{gz,bz2}\t\t\t(tarball)\n>>>   git-htmldocs-1.5.4.4.tar.{gz,bz2}\t\t(preformatted docs)\n>>>   git-manpages-1.5.4.4.tar.{gz,bz2}\t\t(preformatted docs)\n>>>   RPMS/$arch/git-*-1.5.4.4-1.$arch.rpm\t(RPM)\n>> Does it address the following issue, present in git-core-1.5.4.1-1.fc8?\n>>\n>> http://marc.info/?l=git&m=120423022832530&w=2\n> \n> I do not think so.\n> \n> Is it really an issue, or is it just a warning message unread/unfollowed?\n\nIt's not a warning message, it is an annoying delay that has been added \nto almost -every- local pull, impacting my main kernel workflow.\n\nFurther -- as my email demonstrated with examples -- it would repeatedly \n'git gc' on the same repository over and over again, for each 'git pull' \nor 'git rebase' that I did.  That is overly excessive.\n\n\n> I am comparing the last line you quoted from the command output in that\n> message, which suggests the user to run 'git prune', and your comment on\n> the next line in that message that says \"I regularly run 'git gc'\", and\n> scratching my head.  I cannot tell if you regularly run 'git prune' or not\n> from it...\n\nYes, I regularly run both 'git gc' and 'git prune'.\n\nBut since (ref original email) I was doing some rebasing, there are \ninevitably changesets left dangling after such an operation.\n\n\tJeff\n"},{"id":"71557","messageId":"7vmyp7kryp.fsf@gitster.siamese.dyndns.org","threadId":"12274","inReplyTo":"47D459D5.2020101@garzik.org","subject":"Re: [ANNOUNCE] GIT 1.5.4.4","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2008-03-10T06:34:22Z","receivedAt":"2008-03-10T06:34:22Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Jeff Garzik <jeff@garzik.org> writes:\n\n> Yes, I regularly run both 'git gc' and 'git prune'.\n>\n> But since (ref original email) I was doing some rebasing, there are\n> inevitably changesets left dangling after such an operation.\n\nYeah, I'd say it is stupid if \"am\" ran \"gc --auto\" for every patch.  I\nrecall that we had the same issue with git-svn and we made it run once\nevery 1k round, and we probably should do the same for \"am\" and \"rebase\",\nrunning once at the very end.\n\nI notice however that git-am does exactly that.  It runs \"gc --auto\" only\nat the end, and does not run it when it stops upon unapplicable patch.\n\nPerhaps we would want to raise the default \"gc --auto\" limit?  Currently\nwhen it estimates that you have roughly 6700 objects unpacked it runs\n\"repack --prune-packed\", and if there still are that many unpacked objects\nafter that, it suggests you to run \"git prune\" to remove them.  If you are\nrebasing, the commits in the old history that are rewritten will _not_\nimmediately become dangling because they will still be reachable from your\nreflog.  If you are getting the message, these objects were already\ndangling (ancient commits that are not even reachable from your reflog\nentries that are by default kept for 90 days) even before you started your\nrebase or am run.\n\nAfter you finished your day's work on a typical day, what does the output\nfrom \"git count-objects -v\" and \"git fsck-objects\" look like, I wonder?\n"},{"id":"71750","messageId":"47D6D96D.2000302@garzik.org","threadId":"12274","inReplyTo":"7vmyp7kryp.fsf@gitster.siamese.dyndns.org","subject":"Re: [ANNOUNCE] GIT 1.5.4.4","fromName":"Jeff Garzik","fromEmail":"jeff@garzik.org","sentAt":"2008-03-11T19:11:41Z","receivedAt":"2008-03-11T19:11:41Z","isPatch":false,"sender":{"key":"jeff@garzik.org","avatar":null},"body":"Junio C Hamano wrote:\n> Jeff Garzik <jeff@garzik.org> writes:\n> \n>> Yes, I regularly run both 'git gc' and 'git prune'.\n>>\n>> But since (ref original email) I was doing some rebasing, there are\n>> inevitably changesets left dangling after such an operation.\n> \n> Yeah, I'd say it is stupid if \"am\" ran \"gc --auto\" for every patch.  I\n> recall that we had the same issue with git-svn and we made it run once\n> every 1k round, and we probably should do the same for \"am\" and \"rebase\",\n> running once at the very end.\n\n> Perhaps we would want to raise the default \"gc --auto\" limit?  Currently\n\n\nThat seems quite reasonable.  This \"feels\" like a threshold-too-low problem.\n\n\n\n> when it estimates that you have roughly 6700 objects unpacked it runs\n> \"repack --prune-packed\", and if there still are that many unpacked objects\n> after that, it suggests you to run \"git prune\" to remove them.  If you are\n> rebasing, the commits in the old history that are rewritten will _not_\n> immediately become dangling because they will still be reachable from your\n> reflog.  If you are getting the message, these objects were already\n> dangling (ancient commits that are not even reachable from your reflog\n> entries that are by default kept for 90 days) even before you started your\n> rebase or am run.\n\nMy workflow generally looks like this:\n\n\t# repo was created in this manner....  this was done ONCE,\n\t# not every time I apply patches\n\n\tgit clone --reference ../linux-2.6 ../linux-2.6 libata-dev\n\n\n\t# a patch-applying session\n\n\tgit checkout master\n\tgit pull ../linux-2.6\n\tgit fetch --tags ../linux-2.6\t# yes, still necessary...\n\n\tgit branch -D ALL NEXT\n\tgit branch -D upstream-fixes upstream-linus\n\n\tgit checkout -b upstream-fixes master\n\tgit-am --utf8 --signoff -i /g/tmp/mbox\t# repeat many times...\n\tgit branch upstream-linus upstream-fixes\n\n\tgit-checkout sii-lbt && git-rebase master\n\tgit-checkout mv-ahci-pata && git-rebase master\n\tgit-checkout new-eh && git-rebase master\n\tgit branch NEXT master\n\tgit branch ALL new-eh\n\n\tgit checkout master\n\tgit prune\n\tgit push --force --all $URL\n\nThus, 'git prune' is run on a very regular basis, but 'git gc' is not.\n\nHowever, I presume the lack of 'git gc' regularity on libata-dev.git is \nmitigated by the fact that I _do_ run 'git gc' regularly on \nlinux-2.6.git (listed in libata-dev's alternatives, as noted by \ngit-clone statement above)\n\n\n> After you finished your day's work on a typical day, what does the output\n> from \"git count-objects -v\" and \"git fsck-objects\" look like, I wonder?\n\n[jgarzik@pretzel libata-dev]$ git count-objects -v\ncount: 51\nsize: 244\nin-pack: 475\npacks: 4\nprune-packable: 0\ngarbage: 0\n[jgarzik@pretzel libata-dev]$ git fsck-objects\n[jgarzik@pretzel libata-dev]$\n\n\n\n\nAs an aside...  a git-debug-info might be a useful command, wrapping up \neverything you (a git developer) would find interesting from me (a \nhumble and appreciative git user).  Users could attach the output from \ngit-debug-info to emails, when discussing problems in their repositories.\n\n\tJeff\n"}]}