{"thread":{"id":"9349","subject":"1.5.3 plans and preparations","startedAt":"2007-08-02T07:22:20Z","lastAt":"2007-08-02T09:42:37Z","messageCount":3,"participants":["Junio C Hamano","Johannes Schindelin","David Kastrup"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"49394","messageId":"7vfy32z9hv.fsf@assigned-by-dhcp.cox.net","threadId":"9349","inReplyTo":null,"subject":"1.5.3 plans and preparations","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2007-08-02T07:22:20Z","receivedAt":"2007-08-02T07:22:20Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"I'll be pushing out tonight's update and head to bed in about an\nhour now.\n\nThe changes since -rc3 is rather large, as it contains a large\nfixups to one of the topics that is new to 1.5.3 as well as\nupdates to gitk, git-gui, and gitweb.\n\nI would say this should be pretty much what will be in -rc4 this\nweekend.\n\nPlease test it thouroughly and send in obviously correct fixes.\n\n---\n\nGIT v1.5.3 Release Notes (draft)\n========================\n\nUpdates since v1.5.2\n--------------------\n\n* The commit walkers other than http are officially deprecated,\n  but still supported for now.\n\n* The submodule support has Porcelain layer.\n\n* There are a handful pack-objects changes to help you cope better\n  with repositories with pathologically large blobs in them.\n\n* For people who need to import from Perforce, a front-end for\n  fast-import is in contrib/fast-import/.\n\n* Comes with git-gui 0.8.0.\n\n* Comes with updated gitk.\n\n* New commands and options.\n\n  - \"git log --date=<format>\" can use more formats: iso8601, rfc2822.\n\n  - The hunk header output from \"git diff\" family can be customized\n    with the attributes mechanism.  See gitattributes(5) for details.\n\n  - \"git stash\" allows you to quickly save away your work in\n    progress and replay it later on an updated state.\n\n  - \"git rebase\" learned an \"interactive\" mode that let you\n    pick and reorder which commits to rebuild.\n\n  - \"git fsck\" can save its findings in $GIT_DIR/lost-found, without a\n    separate invocation of \"git lost-found\" command.  The blobs stored by\n    lost-found are stored in plain format to allow you to grep in them.\n\n  - $GIT_WORK_TREE environment variable can be used together with\n    $GIT_DIR to work in a subdirectory of a working tree that is\n    not located at \"$GIT_DIR/..\".\n\n  - Giving \"--file=<file>\" option to \"git config\" is the same as\n    running the command with GIT_CONFIG=<file> environment.\n\n  - \"git log\" learned a new option \"--follow\", to follow\n    renaming history of a single file.\n\n  - \"git-filter-branch\" lets you rewrite the revision history of\n    the current branch, creating a new branch. You can specify a\n    number of filters to modify the commits, files and trees.\n\n  - \"git-cvsserver\" learned new options (--base-path, --export-all,\n    --strict-paths) inspired by git-daemon.\n\n  - \"git daemon --base-path-relaxed\" can help migrating a repository URL\n    that did not use to use --base-path to use --base-path.\n\n  - \"git-commit\" can use \"-t templatefile\" option and commit.template\n    configuration variable to prime the commit message given to you in the\n    editor.\n\n  - \"git-submodule\" command helps you manage the projects from\n    the superproject that contain them.\n\n  - In addition to core.compression configuration option,\n    core.loosecompression and pack.compression options can\n    independently tweak zlib compression levels used for loose\n    and packed objects.\n\n  - \"git-ls-tree -l\" shows size of blobs pointed at by the\n    tree entries, similar to \"/bin/ls -l\".\n\n  - \"git-rev-list\" learned --regexp-ignore-case and\n    --extended-regexp options to tweak its matching logic used\n    for --grep fitering.\n\n  - \"git-describe --contains\" is a handier way to call more\n    obscure command \"git-name-rev --tags\".\n\n  - \"git gc --aggressive\" tells the command to spend more cycles\n    to optimize the repository harder.\n\n  - \"git repack\" learned a \"window-memory\" limit which\n    dynamically reduces the window size to stay within the\n    specified memory usage.\n\n  - \"git repack\" can be told to split resulting packs to avoid\n    exceeding limit specified with \"--max-pack-size\".\n\n  - \"git fsck\" gained --verbose option.  This is really really\n    verbose but it might help you identify exact commit that is\n    corrupt in your repository.\n\n  - \"git format-patch\" learned --numbered-files option.  This\n    may be useful for MH users.\n\n  - \"git format-patch\" learned format.subjectprefix configuration\n    variable, which serves the same purpose as \"--subject-prefix\"\n    option.\n\n  - \"git tag -n -l\" shows tag annotations while listing tags.\n\n  - \"git cvsimport\" can optionally use the separate-remote layout.\n\n  - \"git blame\" can be told to see through commits that change\n    whitespaces and indentation levels with \"-w\" option.\n\n  - \"git send-email\" can be told not to thread the messages when\n    sending out more than one patches.\n\n  - \"git config\" learned NUL terminated output format via -z to\n    help scripts.\n\n  - \"git init -q\" makes the command quieter.\n\n* Updated behavior of existing commands.\n\n  - \"gitweb\" can offer multiple snapshot formats.\n\n    ***NOTE*** Unfortunately, this changes the format of the\n    $feature{snapshot}{default} entry in the per-site\n    configuration file 'gitweb_config.perl'.  It used to be a\n    three-element tuple that describe a single format; with the\n    new configuration item format, you only have to say the name\n    of the format ('tgz', 'tbz2' or 'zip').  Please update the\n    your configuration file accordingly.\n\n  - \"git diff\" (but not the plumbing level \"git diff-tree\") now\n    recursively descends into trees by default.\n\n  - The editor to use with many interactive commands can be\n    overridden with GIT_EDITOR environment variable, or if it\n    does not exist, with core.editor configuration variable.  As\n    before, if you have neither, environment variables VISUAL\n    and EDITOR are consulted in this order, and then finally we\n    fall back on \"vi\".\n\n  - \"git rm --cached\" does not complain when removing a newly\n    added file from the index anymore.\n\n  - Options to \"git log\" to affect how --grep/--author options look for\n    given strings now have shorter abbreviations.  -i is for ignore case,\n    and -E is for extended regexp.\n\n  - \"git svn dcommit\" retains local merge information.\n\n  - \"git config\" to set values also honors type flags like --bool\n    and --int.\n\n  - core.quotepath configuration can be used to make textual git\n    output to emit most of the characters in the path literally.\n\n  - \"git mergetool\" chooses its backend more wisely, taking\n    notice of its environment such as use of X, Gnome/KDE, etc.\n\n  - \"gitweb\" shows merge commits a lot nicer than before.  The\n    default view uses more compact --cc format, while the UI\n    allows to choose normal diff with any parent.\n\n  - snapshot files \"gitweb\" creates from a repository at\n    $path/$project/.git are more useful.  We use $project part\n    in the filename, which we used to discard.\n\n  - \"git cvsimport\" creates lightweight tags; there is no\n    interesting information we can record in an annotated tag,\n    and the handcrafted ones the old code created was not\n    properly formed anyway.\n\n  - \"git-push\" pretends that you immediately fetched back from\n    the remote by updating corresponding remote tracking\n    branches if you have any.\n\n  - The diffstat given after a merge (or a pull) honors the\n    color.diff configuration.\n\n  - \"git commit --amend\" is now compatible with various message source\n    options such as -m/-C/-c/-F.\n\n  - \"git-apply --whitespace=strip\" removes blank lines added at\n    the end of the file.\n\n  - \"git-fetch\" over git native protocols with \"-v\" option shows\n    connection status, and the IP address of the other end, to\n    help diagnosing problems.\n\n  - We used to have core.legacyheaders configuration, when\n    set to false, allowed git to write loose objects in a format\n    that mimicks the format used by objects stored in packs.  It\n    turns out that this was not so useful.  Although we will\n    continue to read objects written in that format, we do not\n    honor that configuration anymore and create loose objects in\n    the legacy/traditional format.\n\n  - \"--find-copies-harder\" option to diff family can now be\n    spelled as \"-C -C\" for brevity.\n\n  - \"git-mailsplit\" (hence \"git-am\") can read from Maildir\n    formatted mailboxes.\n\n  - \"git-cvsserver\" does not barf upon seeing \"cvs login\"\n    request.\n\n  - \"pack-objects\" honors \"delta\" attribute set in\n    .gitattributes.  It does not attempt to deltify blobs that\n    come from paths with delta attribute set to false.\n\n  - \"new-workdir\" script (in contrib) can now be used with a\n    bare repository.\n\n  - \"git-mergetool\" learned to use gvimdiff.\n\n  - \"gitview\" (in contrib) has a better blame interface.\n\n  - \"git log\" and friends did not handle a commit log message\n    that is larger than 16kB; they do now.\n\n  - \"--pretty=oneline\" output format for \"git log\" and friends\n    deals with \"malformed\" commit log messages that have more\n    than one lines in the first paragraph better.  We used to\n    show the first line, cutting the title at mid-sentence; we\n    concatenate them into a single line and treat the result as\n    \"oneline\".\n\n  - \"git p4import\" has been demoted to contrib status.  For\n    a superior option, checkout the git-p4 front end to\n    git-fast-import (also in contrib).  The man page and p4\n    rpm have been removed as well.\n\n  - \"git mailinfo\" (hence \"am\") now tries to see if the message\n    is in utf-8 first, instead of assuming iso-8859-1, if\n    incoming e-mail does not say what encoding it is in.\n\n* Builds\n\n  - old-style function definitions (most notably, a function\n    without parameter defined with \"func()\", not \"func(void)\")\n    have been eradicated.\n\n* Performance Tweaks\n\n  - git-pack-objects avoids re-deltification cost by caching\n    small enough delta results it creates while looking for the\n    best delta candidates.\n\n  - git-pack-objects learned a new heuristcs to prefer delta\n    that is shallower in depth over the smallest delta\n    possible.  This improves both overall packfile access\n    performance and packfile density.\n\n  - diff-delta code that is used for packing has been improved\n    to work better on big files.\n\n  - when there are more than one pack files in the repository,\n    the runtime used to try finding an object always from the\n    newest packfile; it now tries the same packfile as we found\n    the object requested the last time, which exploits the\n    locality of references.\n\n  - verifying pack contents done by \"git fsck --full\" got boost\n    by carefully choosing the order to verify objects in them.\n\n\nFixes since v1.5.2\n------------------\n\nAll of the fixes in v1.5.2 maintenance series are included in\nthis release, unless otherwise noted.\n\n* Bugfixes\n\n  - \"gitweb\" had trouble handling non UTF-8 text with older\n    Encode.pm Perl module.\n\n--\nexec >/var/tmp/1\nO=v1.5.3-rc3-119-g50cff52\necho O=`git describe refs/heads/master`\ngit shortlog --no-merges $O..refs/heads/master ^refs/heads/maint\n"},{"id":"49401","messageId":"Pine.LNX.4.64.0708021021060.14781@racer.site","threadId":"9349","inReplyTo":"7vfy32z9hv.fsf@assigned-by-dhcp.cox.net","subject":"Re: 1.5.3 plans and preparations","fromName":"Johannes Schindelin","fromEmail":"johannes.schindelin@gmx.de","sentAt":"2007-08-02T09:26:11Z","receivedAt":"2007-08-02T09:26:11Z","isPatch":false,"sender":{"key":"johannes.schindelin@gmx.de","avatar":"https://avatars.githubusercontent.com/u/127790?v=4"},"body":"Hi,\n\nwhew, what a bunch of changes!\n\nOn Thu, 2 Aug 2007, Junio C Hamano wrote:\n\n>   - \"git-filter-branch\" lets you rewrite the revision history of\n>     the current branch, creating a new branch. You can specify a\n>     number of filters to modify the commits, files and trees.\n\nIt is no longer the current branch, not even by default.  You can rewrite \none or more branches, and you have to specify them.\n\n>   - \"git commit --amend\" is now compatible with various message source\n>     options such as -m/-C/-c/-F.\n\nUrgh.  That makes it much more easy for new users to shoot themselves in \nthe foot.  By \"-C\" and \"-F\" you do not even see the runstatus anymore, \nsince no editor was fired up.  Please say something like \"use with care\", \nor \"do not try this at home\", too.\n\nCiao,\nDscho\n"},{"id":"49402","messageId":"868x8umfw2.fsf@lola.quinscape.zz","threadId":"9349","inReplyTo":"Pine.LNX.4.64.0708021021060.14781@racer.site","subject":"Re: 1.5.3 plans and preparations","fromName":"David Kastrup","fromEmail":"dak@gnu.org","sentAt":"2007-08-02T09:42:37Z","receivedAt":"2007-08-02T09:42:37Z","isPatch":false,"sender":{"key":"dak@gnu.org","avatar":"https://avatars.githubusercontent.com/u/52141349?v=4"},"body":"Johannes Schindelin <Johannes.Schindelin@gmx.de> writes:\n\n> On Thu, 2 Aug 2007, Junio C Hamano wrote:\n>\n>>   - \"git commit --amend\" is now compatible with various message source\n>>     options such as -m/-C/-c/-F.\n>\n> Urgh.  That makes it much more easy for new users to shoot\n> themselves in the foot.  By \"-C\" and \"-F\" you do not even see the\n> runstatus anymore, since no editor was fired up.  Please say\n> something like \"use with care\", or \"do not try this at home\", too.\n\nHm?  How is that specific to \"--amend\"?  Isn't that just a consequence\nof -m/-C/-F, regardless of whether one uses --amend or not?\n\nWhen there are unmerged changes, --amend will not complete anyway.\nAnd of course, if you got things wrong, you can always do another\n--amend, so it is not like there is serious damage involved.\n\nWhat do I overlook here?\n\nAnd before you complain about not getting a copy by Email again: I\nhave to go via Gmane from my work account, since it will for some\nunfathomable reason drop most of my mails (except those I use for\ntesting the problem) from this address.  It is manual work to add a\n\"Cc\" header in this setup (which I often forget), and even then, I\ndoubt this copy will arrive at your place via mail (please tell me if\nit does).\n\n-- \nDavid Kastrup\n"}]}