{"thread":{"id":"2517","subject":"[ANNOUNCE] GIT 0.99.9i aka 1.0rc2","startedAt":"2005-11-15T04:17:02Z","lastAt":"2005-11-15T23:48:58Z","messageCount":3,"participants":["Junio C Hamano","walt"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"11866","messageId":"7vr79isfy9.fsf@assigned-by-dhcp.cox.net","threadId":"2517","inReplyTo":null,"subject":"[ANNOUNCE] GIT 0.99.9i aka 1.0rc2","fromName":"Junio C Hamano","fromEmail":"junkio@cox.net","sentAt":"2005-11-15T04:17:02Z","receivedAt":"2005-11-15T04:17:02Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"GIT 0.99.9i aka 1.0rc2 is found at usual places.\n\nI think the source-tree-wise almost everything is done except:\n\n - http-fetch file descriptor leak fix; I tried Nick's\n   clean-ups, but haven't tried Pasky's patch yet.  Walt reports\n   neither patch fixed the problem.  I wasted the weekend not\n   being able to reproduce this myself, until Pasky reminded me\n   that I have an old special code in git-clone, which was\n   unrelated to this problem, but nevertheless was masking it.\n\n - Preparation for later dash-less \"git frotz\" installation\n   scheme.  As I said earlier, I intend to do 1.0 with a\n   Makefile that installs everything in one directory (either\n   $HOME/bin or /usr/bin) by default, but at least we should\n   encourage our users to get used to spelling the subcommands\n   in dash-less form.  Andreas Ericsson's git wrapper\n   implementation was reviewed favourably on the list, and I'd\n   like to go with it.\n\n - archimport update by Eric Wong has not been reviewed yet;\n   this is difficult for me since I do not use tla anymore, and\n   I am asking help from Martin Langhoff.  I am hoping Eric and\n   Martin would come up with an updated patch that satisfies the\n   needs of both, at that time the update will happen.\n   Personally I do not think 1.0 has to wait for this.\n\nThere is a bit of restructuring in the binary packaging for RPM\n(and probably Debian side needs matching change as well) before\n1.0 can happen.\n\nAnyway, here is the short-log between 0.99.9g and 0.99.9i.\nPlease give it a good beating.\n\n-- >8 -- cut here -- >8 --\n\nAlex Riesen:\n      allow git-update-ref create refs with slashes in names\n\nAndreas Ericsson:\n      git-clone: Keep remote names when cloning unless explicitly told not to.\n      git-clone: Allow cloning into directories other than child of current dir.\n      git-branch: Mention -d and -D in man-page.\n\nChris Wright:\n      specfile cleanups\n\nFredrik Kuivinen:\n      merge-recursive: Indent the output properly\n      merge-recursive: Add copyright notice\n      merge-recursive: Use '~' instead of '_' to separate file names from branch names\n\nJim Radford:\n      Add missing git-core and cvsps RPM dependencies.\n\nJosef Weidendorfer:\n      Let git-clone/git-fetch follow HTTP redirections\n      Bugfix: stop if directory already exists\n      Remove git-rename. git-mv does the same\n\nJunio C Hamano:\n      Documentation: \"host:path/to/repo\" is git native over ssh.\n      Do not lose author name information to locale gotchas.\n      Add --pretty=fuller\n      octopus: do not do AND'ed merge base.\n      RPM: arch submodule needs tla.\n      merge-base: fully contaminate the well.\n      merge-base: avoid unnecessary postprocessing.\n      Add test case for merge-base.\n      git-show-branch: tighten merge-base computation.\n      Fully detect uninteresting commits.\n      t1200: use --topo-order to keep the show-branch output stable.\n      INSTALL: duplicate python requirements from Makefile\n      merge with /dev/null as base, instead of punting O==empty case\n      merge-one-file: use common as base, instead of emptiness.\n      Documentation: git-apply --no-add\n      merge-one-file: use empty- or common-base condintionally in two-stage merge.\n      git-prune: prune redundant packs\n      git-lost+found\n      Rename .git/lost+found to .git/lost-found.\n      Documentation: asciidoc sources are utf-8\n      Ignore built git-lost+found.\n      Debian: build-depend on libexpat-dev.\n      Rename lost+found to lost-found.\n      Separate LDFLAGS and CFLAGS.\n      apply: fix binary patch detection.\n      Update topo-order test.\n\nKai Ruemmler:\n      Fix compilation warnings in pack-redundant.c\n\nLinus Torvalds:\n      Fix git-rev-list \"date order\" with --topo-order\n\nLukas_Sandström:\n      Change 'cache' to 'index' in the docs\n      Add git-pack-intersect\n      Add documentation for git-pack-intersect\n      Add git-pack-intersect to .gitignore\n      Make git-repack use git-pack-intersect.\n      Rename git-pack-intersect to git-pack-redundant\n      Make git-pack-redundant consider alt-odbs\n\nMartin Langhoff:\n      archimport: handle pika escaping\n\nMatthias Urlichs:\n      debian packaging: git-cvs needs cvsps\n      Remove trailing slashes\n      Depend on asciidoc 7 (at least).\n\nNick Hengeveld:\n      Fix for multiple alternates requests in http-fetch\n      Fix fd leak in http-fetch\n\nNikolai Weibull:\n      Document the -n command-line option to git-unpack-objects\n      Document a couple of missing command-line options.\n      Documentation nitpicking\n\nPavel Roskin:\n      Add --no-commit-id option for git-diff-tree, use it in gitk\n      git-clone: quote destination directory name\n\nPetr Baudis:\n      Fix confusing git-update-ref error message\n      Show URL in the \"Getting <foo> list\" http-fetch messages\n\nThomas Matysik:\n      Add expat and expat-devel dependencies (for http-push) to RPM spec.\n      Split gitk into seperate RPM package\n"},{"id":"11899","messageId":"Pine.LNX.4.64.0511150715390.17817@x2.ybpnyarg","threadId":"2517","inReplyTo":"7vr79isfy9.fsf@assigned-by-dhcp.cox.net","subject":"Re: [ANNOUNCE] GIT 0.99.9i aka 1.0rc2","fromName":"walt","fromEmail":"wa1ter@myrealbox.com","sentAt":"2005-11-15T15:26:37Z","receivedAt":"2005-11-15T15:26:37Z","isPatch":false,"sender":{"key":"wa1ter@myrealbox.com","avatar":null},"body":"\n\nOn Mon, 14 Nov 2005, Junio C Hamano wrote:\n\n> GIT 0.99.9i aka 1.0rc2 is found at usual places.\n>\n> I think the source-tree-wise almost everything is done except:\n>\n>  - http-fetch file descriptor leak fix; I tried Nick's\n>    clean-ups, but haven't tried Pasky's patch yet.  Walt reports\n>    neither patch fixed the problem.  I wasted the weekend not\n>    being able to reproduce this myself, until Pasky reminded me\n>    that I have an old special code in git-clone, which was\n>    unrelated to this problem, but nevertheless was masking it.\n\nSo, you're saying that you have *not* fixed it?  Hm.  Using the\nup-to-the-minute repository versions of cogito and git I can now\ndo a successful clone of the git repository with ulimit -n 64,\nwhich I couldn't do two days ago.  When I get home I'll try it\nagain on NetBSD also.\n"},{"id":"11964","messageId":"Pine.NEB.4.63.0511151539360.12158@x9.ybpnyarg","threadId":"2517","inReplyTo":"Pine.LNX.4.64.0511150715390.17817@x2.ybpnyarg","subject":"Re: [ANNOUNCE] GIT 0.99.9i aka 1.0rc2","fromName":"walt","fromEmail":"wa1ter@myrealbox.com","sentAt":"2005-11-15T23:48:58Z","receivedAt":"2005-11-15T23:48:58Z","isPatch":false,"sender":{"key":"wa1ter@myrealbox.com","avatar":null},"body":"On Tue, 15 Nov 2005, walt wrote:\n\n> On Mon, 14 Nov 2005, Junio C Hamano wrote:\n>\n> > I think the source-tree-wise almost everything is done except:\n> >  - http-fetch file descriptor leak fix...\n\n> So, you're saying that you have *not* fixed it?...\n\nI just confirmed the good news on NetBSD.  Out of curiosity I did\nthis test:  I cg-updated and reinstalled cogito, which did *not*\nfix the too-many-open-files error.  Then I cg-updated and rebuilt\ngit which *did* fix the error.  Clearly, something you committed\nin the last two days has fixed this problem.  Dunno what you did,\nbut thanks :o)\n"}]}