{"thread":{"id":"41669","subject":"[ANNOUNCE] Git v2.8.0-rc2","startedAt":"2016-03-10T23:04:44Z","lastAt":"2016-03-22T17:43:25Z","messageCount":22,"participants":["Junio C Hamano","Max Horn","Michael J Gruber","Jiang Xin","Duy Nguyen"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"280594","messageId":"xmqqr3fiaq9f.fsf@gitster.mtv.corp.google.com","threadId":"41669","inReplyTo":null,"subject":"[ANNOUNCE] Git v2.8.0-rc2","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2016-03-10T23:04:44Z","receivedAt":"2016-03-10T23:04:44Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"A release candidate Git v2.8.0-rc2 is now available for testing\nat the usual places. It is comprised of 459 non-merge commits\nsince v2.7.0, contributed by 60 people, 19 of which are new faces.\n\nThe tarballs are found at:\n\n https://www.kernel.org/pub/software/scm/git/testing/\n\nThe following public repositories all have a copy of the\n'v2.8.0-rc2' tag and the 'master' branch that the tag points at:\n\n url = https://kernel.googlesource.com/pub/scm/git/git\n url = git://repo.or.cz/alt-git.git\n url = git://git.sourceforge.jp/gitroot/git-core/git.git\n url = git://git-core.git.sourceforge.net/gitroot/git-core/git-core\n url = https://github.com/gitster/git\n\nNew contributors whose contributions weren't in v2.7.0 are as follows.\nWelcome to the Git development community!\n\n 마누엘, Andrew Wheeler, Changwoo Ryu, Christoph Egger,\n Dan Aloni, Dave Ware, David A. Wheeler, Dickson Wong, Felipe\n Gonçalves Assis, GyuYong Jung, Jon Griffiths, Kazutoshi Satoda,\n Lars Vogel, Martin Amdisen, Matthew Kraai, Paul Wagland, Rob\n Mayoff, Romain Picard, and Victor Leschuk.\n\nReturning contributors who helped this release are as follows.\nThanks for your continued support.\n\n Alexander Kuleshov, Alex Henrie, brian m. carlson, Christian\n Couder, David A. Greene, David Turner, Dennis Kaarsemaker,\n Edmundo Carmona Antoranz, Elia Pinto, Eric Wong, Jacob Keller,\n Jeff King, Jiang Xin, Johannes Schindelin, Johannes Sixt,\n John Keeping, Jonathan Nieder, Junio C Hamano, Karsten Blees,\n Karthik Nayak, Knut Franke, Lars Schneider, Matthieu Moy, Matt\n McCutchen, Michael J Gruber, Mike Hommey, Nguyễn Thái Ngọc\n Duy, Øyvind A. Holm, Patrick Steinhardt, Pat Thoyts, Sebastian\n Schuberth, Shawn O. Pearce, Stefan Beller, Stephen P. Smith,\n SZEDER Gábor, Thomas Ackermann, Thomas Braun, Thomas Gummerer,\n Tobias Klauser, Torsten Bögershausen, and Will Palmer.\n\n----------------------------------------------------------------\n\nGit 2.8 Release Notes (draft)\n=============================\n\nBackward compatibility note\n---------------------------\n\nThe rsync:// transport has been removed.\n\n\nUpdates since v2.7\n------------------\n\nUI, Workflows & Features\n\n * It turns out \"git clone\" over rsync transport has been broken when\n the source repository has packed references for a long time, and\n nobody noticed nor complained about it.\n\n * \"branch --delete\" has \"branch -d\" but \"push --delete\" does not.\n\n * \"git blame\" learned to produce the progress eye-candy when it takes\n too much time before emitting the first line of the result.\n\n * \"git grep\" can now be configured (or told from the command line)\n how many threads to use when searching in the working tree files.\n\n * Some \"git notes\" operations, e.g. \"git log --notes=\", should\n be able to read notes from any tree-ish that is shaped like a notes\n tree, but the notes infrastructure required that the argument must\n be a ref under refs/notes/. Loosen it to require a valid ref only\n when the operation would update the notes (in which case we must\n have a place to store the updated notes tree, iow, a ref).\n\n * \"git grep\" by default does not fall back to its \"--no-index\"\n behaviour outside a directory under Git's control (otherwise the\n user may by mistake end up running a huge recursive search); with a\n new configuration (set in $HOME/.gitconfig--by definition this\n cannot be set in the config file per project), this safety can be\n disabled.\n\n * \"git pull --rebase\" has been extended to allow invoking\n \"rebase -i\".\n\n * \"git p4\" learned to cope with the type of a file getting changed.\n\n * \"git format-patch\" learned to notice format.outputDirectory\n configuration variable. This allows \"-o \" option to be\n omitted on the command line if you always use the same directory in\n your workflow.\n\n * \"interpret-trailers\" has been taught to optionally update a file in\n place, instead of always writing the result to the standard output.\n\n * Many commands that read files that are expected to contain text\n that is generated (or can be edited) by the end user to control\n their behaviour (e.g. \"git grep -f \") have been updated\n to be more tolerant to lines that are terminated with CRLF (they\n used to treat such a line to contain payload that ends with CR,\n which is usually not what the users expect).\n\n * \"git notes merge\" used to limit the source of the merged notes tree\n to somewhere under refs/notes/ hierarchy, which was too limiting\n when inventing a workflow to exchange notes with remote\n repositories using remote-tracking notes trees (located in e.g.\n refs/remote-notes/ or somesuch).\n\n * \"git ls-files\" learned a new \"--eol\" option to help diagnose\n end-of-line problems.\n\n * \"ls-remote\" learned an option to show which branch the remote\n repository advertises as its primary by pointing its HEAD at.\n\n * New http.proxyAuthMethod configuration variable can be used to\n specify what authentication method to use, as a way to work around\n proxies that do not give error response expected by libcurl when\n CURLAUTH_ANY is used. Also, the codepath for proxy authentication\n has been taught to use credential API to store the authentication\n material in user's keyrings.\n\n * Update the untracked cache subsystem and change its primary UI from\n \"git update-index\" to \"git config\".\n\n * There were a few \"now I am doing this thing\" progress messages in\n the TCP connection code that can be triggered by setting a verbose\n option internally in the code, but \"git fetch -v\" and friends never\n passed the verbose option down to that codepath.\n\n * Clean/smudge filters defined in a configuration file of lower\n precedence can now be overridden to be a pass-through no-op by\n setting the variable to an empty string.\n\n * A new \"^{/!-}\" notation can be used to name a\n commit that is reachable from that does not match the\n given .\n\n * The \"user.useConfigOnly\" configuration variable can be used to\n force the user to always set user.email & user.name configuration\n variables, serving as a reminder for those who work on multiple\n projects and do not want to put these in their $HOME/.gitconfig.\n\n * \"git fetch\" and friends that make network connections can now be\n told to only use ipv4 (or ipv6).\n\n * Some authentication methods do not need username or password, but\n libcurl needs some hint that it needs to perform authentication.\n Supplying an empty username and password string is a valid way to\n do so, but you can set the http.[.]emptyAuth configuration\n variable to achieve the same, if you find it cleaner.\n\n * You can now set http.[.]pinnedpubkey to specify the pinned\n public key when building with recent enough versions of libcURL.\n\n * The configuration system has been taught to phrase where it found a\n bad configuration variable in a better way in its error messages.\n \"git config\" learnt a new \"--show-origin\" option to indicate where\n the values come from.\n\n * The \"credential-cache\" daemon process used to run in whatever\n directory it happened to start in, but this made umount(2)ing the\n filesystem that houses the repository harder; now the process\n chdir()s to the directory that house its own socket on startup.\n\n * When \"git submodule update\" did not result in fetching the commit\n object in the submodule that is referenced by the superproject, the\n command learned to retry another fetch, specifically asking for\n that commit that may not be connected to the refs it usually\n fetches.\n\n * \"git merge-recursive\" learned \"--no-renames\" option to disable its\n rename detection logic.\n\n * Across the transition at around Git version 2.0, the user used to\n get a pretty loud warning when running \"git push\" without setting\n push.default configuration variable. We no longer warn, given that\n the transition is over long time ago.\n\n * README has been renamed to README.md and its contents got tweaked\n slightly to make it easier on the eyes.\n\n\nPerformance, Internal Implementation, Development Support etc.\n\n * Add a framework to spawn a group of processes in parallel, and use\n it to run \"git fetch --recurse-submodules\" in parallel.\n\n * A slight update to the Makefile to mark \"phoney\" targets\n as such correctly.\n\n * In-core storage of the reverse index for .pack files (which lets\n you go from a pack offset to an object name) has been streamlined.\n\n * d95138e6 (setup: set env $GIT_WORK_TREE when work tree is set, like\n $GIT_DIR, 2015-06-26) attempted to work around a glitch in alias\n handling by overwriting GIT_WORK_TREE environment variable to\n affect subprocesses when set_git_work_tree() gets called, which\n resulted in a rather unpleasant regression to \"clone\" and \"init\".\n Try to address the same issue by always restoring the environment\n and respawning the real underlying command when handling alias.\n\n * The low-level code that is used to create symbolic references has\n been updated to share more code with the code that deals with\n normal references.\n\n * strbuf_getline() and friends have been redefined to make it easier\n to identify which callsite of (new) strbuf_getline_lf() should\n allow and silently ignore carriage-return at the end of the line to\n help users on DOSsy systems.\n\n * \"git shortlog\" used to accumulate various pieces of information\n regardless of what was asked to be shown in the final output. It\n has been optimized by noticing what need not to be collected\n (e.g. there is no need to collect the log messages when showing\n only the number of changes).\n\n * \"git checkout $branch\" (and other operations that share the same\n underlying machinery) has been optimized.\n\n * Automated tests in Travis CI environment has been optimized by\n persisting runtime statistics of previous \"prove\" run, executing\n tests that take longer before other ones; this reduces the total\n wallclock time.\n\n * Test scripts have been updated to remove assumptions that are not\n portable between Git for POSIX and Git for Windows, or to skip ones\n with expectations that are not satisfiable on Git for Windows.\n\n * Some calls to strcpy(3) triggers a false warning from static\n analysers that are less intelligent than humans, and reducing the\n number of these false hits helps us notice real issues. A few\n calls to strcpy(3) in test-path-utils that are already safe has\n been rewritten to avoid false wanings.\n\n * Some calls to strcpy(3) triggers a false warning from static\n analysers that are less intelligent than humans, and reducing the\n number of these false hits helps us notice real issues. A few\n calls to strcpy(3) in \"git rerere\" that are already safe has been\n rewritten to avoid false wanings.\n\n * The \"name_path\" API was an attempt to reduce the need to construct\n the full path out of a series of path components while walking a\n tree hierarchy, but over time made less efficient because the path\n needs to be flattened, e.g. to be compared with another path that\n is already flat. The API has been removed and its users have been\n rewritten to simplify the overall code complexity.\n\n * Help those who debug http(s) part of the system.\n (merge 0054045 sp/remote-curl-ssl-strerror later to maint).\n\n * The internal API to interact with \"remote.*\" configuration\n variables has been streamlined.\n\n * The ref-filter's format-parsing code has been refactored, in\n preparation for \"branch --format\" and friends.\n\n * Traditionally, the tests that try commands that work on the\n contents in the working tree were named with \"worktree\" in their\n filenames, but with the recent addition of \"git worktree\"\n subcommand, whose tests are also named similarly, it has become\n harder to tell them apart. The traditional tests have been renamed\n to use \"work-tree\" instead in an attempt to differentiate them.\n (merge 5549029 mg/work-tree-tests later to maint).\n\n * Many codepaths forget to check return value from git_config_set();\n the function is made to die() to make sure we do not proceed when\n setting a configuration variable failed.\n (merge 3d18064 ps/config-error later to maint).\n\n * Handling of errors while writing into our internal asynchronous\n process has been made more robust, which reduces flakiness in our\n tests.\n (merge 43f3afc jk/epipe-in-async later to maint).\n\n * There is a new DEVELOPER knob that enables many compiler warning\n options in the Makefile.\n\n * The way the test scripts configure the Apache web server has been\n updated to work also for Apache 2.4 running on RedHat derived\n distros.\n\n * Out of maintenance gcc on OSX 10.6 fails to compile the code in\n 'master'; work it around by using clang by default on the platform.\n\n\nAlso contains various documentation updates and code clean-ups.\n\n\nFixes since v2.7\n----------------\n\nUnless otherwise noted, all the fixes since v2.7 in the maintenance\ntrack are contained in this release (see the maintenance releases'\nnotes for details).\n\n * An earlier change in 2.5.x-era broke users' hooks and aliases by\n exporting GIT_WORK_TREE to point at the root of the working tree,\n interfering when they tried to use a different working tree without\n setting GIT_WORK_TREE environment themselves.\n\n * The \"exclude_list\" structure has the usual \"alloc, nr\" pair of\n fields to be used by ALLOC_GROW(), but clear_exclude_list() forgot\n to reset 'alloc' to 0 when it cleared 'nr' to discard the managed\n array.\n\n * Paths that have been told the index about with \"add -N\" are not\n quite yet in the index, but a few commands behaved as if they\n already are in a harmful way.\n\n * \"git send-email\" was confused by escaped quotes stored in the alias\n files saved by \"mutt\", which has been corrected.\n\n * A few unportable C construct have been spotted by clang compiler\n and have been fixed.\n\n * The documentation has been updated to hint the connection between\n the '--signoff' option and DCO.\n\n * \"git reflog\" incorrectly assumed that all objects that used to be\n at the tip of a ref must be commits, which caused it to segfault.\n\n * The ignore mechanism saw a few regressions around untracked file\n listing and sparse checkout selection areas in 2.7.0; the change\n that is responsible for the regression has been reverted.\n\n * Another try to improve the ignore mechanism that lets you say \"this\n is excluded\" and then later say \"oh, no, this part (that is a\n subset of the previous part) is not excluded\". This has still a\n known limitation, though.\n\n * Some codepaths used fopen(3) when opening a fixed path in $GIT_DIR\n (e.g. COMMIT_EDITMSG) that is meant to be left after the command is\n done. This however did not work well if the repository is set to\n be shared with core.sharedRepository and the umask of the previous\n user is tighter. They have been made to work better by calling\n unlink(2) and retrying after fopen(3) fails with EPERM.\n\n * Asking gitweb for a nonexistent commit left a warning in the server\n log.\n\n Somebody may want to follow this up with an additional test, perhaps?\n IIRC, we do test that no Perl warnings are given to the server log,\n so this should have been caught if our test coverage were good.\n\n * \"git rebase\", unlike all other callers of \"gc --auto\", did not\n ignore the exit code from \"gc --auto\".\n\n * Many codepaths that run \"gc --auto\" before exiting kept packfiles\n mapped and left the file descriptors to them open, which was not\n friendly to systems that cannot remove files that are open. They\n now close the packs before doing so.\n\n * A recent optimization to filter-branch in v2.7.0 introduced a\n regression when --prune-empty filter is used, which has been\n corrected.\n\n * The description for SANITY prerequisite the test suite uses has\n been clarified both in the comment and in the implementation.\n\n * \"git tag\" started listing a tag \"foo\" as \"tags/foo\" when a branch\n named \"foo\" exists in the same repository; remove this unnecessary\n disambiguation, which is a regression introduced in v2.7.0.\n\n * The way \"git svn\" uses auth parameter was broken by Subversion\n 1.9.0 and later.\n\n * The \"split\" subcommand of \"git subtree\" (in contrib/) incorrectly\n skipped merges when it shouldn't, which was corrected.\n\n * A few options of \"git diff\" did not work well when the command was\n run from a subdirectory.\n\n * The command line completion learned a handful of additional options\n and command specific syntax.\n\n * dirname() emulation has been added, as Msys2 lacks it.\n\n * The underlying machinery used by \"ls-files -o\" and other commands\n have been taught not to create empty submodule ref cache for a\n directory that is not a submodule. This removes a ton of wasted\n CPU cycles.\n\n * \"git worktree\" had a broken code that attempted to auto-fix\n possible inconsistency that results from end-users moving a\n worktree to different places without telling Git (the original\n repository needs to maintain backpointers to its worktrees, but\n \"mv\" run by end-users who are not familiar with that fact will\n obviously not adjust them), which actually made things worse\n when triggered.\n\n * The low-level merge machinery has been taught to use CRLF line\n termination when inserting conflict markers to merged contents that\n are themselves CRLF line-terminated.\n\n * \"git push --force-with-lease\" has been taught to report if the push\n needed to force (or fast-forwarded).\n\n * The emulated \"yes\" command used in our test scripts has been\n tweaked not to spend too much time generating unnecessary output\n that is not used, to help those who test on Windows where it would\n not stop until it fills the pipe buffer due to lack of SIGPIPE.\n\n * The documentation for \"git clean\" has been corrected; it mentioned\n that .git/modules/* are removed by giving two \"-f\", which has never\n been the case.\n\n * The vimdiff backend for \"git mergetool\" has been tweaked to arrange\n and number buffers in the order that would match the expectation of\n majority of people who read left to right, then top down and assign\n buffers 1 2 3 4 \"mentally\" to local base remote merge windows based\n on that order.\n\n * \"git show 'HEAD:Foo[BAR]Baz'\" did not interpret the argument as a\n rev, i.e. the object named by the the pathname with wildcard\n characters in a tree object.\n (merge aac4fac nd/dwim-wildcards-as-pathspecs later to maint).\n\n * \"git rev-parse --git-common-dir\" used in the worktree feature\n misbehaved when run from a subdirectory.\n (merge 17f1365 nd/git-common-dir-fix later to maint).\n\n * \"git worktree add -B \" did not work.\n\n * The \"v(iew)\" subcommand of the interactive \"git am -i\" command was\n broken in 2.6.0 timeframe when the command was rewritten in C.\n (merge 708b8cc jc/am-i-v-fix later to maint).\n\n * \"git merge-tree\" used to mishandle \"both sides added\" conflict with\n its own \"create a fake ancestor file that has the common parts of\n what both sides have added and do a 3-way merge\" logic; this has\n been updated to use the usual \"3-way merge with an empty blob as\n the fake common ancestor file\" approach used in the rest of the\n system.\n (merge 907681e jk/no-diff-emit-common later to maint).\n\n * The memory ownership rule of fill_textconv() API, which was a bit\n tricky, has been documented a bit better.\n (merge a64e6a4 jk/more-comments-on-textconv later to maint).\n\n * Update various codepaths to avoid manually-counted malloc().\n (merge 08c95df jk/tighten-alloc later to maint).\n\n * The documentation did not clearly state that the 'simple' mode is\n now the default for \"git push\" when push.default configuration is\n not set.\n (merge f6b1fb3 mm/push-simple-doc later to maint).\n\n * Recent versions of GNU grep are pickier when their input contains\n arbitrary binary data, which some of our tests uses. Rewrite the\n tests to sidestep the problem.\n (merge 3b1442d jk/grep-binary-workaround-in-test later to maint).\n\n * A helper function \"git submodule\" uses since v2.7.0 to list the\n modules that match the pathspec argument given to its subcommands\n (e.g. \"submodule add \") has been fixed.\n (merge 2b56bb7 sb/submodule-module-list-fix later to maint).\n\n * \"git config section.var value\" to set a value in per-repository\n configuration file failed when it was run outside any repository,\n but didn't say the reason correctly.\n (merge 638fa62 js/config-set-in-non-repository later to maint).\n\n * The code to read the pack data using the offsets stored in the pack\n idx file has been made more carefully check the validity of the\n data in the idx.\n (merge 7465feb jk/pack-idx-corruption-safety later to maint).\n\n * Other minor clean-ups and documentation updates\n (merge f459823 ak/extract-argv0-last-dir-sep later to maint).\n (merge 63ca1c0 ak/git-strip-extension-from-dashed-command later to maint).\n (merge 4867f11 ps/plug-xdl-merge-leak later to maint).\n (merge 4938686 dt/initial-ref-xn-commit-doc later to maint).\n (merge 9537f21 ma/update-hooks-sample-typofix later to maint).\n\n----------------------------------------------------------------\n\nChanges since v2.7.0 are as follows:\n\nAlex Henrie (1):\n stripspace: call U+0020 a \"space\" instead of a \"blank\"\n\nAlexander Kuleshov (3):\n format-patch: introduce format.outputDirectory configuration\n exec_cmd.c: use find_last_dir_sep() for code simplification\n git.c: simplify stripping extension of a file in handle_builtin()\n\nAndrew Wheeler (1):\n push: fix ref status reporting for --force-with-lease\n\nChangwoo Ryu (1):\n l10n: ko.po: Add Korean translation\n\nChristian Couder (11):\n dir: free untracked cache when removing it\n update-index: use enum for untracked cache options\n update-index: add --test-untracked-cache\n update-index: add untracked cache notifications\n update-index: move 'uc' var declaration\n dir: add {new,add}_untracked_cache()\n dir: add remove_untracked_cache()\n dir: simplify untracked cache \"ident\" field\n config: add core.untrackedCache\n test-dump-untracked-cache: don't modify the untracked cache\n t7063: add tests for core.untrackedCache\n\nChristoph Egger (1):\n http: implement public key pinning\n\nDan Aloni (1):\n ident: add user.useConfigOnly boolean for when ident shouldn't be guessed\n\nDave Ware (1):\n contrib/subtree: fix \"subtree split\" skipped-merge bug\n\nDavid A. Greene (1):\n contrib/subtree: Make testing easier\n\nDavid A. Wheeler (1):\n Expand documentation describing --signoff\n\nDavid Turner (3):\n do_compare_entry: use already-computed path\n unpack-trees: fix accidentally quadratic behavior\n refs: document transaction semantics\n\nDennis Kaarsemaker (1):\n reflog-walk: don't segfault on non-commit sha1's in the reflog\n\nDickson Wong (1):\n mergetool: reorder vim/gvim buffers in three-way diffs\n\nEdmundo Carmona Antoranz (1):\n blame: add support for --[no-]progress option\n\nElia Pinto (92):\n Makefile: add missing phony target\n contrib/examples/git-commit.sh: use the $( ... ) construct for command substitution\n contrib/examples/git-fetch.sh: use the $( ... ) construct for command substitution\n contrib/examples/git-merge.sh: use the $( ... ) construct for command substitution\n contrib/examples/git-repack.sh: use the $( ... ) construct for command substitution\n contrib/examples/git-revert.sh: use the $( ... ) construct for command substitution\n contrib/thunderbird-patch-inline/appp.sh: use the $( ... ) construct for command substitution\n git-gui/po/glossary/txt-to-pot.sh: use the $( ... ) construct for command substitution\n t/lib-httpd.sh: use the $( ... ) construct for command substitution\n test-sha1.sh: use the $( ... ) construct for command substitution\n unimplemented.sh: use the $( ... ) construct for command substitution\n t/t1100-commit-tree-options.sh: use the $( ... ) construct for command substitution\n t/t1401-symbolic-ref.sh: use the $( ... ) construct for command substitution\n t/t1410-reflog.sh: use the $( ... ) construct for command substitution\n t/t1511-rev-parse-caret.sh: use the $( ... ) construct for command substitution\n t/t1512-rev-parse-disambiguation.sh: use the $( ... ) construct for command substitution\n t/t1700-split-index.sh: use the $( ... ) construct for command substitution\n t/t2025-worktree-add.sh: use the $( ... ) construct for command substitution\n t/t2102-update-index-symlinks.sh: use the $( ... ) construct for command substitution\n t/t3030-merge-recursive.sh: use the $( ... ) construct for command substitution\n t/t3100-ls-tree-restrict.sh: use the $( ... ) construct for command substitution\n t/t3101-ls-tree-dirname.sh: use the $( ... ) construct for command substitution\n t/t3210-pack-refs.sh: use the $( ... ) construct for command substitution\n t/t3403-rebase-skip.sh: use the $( ... ) construct for command substitution\n t/t3511-cherry-pick-x.sh: use the $( ... ) construct for command substitution\n t/t3600-rm.sh: use the $( ... ) construct for command substitution\n t/t3700-add.sh: use the $( ... ) construct for command substitution\n t/t5100-mailinfo.sh: use the $( ... ) construct for command substitution\n t/t5300-pack-object.sh: use the $( ... ) construct for command substitution\n t/t5301-sliding-window.sh: use the $( ... ) construct for command substitution\n t/t5302-pack-index.sh: use the $( ... ) construct for command substitution\n t/t5303-pack-corruption-resilience.sh: use the $( ... ) construct for command substitution\n t/t5304-prune.sh: use the $( ... ) construct for command substitution\n t/t5305-include-tag.sh: use the $( ... ) construct for command substitution\n t/t5500-fetch-pack.sh: use the $( ... ) construct for command substitution\n t/t5505-remote.sh: use the $( ... ) construct for command substitution\n t/t5506-remote-groups.sh: use the $( ... ) construct for command substitution\n t/t5510-fetch.sh: use the $( ... ) construct for command substitution\n t/t5515-fetch-merge-logic.sh: use the $( ... ) construct for command substitution\n t/t5516-fetch-push.sh: use the $( ... ) construct for command substitution\n t/t5517-push-mirror.sh: use the $( ... ) construct for command substitution\n t/t5522-pull-symlink.sh: use the $( ... ) construct for command substitution\n t/t5530-upload-pack-error.sh: use the $( ... ) construct for command substitution\n t/t5532-fetch-proxy.sh: use the $( ... ) construct for command substitution\n t/t5537-fetch-shallow.sh: use the $( ... ) construct for command substitution\n t/t5538-push-shallow.sh: use the $( ... ) construct for command substitution\n t/t5550-http-fetch-dumb.sh: use the $( ... ) construct for command substitution\n t/t5570-git-daemon.sh: use the $( ... ) construct for command substitution\n t/t5601-clone.sh: use the $( ... ) construct for command substitution\n t/t5700-clone-reference.sh: use the $( ... ) construct for command substitution\n t/t5710-info-alternate.sh: use the $( ... ) construct for command substitution\n t/t5900-repo-selection.sh: use the $( ... ) construct for command substitution\n t/t6001-rev-list-graft.sh: use the $( ... ) construct for command substitution\n t/t6002-rev-list-bisect.sh: use the $( ... ) construct for command substitution\n t/t6015-rev-list-show-all-parents.sh: use the $( ... ) construct for command substitution\n t/t6032-merge-large-rename.sh: use the $( ... ) construct for command substitution\n t/t6132-pathspec-exclude.sh: use the $( ... ) construct for command substitution\n t/t7001-mv.sh: use the $( ... ) construct for command substitution\n t/t7003-filter-branch.sh: use the $( ... ) construct for command substitution\n t/t7004-tag.sh: use the $( ... ) construct for command substitution\n t/t7006-pager.sh: use the $( ... ) construct for command substitution\n t/t7103-reset-bare.sh: use the $( ... ) construct for command substitution\n t/t7406-submodule-update.sh: use the $( ... ) construct for command substitution\n t/t7408-submodule-reference.sh: use the $( ... ) construct for command substitution\n t/t7504-commit-msg-hook.sh: use the $( ... ) construct for command substitution\n t/t7505-prepare-commit-msg-hook.sh: use the $( ... ) construct for command substitution\n t/t7602-merge-octopus-many.sh: use the $( ... ) construct for command substitution\n t/t7700-repack.sh: use the $( ... ) construct for command substitution\n t/t8003-blame-corner-cases.sh: use the $( ... ) construct for command substitution\n t/t9001-send-email.sh: use the $( ... ) construct for command substitution\n t9100-git-svn-basic.sh: use the $( ... ) construct for command substitution\n t9101-git-svn-props.sh: use the $( ... ) construct for command substitution\n t9104-git-svn-follow-parent.sh: use the $( ... ) construct for command substitution\n t9105-git-svn-commit-diff.sh: use the $( ... ) construct for command substitution\n t9107-git-svn-migrate.sh: use the $( ... ) construct for command substitution\n t9108-git-svn-glob.sh: use the $( ... ) construct for command substitution\n t9109-git-svn-multi-glob.sh: use the $( ... ) construct for command substitution\n t9110-git-svn-use-svm-props.sh: use the $( ... ) construct for command substitution\n t9114-git-svn-dcommit-merge.sh: use the $( ... ) construct for command substitution\n t9118-git-svn-funky-branch-names.sh: use the $( ... ) construct for command substitution\n t9119-git-svn-info.sh: use the $( ... ) construct for command substitution\n t9129-git-svn-i18n-commitencoding.sh: use the $( ... ) construct for command substitution\n t9130-git-svn-authors-file.sh: use the $( ... ) construct for command substitution\n t9132-git-svn-broken-symlink.sh: use the $( ... ) construct for command substitution\n t9137-git-svn-dcommit-clobber-series.sh: use the $( ... ) construct for command substitution\n t9138-git-svn-authors-prog.sh: use the $( ... ) construct for command substitution\n t9145-git-svn-master-branch.sh: use the $( ... ) construct for command substitution\n t9150-svk-mergetickets.sh: use the $( ... ) construct for command substitution\n t9300-fast-import.sh: use the $( ... ) construct for command substitution\n t9350-fast-export.sh: use the $( ... ) construct for command substitution\n t9501-gitweb-standalone-http-status.sh: use the $( ... ) construct for command substitution\n t9901-git-web--browse.sh: use the $( ... ) construct for command substitution\n\nEric Wong (8):\n git-send-email: do not double-escape quotes from mutt\n for-each-ref: document `creatordate` and `creator` fields\n git-svn: fix auth parameter handling on SVN 1.9.0+\n pass transport verbosity down to git_connect\n connect & http: support -4 and -6 switches for remote operations\n t5570: add tests for \"git {clone,fetch,pull} -v\"\n git-svn: hoist out utf8 prep from t9129 to lib-git-svn\n tests: remove no-op full-svn-test target\n\nFelipe Gonçalves Assis (7):\n merge-recursive: option to disable renames\n merge-recursive: more consistent interface\n merge-strategies.txt: fix typo\n merge-recursive: find-renames resets threshold\n t3034: add rename threshold tests\n t3034: test option to disable renames\n t3034: test deprecated interface\n\nGyuYong Jung (1):\n git-cvsserver.perl: fix typo\n\nJacob Keller (1):\n notes: allow merging from arbitrary references\n\nJeff King (79):\n pack-revindex: drop hash table\n pack-revindex: store entries directly in packed_git\n create_symref: modernize variable names\n create_symref: use existing ref-lock code\n create_symref: write reflog while holding lock\n run-command: don't warn on SIGPIPE deaths\n avoid shifting signed integers 31 bits\n bswap: add NO_UNALIGNED_LOADS define\n checkout,clone: check return value of create_symref\n lock_ref_sha1_basic: always fill old_oid while holding lock\n lock_ref_sha1_basic: handle REF_NODEREF with invalid refs\n rebase: ignore failures from \"gc --auto\"\n shortlog: match both \"Author:\" and \"author\" on stdin\n shortlog: use strbufs to read from stdin\n shortlog: replace hand-parsing of author with pretty-printer\n shortlog: optimize \"--summary\" mode\n shortlog: optimize out useless \"\" normalization\n shortlog: optimize out useless string list\n shortlog: don't warn on empty author\n filter-branch: resolve $commit^{tree} in no-index case\n clean: make is_git_repository a public function\n resolve_gitlink_ref: ignore non-repository paths\n t6300: use test_atom for some un-modern tests\n tag: do not show ambiguous tag names as \"tags/foo\"\n transport: drop support for git-over-rsync\n give \"nbuf\" strbuf a more meaningful name\n checkout-index: simplify \"-z\" option parsing\n checkout-index: handle \"--no-prefix\" option\n checkout-index: handle \"--no-index\" option\n checkout-index: disallow \"--no-stage\" option\n apply, ls-files: simplify \"-z\" parsing\n fmt_ident: refactor strictness checks\n test-path-utils: use xsnprintf in favor of strcpy\n rerere: replace strcpy with xsnprintf\n checkout: reorder check_filename conditional\n check_filename: tighten dwim-wildcard ambiguity\n get_sha1: don't die() on bogus search strings\n http-push: stop using name_path\n show_object_with_name: simplify by using path_name()\n list-objects: convert name_path to a strbuf\n list-objects: drop name_path entirely\n list-objects: pass full pathname to callbacks\n git-config: better document default behavior for `--include`\n ref-filter: use string_list_split over strbuf_split\n reflog_expire_cfg: NUL-terminate pattern field\n add helpers for detecting size_t overflow\n tree-diff: catch integer overflow in combine_diff_path allocation\n diff: clarify textconv interface\n harden REALLOC_ARRAY and xcalloc against size_t overflow\n add helpers for allocating flex-array structs\n argv-array: add detach function\n convert manual allocations to argv_array\n convert trivial cases to ALLOC_ARRAY\n use xmallocz to avoid size arithmetic\n convert trivial cases to FLEX_ARRAY macros\n use st_add and st_mult for allocation size computation\n prepare_{git,shell}_cmd: use argv_array\n write_untracked_extension: use FLEX_ALLOC helper\n fast-import: simplify allocation in start_packfile\n fetch-pack: simplify add_sought_entry\n test-path-utils: fix normalize_path_copy output buffer size\n sequencer: simplify memory allocation of get_message\n git-compat-util: drop mempcpy compat code\n transport_anonymize_url: use xstrfmt\n diff_populate_gitlink: use a strbuf\n convert ewah/bitmap code to use xmalloc\n ewah: convert to REALLOC_ARRAY, etc\n merge-one-file: use empty blob for add/add base\n merge-tree: drop generate_common strategy\n xdiff: drop XDL_EMIT_COMMON\n t5313: test bounds-checks of corrupted/malicious pack/idx files\n nth_packed_object_offset: bounds-check extended offset\n use_pack: handle signed off_t overflow\n write_or_die: handle EPIPE in async threads\n fetch-pack: ignore SIGPIPE in sideband demuxer\n test_must_fail: report number of unexpected signal\n t5504: handle expected output from SIGPIPE death\n compat/mingw: brown paper bag fix for 50a6c8e\n t9700: fix test for perl older than 5.14\n\nJiang Xin (1):\n http: honor no_http env variable to bypass proxy\n\nJohannes Schindelin (47):\n commit: allow editing the commit message even in shared repos\n Handle more file writes correctly in shared repos\n Refactor skipping DOS drive prefixes\n compat/basename: make basename() conform to POSIX\n compat/basename.c: provide a dirname() compatibility function\n t0060: verify that basename() and dirname() work as expected\n config.mak.uname: support MSys2\n config.mak.uname: supporting 64-bit MSys2\n fetch: release pack files before garbage-collecting\n am: release pack files before garbage-collecting\n merge: release pack files before garbage-collecting\n receive-pack: release pack files before garbage-collecting\n pull: allow interactive rebase with --rebase=interactive\n remote: handle the config setting branch.*.rebase=interactive\n completion: add missing branch.*.rebase values\n nedmalloc: allow compiling with MSys2's compiler\n compat/mingw: support MSys2-based MinGW build\n compat/winansi: support compiling with MSys2\n t0060: loosen overly strict expectations\n mingw: avoid redefining S_* constants\n mingw: avoid warnings when casting HANDLEs to int\n mingw: squash another warning about a cast\n mingw: uglify (a, 0) definitions to shut up warnings\n mingw: let's use gettext with MSYS2\n mingw: do not trust MSYS2's MinGW gettext.sh\n Git.pm: stop assuming that absolute paths start with a slash\n mingw: prepare the TMPDIR environment variable for shell scripts\n mingw: let lstat() fail with errno == ENOTDIR when appropriate\n merge-file: let conflict markers match end-of-line style of the context\n merge-file: ensure that conflict sections match eol style\n mingw: fix t5601-clone.sh\n mingw: accomodate t0060-path-utils for MSYS2\n mingw: disable mkfifo-based tests\n tests: turn off git-daemon tests if FIFOs are not available\n mingw: skip test in t1508 that fails due to path conversion\n mingw: fix t9700's assumption about directory separators\n mingw: work around pwd issues in the tests\n mingw: mark t9100's test cases with appropriate prereqs\n mingw: avoid illegal filename in t9118\n mingw: handle the missing POSIXPERM prereq in t9124\n mingw: skip a test in t9130 that cannot pass on Windows\n mingw: do not bother to test funny file names\n test-lib: limit the output of the yes utility\n gitignore: ignore generated test-fake-ssh executable\n t5505: 'remote add x y' should work when url.y.insteadOf = x\n git config: report when trying to modify a non-existing repo config\n Mark win32's pthread_exit() as NORETURN\n\nJohannes Sixt (3):\n t/t5100: no need to use 'echo' command substitutions for globbing\n mingw: avoid linking to the C library's isalpha()\n t0001: fix GIT_* environment variable check under --valgrind\n\nJohn Keeping (3):\n completion: add missing git-rebase options\n t8005: avoid grep on non-ASCII data\n t9200: avoid grep on non-ASCII data\n\nJon Griffiths (3):\n credential-cache--daemon: refactor check_socket_directory\n credential-cache--daemon: disallow relative socket path\n credential-cache--daemon: change to the socket dir on startup\n\nJonathan Nieder (1):\n submodule.c: write \"Fetching submodule \" to stderr\n\nJunio C Hamano (46):\n First batch for post 2.7 cycle\n strbuf: miniscule style fix\n strbuf: make strbuf_getline_crlf() global\n strbuf: introduce strbuf_getline_{lf,nul}()\n mktree: there are only two possible line terminations\n check-attr: there are only two possible line terminations\n check-ignore: there are only two possible line terminations\n update-index: there are only two possible line terminations\n checkout-index: there are only two possible line terminations\n strbuf: give strbuf_getline() to the \"most text friendly\" variant\n hash-object: read --stdin-paths with strbuf_getline()\n revision: read --stdin with strbuf_getline()\n rev-parse: read parseopt spec with strbuf_getline()\n ident.c: read /etc/mailname with strbuf_getline()\n remote.c: read $GIT_DIR/remotes/* with strbuf_getline()\n clone/sha1_file: read info/alternates with strbuf_getline()\n transport-helper: read helper response with strbuf_getline()\n cat-file: read batch stream with strbuf_getline()\n column: read lines with strbuf_getline()\n send-pack: read list of refs with strbuf_getline()\n grep: read -f file with strbuf_getline()\n test-sha1-array: read command stream with strbuf_getline()\n test-lib: clarify and tighten SANITY\n Second batch for 2.8 cycle\n Third batch for 2.8 cycle\n git: remove an early return from save_env_before_alias()\n git: protect against unbalanced calls to {save,restore}_env()\n git: simplify environment save/restore logic\n Fourth batch for 2.8.cycle\n Getting closer to 2.7.1\n restore_env(): free the saved environment variable once we are done\n Fifth batch for 2.8 cycle\n Git 2.7.1\n Sixth batch for the 2.8 cycle\n pager: lose a separate argv[]\n pager: factor out a helper to prepare a child process to run the pager\n am -i: fix \"v\"iew\n Start preparing for 2.7.2\n Seventh batch for the 2.8 cycle\n Git 2.7.2\n Eighth batch for 2.8\n Git 2.8-rc0\n Git 2.8-rc1\n gitignore: document that unignoring a directory unignores everything in it\n Git 2.7.3\n Git 2.8-rc2\n\nKarsten Blees (1):\n mingw: factor out Windows specific environment setup\n\nKarthik Nayak (10):\n ref-filter: bump 'used_atom' and related code to the top\n ref-filter: introduce struct used_atom\n ref-filter: introduce parsing functions for each valid atom\n ref-filter: introduce color_atom_parser()\n ref-filter: introduce parse_align_position()\n ref-filter: introduce align_atom_parser()\n ref-filter: align: introduce long-form syntax\n ref-filter: introduce remote_ref_atom_parser()\n ref-filter: introduce contents_atom_parser()\n ref-filter: introduce objectname_atom_parser()\n\nKazutoshi Satoda (2):\n git-svn: enable \"svn.pathnameencoding\" on dcommit\n git-svn: apply \"svn.pathnameencoding\" before URL encoding\n\nKnut Franke (2):\n http: allow selection of proxy authentication method\n http: use credential API to handle proxy authentication\n\nLars Schneider (8):\n travis-ci: run previously failed tests first, then slowest to fastest\n travis-ci: explicity use container-based infrastructure\n convert: treat an empty string for clean/smudge filters as \"cat\"\n t: do not hide Git's exit code in tests using 'nul_to_q'\n rename git_config_from_buf to git_config_from_mem\n config: add 'origin_type' to config_source struct\n config: add '--show-origin' option to print the origin of a config value\n add DEVELOPER makefile knob to check for acknowledged warnings\n\nLars Vogel (1):\n git-add doc: do not say working directory when you mean working tree\n\nMartin Amdisen (1):\n templates/hooks: fix minor typo in the sample update-hook\n\nMatt McCutchen (1):\n Documentation/git-clean.txt: don't mention deletion of .git/modules/*\n\nMatthew Kraai (1):\n Documentation: remove unnecessary backslashes\n\nMatthieu Moy (7):\n Documentation/git-push: document that 'simple' is the default\n README: use markdown syntax\n README.md: add hyperlinks on filenames\n README.md: move the link to git-scm.com up\n README.md: don't call git stupid in the title\n README.md: move down historical explanation about the name\n push: remove \"push.default is unset\" warning message\n\nMichael J Gruber (4):\n t9100: fix breakage when SHELL_PATH is not /bin/sh\n tests: rename work-tree tests to *work-tree*\n t/lib-httpd: load mod_unixd\n t5510: do not leave changed cwd\n\nMike Hommey (1):\n notes: allow treeish expressions as notes ref\n\nNguyễn Thái Ngọc Duy (25):\n blame: remove obsolete comment\n add and use a convenience macro ce_intent_to_add()\n Revert \"setup: set env $GIT_WORK_TREE when work tree is set, like $GIT_DIR\"\n git.c: make it clear save_env() is for alias handling only\n setup.c: re-fix d95138e (setup: set env $GIT_WORK_TREE when ..\n git.c: make sure we do not leak GIT_* to alias scripts\n grep: make it clear i-t-a entries are ignored\n dir.c: clean the entire struct in clear_exclude_list()\n Revert \"dir.c: don't exclude whole dir prematurely if neg pattern may match\"\n worktree.c: fix indentation\n diff-no-index: do not take a redundant prefix argument\n diff: make -O and --output work in subdirectory\n worktree: stop supporting moving worktrees manually\n rev-parse: take prefix into account in --git-common-dir\n dir.c: fix match_pathname()\n dir.c: support tracing exclude\n dir.c: support marking some patterns already matched\n dir.c: don't exclude whole dir prematurely\n worktree: fix \"add -B\"\n worktree add -B: do the checkout test before update branch\n sha1_file.c: mark strings for translation\n builtin/checkout.c: mark strings for translation\n builtin/clone.c: mark strings for translation\n ref-filter.c: mark strings for translation\n trailer.c: mark strings for translation\n\nPat Thoyts (1):\n t0008: avoid absolute path\n\nPatrick Steinhardt (18):\n push: add '--delete' flag to synopsis\n push: add '-d' as shorthand for '--delete'\n config: introduce set_or_die wrappers\n branch: report errors in tracking branch setup\n branch: die on config error when unsetting upstream\n branch: die on config error when editing branch description\n submodule: die on config error when linking modules\n submodule--helper: die on config error when cloning module\n remote: die on config error when setting URL\n remote: die on config error when setting/adding branches\n remote: die on config error when manipulating remotes\n clone: die on config error in cmd_clone\n init-db: die on config errors when initializing empty repo\n sequencer: die on config error when saving replay opts\n compat: die when unable to set core.precomposeunicode\n config: rename git_config_set to git_config_set_gently\n config: rename git_config_set_or_die to git_config_set\n xdiff/xmerge: fix memory leak in xdl_merge\n\nPaul Wagland (2):\n completion: complete show-branch \"--date-order\"\n completion: update completion arguments for stash\n\nRob Mayoff (1):\n contrib/subtree: unwrap tag refs\n\nRomain Picard (1):\n git-p4.py: add support for filetype change\n\nSZEDER Gábor (2):\n t6050-replace: make failing editor test more robust\n completion: fix mis-indentation in _git_stash()\n\nSebastian Schuberth (3):\n docs: clarify that passing --depth to git-clone implies --single-branch\n docs: say \"commits\" in the --depth option wording for git-clone\n docs: clarify that --depth for git-fetch works with newly initialized repos\n\nShawn O. Pearce (1):\n remote-curl: include curl_errorstr on SSL setup failures\n\nStefan Beller (10):\n xread: poll on non blocking fds\n strbuf: add strbuf_read_once to read without blocking\n sigchain: add command to pop all common signals\n run-command: add an asynchronous parallel child processor\n fetch_populated_submodules: use new parallel job processing\n submodules: allow parallel fetching, add tests and documentation\n submodule helper list: respect correct path prefix\n submodule: try harder to fetch needed sha1 by direct fetching sha1\n run-command: do not pass child process data into callbacks\n Documentation: reword rebase summary\n\nStephen P. Smith (4):\n user-manual: remove temporary branch entry from todo list\n glossary: define the term shallow clone\n user-manual: add section documenting shallow clones\n user-manual: add addition gitweb information\n\nThomas Ackermann (1):\n documentation: fix some typos\n\nThomas Braun (1):\n completion: complete \"diff --word-diff-regex=\"\n\nThomas Gummerer (11):\n t7810: correct --no-index test\n builtin/grep: add grep.fallbackToNoIndex config\n ls-remote: document --quiet option\n ls-remote: document --refs option\n ls-remote: fix synopsis\n ls-remote: use parse-options api\n ls-remote: add support for showing symrefs\n remote: use parse_config_key\n remote: simplify remote_is_configured()\n remote: actually check if remote exits\n remote: use remote_is_configured() for add and rename\n\nTobias Klauser (2):\n trailer: allow to write to files other than stdout\n interpret-trailers: add option for in-place editing\n\nTorsten Bögershausen (9):\n ls-files: add eol diagnostics\n t0027: add tests for get_stream_filter()\n convert.c: remove unused parameter 'path'\n convert.c: remove input_crlf_action()\n convert.c: use text_eol_is_crlf()\n convert.c: refactor crlf_action\n convert.c: simplify text_stat\n convert.c: correct attr_action()\n config.mak.uname: use clang for Mac OS X 10.6\n\nVictor Leschuk (3):\n grep: allow threading even on a single-core machine\n grep: slight refactoring to the code that disables threading\n grep: add --threads= option and grep.threads configuration\n\nWill Palmer (2):\n test for '!' handling in rev-parse's named commits\n object name: introduce '^{/!-}' notation\n\nbrian m. carlson (1):\n http: add option to try authentication without username\n\nØyvind A. Holm (1):\n gitweb: squelch \"uninitialized value\" warning\n\n마누엘 (1):\n mingw: try to delete target directory before renaming\n"},{"id":"280660","messageId":"006DA180-E4B6-404E-9377-F2BB41F65591@quendi.de","threadId":"41669","inReplyTo":"xmqqr3fiaq9f.fsf@gitster.mtv.corp.google.com","subject":"Re: [ANNOUNCE] Git v2.8.0-rc2","fromName":"Max Horn","fromEmail":"max@quendi.de","sentAt":"2016-03-12T09:11:15Z","receivedAt":"2016-03-12T09:11:15Z","isPatch":false,"sender":{"key":"max@quendi.de","avatar":"https://avatars.githubusercontent.com/u/241512?v=4"},"body":"\n> On 11 Mar 2016, at 00:04, Junio C Hamano wrote:\n> \n> A release candidate Git v2.8.0-rc2 is now available for testing\n> at the usual places. It is comprised of 459 non-merge commits\n> since v2.7.0, contributed by 60 people, 19 of which are new faces.\n> \n[...]\n\n> Updates since v2.7\n> ------------------\n> \n> UI, Workflows & Features\n> \n> * It turns out \"git clone\" over rsync transport has been broken when\n> the source repository has packed references for a long time, and\n> nobody noticed nor complained about it.\n> \n> * \"branch --delete\" has \"branch -d\" but \"push --delete\" does not.\n\nThis states a problem, but not whether (and how) it was resolved?\n\n[...]\n\n> * Across the transition at around Git version 2.0, the user used to\n> get a pretty loud warning when running \"git push\" without setting\n> push.default configuration variable. We no longer warn, given that\n> the transition is over long time ago.\n\nThat last sentence sounds weird... perhaps \"the transition was completed a long time ago\" ? Or \"the transition ended a long time ago\" ?\n\n> \n> * README has been renamed to README.md and its contents got tweaked\n> slightly to make it easier on the eyes.\n> \n> \n> Performance, Internal Implementation, Development Support etc.\n> \n> * Add a framework to spawn a group of processes in parallel, and use\n> it to run \"git fetch --recurse-submodules\" in parallel.\n> \n> * A slight update to the Makefile to mark \"phoney\" targets\n> as such correctly.\n\nphoney -> phony ? (it seems in british english you can write \"phoney\", but according to a quickly Google search this is rarely used in relation to Makefiles; no surprise, given that the corresponding syntax element is \".PHONY\" and you are not allowed to spell it differently ;-)\n\n[...]\n\n> \n> * Some calls to strcpy(3) triggers a false warning from static\n> analysers that are less intelligent than humans, and reducing the\n> number of these false hits helps us notice real issues. A few\n> calls to strcpy(3) in test-path-utils that are already safe has\n> been rewritten to avoid false wanings.\n> \n> * Some calls to strcpy(3) triggers a false warning from static\n> analysers that are less intelligent than humans, and reducing the\n> number of these false hits helps us notice real issues. A few\n> calls to strcpy(3) in \"git rerere\" that are already safe has been\n> rewritten to avoid false wanings.\n\nThe above two messages are very similar, only the end differs a bit. Perhaps\nmerge them into one? I.e.. \"A few calls to strcpy(3) in test-path-utils and \"git rerere\" that [...]\"\n\nAlso: wanings -> warnings\n\n[...]\n\n> \n> * Asking gitweb for a nonexistent commit left a warning in the server\n> log.\n> \n> Somebody may want to follow this up with an additional test, perhaps?\n> IIRC, we do test that no Perl warnings are given to the server log,\n> so this should have been caught if our test coverage were good.\n\nThat last paragraph seems odd for a changelog?\n\n[...]\n\n> * The underlying machinery used by \"ls-files -o\" and other commands\n> have been taught not to create empty submodule ref cache for a\n\nhave -> has (the machinery is singular)\n\n\n\n\n\nCheers,\nMax"},{"id":"280711","messageId":"56E6D8C4.2010205@drmicha.warpmail.net","threadId":"41669","inReplyTo":"xmqqr3fiaq9f.fsf@gitster.mtv.corp.google.com","subject":"Re: [ANNOUNCE] Git v2.8.0-rc2","fromName":"Michael J Gruber","fromEmail":"git@drmicha.warpmail.net","sentAt":"2016-03-14T15:29:08Z","receivedAt":"2016-03-14T15:29:08Z","isPatch":false,"sender":{"key":"git@grubix.eu","avatar":"https://avatars.githubusercontent.com/u/233215?v=4"},"body":"Hi Junio,\n\nHave you pulled git.pot for 2.8.0 already? [1]\n\nMaybe I'm not up-to-date on the release cycle timing regarding l10n, but\nI was trying to fix a mixed translation/non-translation issue, and it\nturned out the reason is probably that it is not in \"git.pot\":\n\n\"behind \" is in git.pot\n\"ahead \" is not in git.pot\n\nNow, I can recreate git.pot locally, of course, but localisations such\nas de.po seem to have those translation commented out because they're\nmissing from git.pot.\n\nBasically, translations don't get proper testing at this point in the\nrelease cycle.\n\nIn fact, there seems to be another weird problem around that string.\nI'll follow up with a patch in a minute.\n\nMichael\n\n[1] http://permalink.gmane.org/gmane.comp.version-control.git/287774\n"},{"id":"280712","messageId":"3b88641beff6b34ae8b86bf353e6fbb8f9e636f1.1457969307.git.git@drmicha.warpmail.net","threadId":"41669","inReplyTo":"56E6D8C4.2010205@drmicha.warpmail.net","subject":"[PATCH] wt-status: allow \"ahead \" to be picked up by l10n","fromName":"Michael J Gruber","fromEmail":"git@drmicha.warpmail.net","sentAt":"2016-03-14T15:30:33Z","receivedAt":"2016-03-14T15:30:33Z","isPatch":true,"sender":{"key":"git@grubix.eu","avatar":"https://avatars.githubusercontent.com/u/233215?v=4"},"body":"The extra pair of parentheses keeps the l10n engine from picking up the\nstring. Remove them so that \"ahead \" ends up in git.pot.\n\nSigned-off-by: Michael J Gruber \n---\n wt-status.c | 4 ++--\n 1 file changed, 2 insertions(+), 2 deletions(-)\n\ndiff --git a/wt-status.c b/wt-status.c\nindex ab4f80d..ef74864 100644\n--- a/wt-status.c\n+++ b/wt-status.c\n@@ -1677,10 +1677,10 @@ static void wt_shortstatus_print_tracking(struct wt_status *s)\n \t\tcolor_fprintf(s->fp, header_color, LABEL(N_(\"behind \")));\n \t\tcolor_fprintf(s->fp, branch_color_remote, \"%d\", num_theirs);\n \t} else if (!num_theirs) {\n-\t\tcolor_fprintf(s->fp, header_color, LABEL(N_((\"ahead \"))));\n+\t\tcolor_fprintf(s->fp, header_color, LABEL(N_(\"ahead \")));\n \t\tcolor_fprintf(s->fp, branch_color_local, \"%d\", num_ours);\n \t} else {\n-\t\tcolor_fprintf(s->fp, header_color, LABEL(N_((\"ahead \"))));\n+\t\tcolor_fprintf(s->fp, header_color, LABEL(N_(\"ahead \")));\n \t\tcolor_fprintf(s->fp, branch_color_local, \"%d\", num_ours);\n \t\tcolor_fprintf(s->fp, header_color, \", %s\", LABEL(N_(\"behind \")));\n \t\tcolor_fprintf(s->fp, branch_color_remote, \"%d\", num_theirs);\n-- \n2.8.0.rc2.197.g3b88641.dirty\n"},{"id":"280715","messageId":"xmqqoaahaw99.fsf@gitster.mtv.corp.google.com","threadId":"41669","inReplyTo":"56E6D8C4.2010205@drmicha.warpmail.net","subject":"Re: [ANNOUNCE] Git v2.8.0-rc2","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2016-03-14T15:56:34Z","receivedAt":"2016-03-14T15:56:34Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Michael J Gruber writes:\n\n> Have you pulled git.pot for 2.8.0 already? [1]\n\nThanks for bringing this up.\n\nNo, I never pull the \"git.pot round N for release\" myself, as I am\nnot working on any translations, and I expect that translation teams\nwork off of i18n coordinator's tree (i.e. Jiang's tree), not from\nmine. The updated git.pot would come to me along with the new\ntranslations via his tree.\n\nBut if it makes it easier for translations teams and the i18n\ncoordinator to work together if I also pulled the git.pot update\nmyself, I'll do so. I just didn't know (and still don't know) if\nthat makes things easier for you guys, or if that risks making\nthings more confusing, having to or being able to pull from two\ntrees that are not necessarily in sync down to the minute.\n\n> [1] http://permalink.gmane.org/gmane.comp.version-control.git/287774\n"},{"id":"280716","messageId":"xmqqk2l5aw8a.fsf@gitster.mtv.corp.google.com","threadId":"41669","inReplyTo":"3b88641beff6b34ae8b86bf353e6fbb8f9e636f1.1457969307.git.git@drmicha.warpmail.net","subject":"Re: [PATCH] wt-status: allow \"ahead \" to be picked up by l10n","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2016-03-14T15:57:09Z","receivedAt":"2016-03-14T15:57:09Z","isPatch":true,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Michael J Gruber writes:\n\n> The extra pair of parentheses keeps the l10n engine from picking up the\n> string. Remove them so that \"ahead \" ends up in git.pot.\n>\n> Signed-off-by: Michael J Gruber \n> ---\n\nThanks, will apply.\n\n> wt-status.c | 4 ++--\n> 1 file changed, 2 insertions(+), 2 deletions(-)\n>\n> diff --git a/wt-status.c b/wt-status.c\n> index ab4f80d..ef74864 100644\n> --- a/wt-status.c\n> +++ b/wt-status.c\n> @@ -1677,10 +1677,10 @@ static void wt_shortstatus_print_tracking(struct wt_status *s)\n> \t\tcolor_fprintf(s->fp, header_color, LABEL(N_(\"behind \")));\n> \t\tcolor_fprintf(s->fp, branch_color_remote, \"%d\", num_theirs);\n> \t} else if (!num_theirs) {\n> -\t\tcolor_fprintf(s->fp, header_color, LABEL(N_((\"ahead \"))));\n> +\t\tcolor_fprintf(s->fp, header_color, LABEL(N_(\"ahead \")));\n> \t\tcolor_fprintf(s->fp, branch_color_local, \"%d\", num_ours);\n> \t} else {\n> -\t\tcolor_fprintf(s->fp, header_color, LABEL(N_((\"ahead \"))));\n> +\t\tcolor_fprintf(s->fp, header_color, LABEL(N_(\"ahead \")));\n> \t\tcolor_fprintf(s->fp, branch_color_local, \"%d\", num_ours);\n> \t\tcolor_fprintf(s->fp, header_color, \", %s\", LABEL(N_(\"behind \")));\n> \t\tcolor_fprintf(s->fp, branch_color_remote, \"%d\", num_theirs);\n"},{"id":"280722","messageId":"xmqqziu19cjz.fsf@gitster.mtv.corp.google.com","threadId":"41669","inReplyTo":"xmqqoaahaw99.fsf@gitster.mtv.corp.google.com","subject":"Re: [ANNOUNCE] Git v2.8.0-rc2","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2016-03-14T17:47:28Z","receivedAt":"2016-03-14T17:47:28Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Junio C Hamano writes:\n\n> But if it makes it easier for translations teams and the i18n\n> coordinator to work together if I also pulled the git.pot update\n> myself, I'll do so. I just didn't know (and still don't know) if\n> that makes things easier for you guys, or if that risks making\n> things more confusing, having to or being able to pull from two\n> trees that are not necessarily in sync down to the minute.\n\nSo, please just tell me to pull it myself too, if it makes the life\nof i18n team and the coordinator easier.\n\nThanks.\n"},{"id":"280724","messageId":"xmqqvb4p9bgz.fsf@gitster.mtv.corp.google.com","threadId":"41669","inReplyTo":"006DA180-E4B6-404E-9377-F2BB41F65591@quendi.de","subject":"Re: [ANNOUNCE] Git v2.8.0-rc2","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2016-03-14T18:10:52Z","receivedAt":"2016-03-14T18:10:52Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Max Horn writes:\n\n>> On 11 Mar 2016, at 00:04, Junio C Hamano wrote:\n>> ...\n>> * \"branch --delete\" has \"branch -d\" but \"push --delete\" does not.\n>\n> This states a problem, but not whether (and how) it was resolved?\n\nThanks; how about\n\n * \"push\" learned that its \"--delete\" option can be shortened to\n \"-d\", just like \"branch --delete\" and \"branch -d\" are the same\n thing.\n\n>> * Across the transition at around Git version 2.0, the user used to\n>> get a pretty loud warning when running \"git push\" without setting\n>> push.default configuration variable. We no longer warn, given that\n>> the transition is over long time ago.\n>\n> That last sentence sounds weird... perhaps \"the transition was\n> completed a long time ago\" ? Or \"the transition ended a long time\n> ago\" ?\n\n * ... We no longer warn because the transition was completed a long\n time ago.\n\n>> * A slight update to the Makefile to mark \"phoney\" targets\n>> as such correctly.\n>\n> phoney -> phony?\n\nThanks for sharp eyes.\n\n>> * Some calls to strcpy(3) triggers a false warning from static\n>> analysers that are less intelligent than humans, and reducing the\n>> number of these false hits helps us notice real issues. A few\n>> calls to strcpy(3) in test-path-utils that are already safe has\n>> been rewritten to avoid false wanings.\n>> \n>> * Some calls to strcpy(3) triggers a false warning from static\n>> analysers that are less intelligent than humans, and reducing the\n>> number of these false hits helps us notice real issues. A few\n>> calls to strcpy(3) in \"git rerere\" that are already safe has been\n>> rewritten to avoid false wanings.\n>\n> The above two messages are very similar, only the end differs a bit.\n\nThat's deliberate as they are two different fixes to a similar\nproblems. We can just omit the names of the components to make them\ninto one entry if we really wanted to.\n\n>> * Asking gitweb for a nonexistent commit left a warning in the server\n>> log.\n>> \n>> Somebody may want to follow this up with an additional test, perhaps?\n>> IIRC, we do test that no Perl warnings are given to the server log,\n>> so this should have been caught if our test coverage were good.\n>\n> That last paragraph seems odd for a changelog?\n\nI do not think it is wrong to leave a reminder to ourselves that\nthere are yet more work need to be done.\n\n>> * The underlying machinery used by \"ls-files -o\" and other commands\n>> have been taught not to create empty submodule ref cache for a\n>\n> have -> has (the machinery is singular)\n\nAgain, thanks for sharp eyes.\n"},{"id":"280788","messageId":"CANYiYbGkieYRVCEgHCADKwDFXmXexyCV0C1GquDv8FbX+3dYYw@mail.gmail.com","threadId":"41669","inReplyTo":"56E6D8C4.2010205@drmicha.warpmail.net","subject":"Re: [ANNOUNCE] Git v2.8.0-rc2","fromName":"Jiang Xin","fromEmail":"worldhello.net@gmail.com","sentAt":"2016-03-15T16:42:35Z","receivedAt":"2016-03-15T16:42:35Z","isPatch":false,"sender":{"key":"worldhello.net@gmail.com","avatar":"https://avatars.githubusercontent.com/u/183860?v=4"},"body":"2016-03-14 23:29 GMT+08:00 Michael J Gruber :\n> Hi Junio,\n>\n> Have you pulled git.pot for 2.8.0 already? [1]\n>\n> Maybe I'm not up-to-date on the release cycle timing regarding l10n, but\n> I was trying to fix a mixed translation/non-translation issue, and it\n> turned out the reason is probably that it is not in \"git.pot\":\n>\n> \"behind \" is in git.pot\n> \"ahead \" is not in git.pot\n\nYes, I find your commit has just been merged in the master branch, and\nit brings one extra message to be translated in Git v2.8.0 l10n.\n\nSo let's start round #3 of l10n for Git 2.8.0 to make it perfect.\n\n> Now, I can recreate git.pot locally, of course, but localisations such\n> as de.po seem to have those translation commented out because they're\n> missing from git.pot.\n>\n> Basically, translations don't get proper testing at this point in the\n> release cycle.\n\nI will send a pull request to Junio right now, and you can test it on\na updated git.pot.\n\n\n-- \nJiang Xin\n"},{"id":"280898","messageId":"56E96096.4020108@drmicha.warpmail.net","threadId":"41669","inReplyTo":"xmqqziu19cjz.fsf@gitster.mtv.corp.google.com","subject":"Re: [ANNOUNCE] Git v2.8.0-rc2","fromName":"Michael J Gruber","fromEmail":"git@drmicha.warpmail.net","sentAt":"2016-03-16T13:33:10Z","receivedAt":"2016-03-16T13:33:10Z","isPatch":false,"sender":{"key":"git@grubix.eu","avatar":"https://avatars.githubusercontent.com/u/233215?v=4"},"body":"Junio C Hamano venit, vidit, dixit 14.03.2016 18:47:\n> Junio C Hamano writes:\n> \n>> But if it makes it easier for translations teams and the i18n\n>> coordinator to work together if I also pulled the git.pot update\n>> myself, I'll do so. I just didn't know (and still don't know) if\n>> that makes things easier for you guys, or if that risks making\n>> things more confusing, having to or being able to pull from two\n>> trees that are not necessarily in sync down to the minute.\n> \n> So, please just tell me to pull it myself too, if it makes the life\n> of i18n team and the coordinator easier.\n> \n> Thanks.\n> \n\nI don't know about the workflow in general. I'll write up what triggered\nmy question: I was looking at the FAQ \"how do I display the current\nbranch in git\" and into ways to provide some ui (think \"git status -sb\",\nthe \"+\"-line in \"git branch\"), when I found the problematic output. The\nmultiple parentheses looked suspicious to me, but given the many levels\nof macro expansion I wasn't sure, and simply patching the parentheses\ndidn't help either. It needed a combination of \"make pot\" and \"msgmerge\n...\", and the fact that the last merge of git.pot was from 2.7.0-rc\ntriggered my request to merge what we have.\n\nIn hindsight, what happened must have been like this:\n\n\"ahead \" was marked properly for l10n and translated in the past.\n\n7a76c28 (status: disable translation when --porcelain is used,\n2014-03-20) introduced those extra parentheses. Matthieu probably didn't\nrerun \"make pot\" and \"msgmerge\" so that he didn't notice the consequences.\n\nWhen Jian ran \"make pot\" the \"ahead \"-entry got removed from git.pot:\n5e078fc (l10n: git.pot: v2.0.0 round 1 (45 new, 28 removed), 2014-04-19)\n\nWhen translators ran \"msgmerge\" with the new git.pot the existing \"ahead\n\"-entry got commented out, for example here for de.po:\n74c17bb (l10n: de.po: translate 45 new messages, 2014-04-01)\n\nI'm actually wondering why I didn't notice this much earlier. I don't\nknow which workflow would have prevented this either. Maybe, since we\nhave \"make pot\", we should also have \"make l10n\" or something to make it\n(even) easier for non-l10n-experts to check whether they introduced any\nproblems.\n\nStrictly speaking, every source file with i18n markup should trigger a\n\"make pot\" (and make l10n) when changed, but there's probably a good\nreason why we don't do that.\n\nMichael\n"},{"id":"280899","messageId":"CACsJy8CyLGtYpPnwdhS-AD4d2pcU7RVv=OxPEePcHYhFpDh4pg@mail.gmail.com","threadId":"41669","inReplyTo":"56E96096.4020108@drmicha.warpmail.net","subject":"Re: [ANNOUNCE] Git v2.8.0-rc2","fromName":"Duy Nguyen","fromEmail":"pclouds@gmail.com","sentAt":"2016-03-16T13:40:54Z","receivedAt":"2016-03-16T13:40:54Z","isPatch":false,"sender":{"key":"pclouds@gmail.com","avatar":"https://avatars.githubusercontent.com/u/720?v=4"},"body":"On Wed, Mar 16, 2016 at 8:33 PM, Michael J Gruber\n wrote:\n> In hindsight, what happened must have been like this:\n>\n> \"ahead \" was marked properly for l10n and translated in the past.\n>\n> 7a76c28 (status: disable translation when --porcelain is used,\n> 2014-03-20) introduced those extra parentheses. Matthieu probably didn't\n> rerun \"make pot\" and \"msgmerge\" so that he didn't notice the consequences.\n\n.po(t) files are _very_ noisy because they record line numbers and\nthat pretty much guarantees huge diffs. Even I as a former translator\ndo not want to \"make pot\" and see the differences. If only msgmerge\nsupports an output without line numbers...\n-- \nDuy\n"},{"id":"280908","messageId":"56E97CA8.8070400@drmicha.warpmail.net","threadId":"41669","inReplyTo":"CACsJy8CyLGtYpPnwdhS-AD4d2pcU7RVv=OxPEePcHYhFpDh4pg@mail.gmail.com","subject":"Re: [ANNOUNCE] Git v2.8.0-rc2","fromName":"Michael J Gruber","fromEmail":"git@drmicha.warpmail.net","sentAt":"2016-03-16T15:32:56Z","receivedAt":"2016-03-16T15:32:56Z","isPatch":false,"sender":{"key":"git@grubix.eu","avatar":"https://avatars.githubusercontent.com/u/233215?v=4"},"body":"Duy Nguyen venit, vidit, dixit 16.03.2016 14:40:\n> On Wed, Mar 16, 2016 at 8:33 PM, Michael J Gruber\n> wrote:\n>> In hindsight, what happened must have been like this:\n>>\n>> \"ahead \" was marked properly for l10n and translated in the past.\n>>\n>> 7a76c28 (status: disable translation when --porcelain is used,\n>> 2014-03-20) introduced those extra parentheses. Matthieu probably didn't\n>> rerun \"make pot\" and \"msgmerge\" so that he didn't notice the consequences.\n> \n> .po(t) files are _very_ noisy because they record line numbers and\n> that pretty much guarantees huge diffs. Even I as a former translator\n> do not want to \"make pot\" and see the differences. If only msgmerge\n> supports an output without line numbers...\n> \n\necho '*.po diff=po' >>.gitattributes\necho '*.pot diff=po' >>.gitattributes\ngit config diff.po.textconv \"msgcat --indent --no-location\"\n\nWith or without the indent, that gives a pretty clean diff. [It's\nunfortunate that one half of that config is in-tree, one-half is not.]\n\nThe way it currently is, \"non-l10n\" coders don't even notice when their\nchanges affect the set of marked strings. But I guess ususally that's\nnot a problem.\n\nFor our l10n coordinator it is impossible to decide whether a \"git.pot\"\ndiff between two runs looks the way it is supposed to look - he would\nhave to be aware of all code changes that intentionally change the\nmarked strings. (Likewise, for the translators.)\n\nSo, really, the \"actual coders\" know best whether their changes should\naffect l10n or not, so they should be made more aware of it. Forcing\n\"make pot\" (and maybe more) on everyone sounds a bit harsh, but what\nelse can we do?\n\nMichael\n"},{"id":"280915","messageId":"xmqqmvpy5qru.fsf@gitster.mtv.corp.google.com","threadId":"41669","inReplyTo":"56E97CA8.8070400@drmicha.warpmail.net","subject":"Re: [ANNOUNCE] Git v2.8.0-rc2","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2016-03-16T16:30:45Z","receivedAt":"2016-03-16T16:30:45Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Michael J Gruber writes:\n\n> echo '*.po diff=po' >>.gitattributes\n> echo '*.pot diff=po' >>.gitattributes\n> git config diff.po.textconv \"msgcat --indent --no-location\"\n>\n> With or without the indent, that gives a pretty clean diff. [It's\n> unfortunate that one half of that config is in-tree, one-half is not.]\n\nThat's a good tip. [By the way, it is not unfortunate that these are\nseparated to two places, but quite the opposite. Attributes define\n\"what kind of things\" they are, and configuration defines \"how\" each\nkind of things are handled. \"msgcat\" may have to be invoked\ndifferently from yours on other people's systems, and one level of\nindirection is a reasonable way to allow customizing \"how\" part\nwithout forcing people to rewrite all of THIS in \"for *.po do THIS,\nfor *.pot do THIS too\". You should be thankful for this separation.]\n\n> So, really, the \"actual coders\" know best whether their changes should\n> affect l10n or not, so they should be made more aware of it. Forcing\n> \"make pot\" (and maybe more) on everyone sounds a bit harsh, but what\n> else can we do?\n\nI am not sure what problem you are trying to solve. Do you want to\nmake sure mismarking such as N_((\"foo\")) is caught by the person who\nchanges \"foo\" into N_((\"foo\"))?\n\n\"make pot\" alone would obviously not help, and you would definitely\nneed \"maybe more\" but I'd imagine that would involve checking the\ndiff in the code part i.e. \"we have a new N_(...)\" against the\ndifferences in git.pot files you would obtain by running \"make pot\"\nbefore the code change and after the code change, i.e. \"there is no\nnew mention of \"foo\"\".\n\nI do not think you are suggesting to commit the result of \"make pot\"\nalong with code changes, but if you are, please don't ;-)\n"},{"id":"281044","messageId":"56EAC47F.6000708@drmicha.warpmail.net","threadId":"41669","inReplyTo":"xmqqmvpy5qru.fsf@gitster.mtv.corp.google.com","subject":"Re: [ANNOUNCE] Git v2.8.0-rc2","fromName":"Michael J Gruber","fromEmail":"git@drmicha.warpmail.net","sentAt":"2016-03-17T14:51:43Z","receivedAt":"2016-03-17T14:51:43Z","isPatch":false,"sender":{"key":"git@grubix.eu","avatar":"https://avatars.githubusercontent.com/u/233215?v=4"},"body":"Junio C Hamano venit, vidit, dixit 16.03.2016 17:30:\n> Michael J Gruber writes:\n> \n>> echo '*.po diff=po' >>.gitattributes\n>> echo '*.pot diff=po' >>.gitattributes\n>> git config diff.po.textconv \"msgcat --indent --no-location\"\n>>\n>> With or without the indent, that gives a pretty clean diff. [It's\n>> unfortunate that one half of that config is in-tree, one-half is not.]\n> \n> That's a good tip. [By the way, it is not unfortunate that these are\n> separated to two places, but quite the opposite. Attributes define\n> \"what kind of things\" they are, and configuration defines \"how\" each\n> kind of things are handled. \"msgcat\" may have to be invoked\n> differently from yours on other people's systems, and one level of\n> indirection is a reasonable way to allow customizing \"how\" part\n> without forcing people to rewrite all of THIS in \"for *.po do THIS,\n> for *.pot do THIS too\". You should be thankful for this separation.]\n\nI know why we have that. It's just unfortunate that we can't even\nprovide a default textconfig config easily - I know very well we can't\nhave that securely.\n\n\"Unfortunate\" is meant in this context in the sense: I want to make it\neasy and convenient for non-l10n-people to watch out for l10-affecting\nchanges.\n\n>> So, really, the \"actual coders\" know best whether their changes should\n>> affect l10n or not, so they should be made more aware of it. Forcing\n>> \"make pot\" (and maybe more) on everyone sounds a bit harsh, but what\n>> else can we do?\n> \n> I am not sure what problem you are trying to solve. Do you want to\n> make sure mismarking such as N_((\"foo\")) is caught by the person who\n> changes \"foo\" into N_((\"foo\"))?\n\nYes. That and similar ones.\n\n> \"make pot\" alone would obviously not help, and you would definitely\n> need \"maybe more\" but I'd imagine that would involve checking the\n> diff in the code part i.e. \"we have a new N_(...)\" against the\n> differences in git.pot files you would obtain by running \"make pot\"\n> before the code change and after the code change, i.e. \"there is no\n> new mention of \"foo\"\".\n> \n> I do not think you are suggesting to commit the result of \"make pot\"\n> along with code changes, but if you are, please don't ;-)\n\nAs I said: I assume there's a good reason we don't do that, and that's\nwhy I'm not suggesting it.\n\nOn the other hand, it means that the in-tree git.pot does not correlate\nwith the in-tree source code at all, which feels really weird[*]. And it\nmakes it difficult to check the impact of your code changes: You can't\nsimply run \"make pot\" and check the diff - because the in-tree git.pot\ndoes not reflect the state before your changes at all.\n\n[*] It just feels wrong to me that a \"make pot\" in a clean checkout\nleads to dirty index.\n\nSo, in short, I do believe there is a good reason for the \"out of sync\"\ngit.pot. That doesn't make the negative side effect that I describe any\nless true, and I'm looking for ways to ammeliorate that. Something as\neasy as \"make check\" or \"make test-lint\".\n\nMichael\n"},{"id":"281046","messageId":"2296423d1c46365f1fdeaa12cd1917bc859ff8d9.1458227678.git.git@drmicha.warpmail.net","threadId":"41669","inReplyTo":"56EAC47F.6000708@drmicha.warpmail.net","subject":"[RFC/PATCH] Makefile: allow po generation through po target","fromName":"Michael J Gruber","fromEmail":"git@drmicha.warpmail.net","sentAt":"2016-03-17T15:14:52Z","receivedAt":"2016-03-17T15:14:52Z","isPatch":true,"sender":{"key":"git@grubix.eu","avatar":"https://avatars.githubusercontent.com/u/233215?v=4"},"body":"The main Makefile has a \"pot\" target that recreates the git.pot file of\nstrings which are marked for translation.\n\nAdd a \"po\" target that recreates the $(LANGUAGE).po files which contain\nthe translations (or stubs).\n\nSigned-off-by: Michael J Gruber \n---\n\nNotes:\n This makes it easier to recreate po (and mo) without reading po/README.\n Alternatively, one may think about a Makefile in po/ which does both pot\n and po updates, just like we have makefiles in t/ and Ducumentation/.\n \n This doesn't give you proper before-after diffs yet, but at least the after\n state.\n\n Makefile | 8 ++++++++\n 1 file changed, 8 insertions(+)\n\ndiff --git a/Makefile b/Makefile\nindex 24bef8d..bb60a34 100644\n--- a/Makefile\n+++ b/Makefile\n@@ -445,6 +445,7 @@ TCL_PATH = tclsh\n TCLTK_PATH = wish\n XGETTEXT = xgettext\n MSGFMT = msgfmt\n+MSGMERGE = msgmerge\n CURL_CONFIG = curl-config\n PTHREAD_LIBS = -lpthread\n PTHREAD_CFLAGS =\n@@ -1514,6 +1515,7 @@ ifndef V\n \tQUIET_LNCP = @echo ' ' LN/CP $@;\n \tQUIET_XGETTEXT = @echo ' ' XGETTEXT $@;\n \tQUIET_MSGFMT = @echo ' ' MSGFMT $@;\n+\tQUIET_MSGMERGE = @echo ' ' MSGMERGE $@;\n \tQUIET_GCOV = @echo ' ' GCOV $@;\n \tQUIET_SP = @echo ' ' SP $<;\n \tQUIET_RC = @echo ' ' RC $@;\n@@ -2088,6 +2090,9 @@ pot: po/git.pot\n POFILES := $(wildcard po/*.po)\n MOFILES := $(patsubst po/%.po,po/build/locale/%/LC_MESSAGES/git.mo,$(POFILES))\n \n+.PHONY: po\n+po: $(POFILES)\n+\n ifndef NO_GETTEXT\n all:: $(MOFILES)\n endif\n@@ -2095,6 +2100,9 @@ endif\n po/build/locale/%/LC_MESSAGES/git.mo: po/%.po\n \t$(QUIET_MSGFMT)mkdir -p $(dir $@) && $(MSGFMT) -o $@ $<\n \n+po/%.po: po/git.pot\n+\t$(QUIET_MSGMERGE)$(MSGMERGE) --add-location --backup=off -q -U $@ $<\n+\n FIND_SOURCE_FILES = ( git ls-files '*.[hcS]' 2>/dev/null || \\\n \t\t\t$(FIND) . \\( -name .git -type d -prune \\) \\\n \t\t\t\t-o \\( -name '*.[hcS]' -type f -print \\) )\n-- \n2.8.0.rc2.197.g3b88641.dirty\n"},{"id":"281047","messageId":"xmqqio0l13nz.fsf@gitster.mtv.corp.google.com","threadId":"41669","inReplyTo":"56EAC47F.6000708@drmicha.warpmail.net","subject":"Re: [ANNOUNCE] Git v2.8.0-rc2","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2016-03-17T16:15:44Z","receivedAt":"2016-03-17T16:15:44Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Michael J Gruber writes:\n\n> So, in short, I do believe there is a good reason for the \"out of sync\"\n> git.pot. That doesn't make the negative side effect that I describe any\n> less true, and I'm looking for ways to ammeliorate that. Something as\n> easy as \"make check\" or \"make test-lint\".\n\nYes, I agree fully with the last sentence. I think the task\nprobably needs two building blocks:\n\n - A tool to generate up-to-date git.pot contents and store into a\n given filename.\n\n - A tool that can be used to compare two versions of a .pot file,\n stored as two separate files in the filesystem, without the noise\n that comes from auto-generator (e.g. line numbers, the difference\n of line-wrapping the same messge).\n\nWith that, a user can (1) run the former and save the result in\ngit.pot-before-change before making any change (2) do her work, (3)\nrun the former again and save the result in git.pot-after-change,\nand (4) run the latter between these two.\n\nThis is a bit hard to drive from the Makefile, though, as our\nMakefile cannot assume people are using Git (they may be working off\nof a tarball extract to produce a one-liner patch), and there is the\nissue of \"then how do we ensure that the user runs the former first\non prestine before starting to hack so that later the result can be\ncompared?\"\n\nBut just like the version number generation, it is OK for some\ntargets to be optional, so perhaps it is OK for this \"view pot\nchanges\" task is limited to those who work on their own clone of\nGit. If we add that assumption, then the Makefile target for\n\"view-pot-changes\" would\n\n - see if untracked file git.pot-$(git rev-parse HEAD) exists (treat\n this as a build artifact, like *.o files). If it does not exist:\n\n - check out HEAD to a temporary location elsewhere on the\n filesystem;\n\n - run the first tool and store the result in the above file.\n\n - run the first tool in the working tree and store the result in\n another file git.pot-now (treat this file as a build artifact,\n too).\n\n - run the second tool on these two git.pot files.\n\nAdd \"git.pot-*\" to our .gitignore file, and make sure \"make clean\"\nremoves them.\n"},{"id":"281091","messageId":"xmqq8u1gzpz3.fsf@gitster.mtv.corp.google.com","threadId":"41669","inReplyTo":"2296423d1c46365f1fdeaa12cd1917bc859ff8d9.1458227678.git.git@drmicha.warpmail.net","subject":"Re: [RFC/PATCH] Makefile: allow po generation through po target","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2016-03-17T22:42:08Z","receivedAt":"2016-03-17T22:42:08Z","isPatch":true,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Michael J Gruber writes:\n\n> The main Makefile has a \"pot\" target that recreates the git.pot file of\n> strings which are marked for translation.\n>\n> Add a \"po\" target that recreates the $(LANGUAGE).po files which contain\n> the translations (or stubs).\n>\n> Signed-off-by: Michael J Gruber \n> ---\n>\n> Notes:\n> This makes it easier to recreate po (and mo) without reading po/README.\n> Alternatively, one may think about a Makefile in po/ which does both pot\n> and po updates, just like we have makefiles in t/ and Ducumentation/.\n> \n> This doesn't give you proper before-after diffs yet, but at least the after\n> state.\n\nMore seriously, the \"before state\" does not exist anywhere, because\n*.po and *.pot are expected not to be in sync with the source, and\nafter a code developer runs \"make po\", because she does not know all\nthe languages we have *.po for, she has to \"git checkout\" to erase\nthe changes made by \"make po\".\n\nSo while your starting discussion (i.e. RFC-ness of this patch) is\nvery much appreciated, I do not think this is a good change we would\nwant to base further work on top. For \"before-after-diff\" purposes,\nthe targets for before and after should drop their output in an\nuntracked build artifact, instead of overwriting tracked files, I\nwould think.\n"},{"id":"281268","messageId":"CANYiYbFa5i-E0dYYj2dm4pHmQwLJfj3UBc3OspQz93HTP3C3Ng@mail.gmail.com","threadId":"41669","inReplyTo":"56E96096.4020108@drmicha.warpmail.net","subject":"Re: [ANNOUNCE] Git v2.8.0-rc2","fromName":"Jiang Xin","fromEmail":"worldhello.net@gmail.com","sentAt":"2016-03-20T09:45:19Z","receivedAt":"2016-03-20T09:45:19Z","isPatch":false,"sender":{"key":"worldhello.net@gmail.com","avatar":"https://avatars.githubusercontent.com/u/183860?v=4"},"body":"2016-03-16 21:33 GMT+08:00 Michael J Gruber :\n> Junio C Hamano venit, vidit, dixit 14.03.2016 18:47:\n>> Junio C Hamano writes:\n>>\n>>> But if it makes it easier for translations teams and the i18n\n>>> coordinator to work together if I also pulled the git.pot update\n>>> myself, I'll do so. I just didn't know (and still don't know) if\n>>> that makes things easier for you guys, or if that risks making\n>>> things more confusing, having to or being able to pull from two\n>>> trees that are not necessarily in sync down to the minute.\n>>\n>> So, please just tell me to pull it myself too, if it makes the life\n>> of i18n team and the coordinator easier.\n>>\n>> Thanks.\n>>\n>\n> I don't know about the workflow in general. I'll write up what triggered\n> my question: I was looking at the FAQ \"how do I display the current\n> branch in git\" and into ways to provide some ui (think \"git status -sb\",\n> the \"+\"-line in \"git branch\"), when I found the problematic output. The\n> multiple parentheses looked suspicious to me, but given the many levels\n> of macro expansion I wasn't sure, and simply patching the parentheses\n> didn't help either. It needed a combination of \"make pot\" and \"msgmerge\n> ...\", and the fact that the last merge of git.pot was from 2.7.0-rc\n> triggered my request to merge what we have.\n>\n> In hindsight, what happened must have been like this:\n>\n> \"ahead \" was marked properly for l10n and translated in the past.\n>\n> 7a76c28 (status: disable translation when --porcelain is used,\n> 2014-03-20) introduced those extra parentheses. Matthieu probably didn't\n> rerun \"make pot\" and \"msgmerge\" so that he didn't notice the consequences.\n>\n> When Jian ran \"make pot\" the \"ahead \"-entry got removed from git.pot:\n> 5e078fc (l10n: git.pot: v2.0.0 round 1 (45 new, 28 removed), 2014-04-19)\n>\n> When translators ran \"msgmerge\" with the new git.pot the existing \"ahead\n> \"-entry got commented out, for example here for de.po:\n> 74c17bb (l10n: de.po: translate 45 new messages, 2014-04-01)\n>\n> I'm actually wondering why I didn't notice this much earlier. I don't\n> know which workflow would have prevented this either. Maybe, since we\n> have \"make pot\", we should also have \"make l10n\" or something to make it\n> (even) easier for non-l10n-experts to check whether they introduced any\n> problems.\n>\n> Strictly speaking, every source file with i18n markup should trigger a\n> \"make pot\" (and make l10n) when changed, but there's probably a good\n> reason why we don't do that.\n\nHow about this? We can make a website (host on github) to show i18n\nchanges. Homepage is just a markdown page, list no-merge commits\nwhich introduced i18n changes. So we don't have to change the workflow.\n\n1. A local git clone, keep update with upstream.\n2. Give a starting point, and generate a no-merge commit list.\n3. Scan through local git clone and generate a pot file for each\n commit and its parent commit, and save them in cache (by hard\n link to save disk space).\n4. Generate diffs for each commit.\n5. Generate the MarkDown web page through a template.\n6. Commit the changes of the markdown page, and push to github.\n\n-- \nJiang Xin\n"},{"id":"281280","messageId":"CAA19uiSUV0C=WQAhgum9MM8r8NixMF8O0XOFxzywSJtBEcGNmQ@mail.gmail.com","threadId":"41669","inReplyTo":"CANYiYbFa5i-E0dYYj2dm4pHmQwLJfj3UBc3OspQz93HTP3C3Ng@mail.gmail.com","subject":"Re: [ANNOUNCE] Git v2.8.0-rc2","fromName":"Michael J Gruber","fromEmail":"git@drmicha.warpmail.net","sentAt":"2016-03-20T15:11:31Z","receivedAt":"2016-03-20T15:11:31Z","isPatch":false,"sender":{"key":"git@grubix.eu","avatar":"https://avatars.githubusercontent.com/u/233215?v=4"},"body":"2016-03-20 10:45 GMT+01:00 Jiang Xin :\n> 2016-03-16 21:33 GMT+08:00 Michael J Gruber :\n>> Junio C Hamano venit, vidit, dixit 14.03.2016 18:47:\n>>> Junio C Hamano writes:\n>>>\n>>>> But if it makes it easier for translations teams and the i18n\n>>>> coordinator to work together if I also pulled the git.pot update\n>>>> myself, I'll do so. I just didn't know (and still don't know) if\n>>>> that makes things easier for you guys, or if that risks making\n>>>> things more confusing, having to or being able to pull from two\n>>>> trees that are not necessarily in sync down to the minute.\n>>>\n>>> So, please just tell me to pull it myself too, if it makes the life\n>>> of i18n team and the coordinator easier.\n>>>\n>>> Thanks.\n>>>\n>>\n>> I don't know about the workflow in general. I'll write up what triggered\n>> my question: I was looking at the FAQ \"how do I display the current\n>> branch in git\" and into ways to provide some ui (think \"git status -sb\",\n>> the \"+\"-line in \"git branch\"), when I found the problematic output. The\n>> multiple parentheses looked suspicious to me, but given the many levels\n>> of macro expansion I wasn't sure, and simply patching the parentheses\n>> didn't help either. It needed a combination of \"make pot\" and \"msgmerge\n>> ...\", and the fact that the last merge of git.pot was from 2.7.0-rc\n>> triggered my request to merge what we have.\n>>\n>> In hindsight, what happened must have been like this:\n>>\n>> \"ahead \" was marked properly for l10n and translated in the past.\n>>\n>> 7a76c28 (status: disable translation when --porcelain is used,\n>> 2014-03-20) introduced those extra parentheses. Matthieu probably didn't\n>> rerun \"make pot\" and \"msgmerge\" so that he didn't notice the consequences.\n>>\n>> When Jian ran \"make pot\" the \"ahead \"-entry got removed from git.pot:\n>> 5e078fc (l10n: git.pot: v2.0.0 round 1 (45 new, 28 removed), 2014-04-19)\n>>\n>> When translators ran \"msgmerge\" with the new git.pot the existing \"ahead\n>> \"-entry got commented out, for example here for de.po:\n>> 74c17bb (l10n: de.po: translate 45 new messages, 2014-04-01)\n>>\n>> I'm actually wondering why I didn't notice this much earlier. I don't\n>> know which workflow would have prevented this either. Maybe, since we\n>> have \"make pot\", we should also have \"make l10n\" or something to make it\n>> (even) easier for non-l10n-experts to check whether they introduced any\n>> problems.\n>>\n>> Strictly speaking, every source file with i18n markup should trigger a\n>> \"make pot\" (and make l10n) when changed, but there's probably a good\n>> reason why we don't do that.\n>\n> How about this? We can make a website (host on github) to show i18n\n> changes. Homepage is just a markdown page, list no-merge commits\n> which introduced i18n changes. So we don't have to change the workflow.\n>\n> 1. A local git clone, keep update with upstream.\n> 2. Give a starting point, and generate a no-merge commit list.\n> 3. Scan through local git clone and generate a pot file for each\n> commit and its parent commit, and save them in cache (by hard\n> link to save disk space).\n> 4. Generate diffs for each commit.\n> 5. Generate the MarkDown web page through a template.\n> 6. Commit the changes of the markdown page, and push to github.\n>\n> --\n> Jiang Xin\n\nI think this is a general question about how to track build products.\nThe proper place may be in a tree that is referenced from a note or\nso.\nMaybe I shouldn't consider git.pot a build product - I don't know, as\nI honestly don't know why we treat it the way we do.\nI could understand if we don't track it all, or if we regnerate it\nfrom \"make all\" and track the result. We do track it but let it go\nstale. There must be a good reason for that, it's just that I don't\nknow it.\n\nMaybe the problem that I'm trying to solve appears rare enough to not\nwarrant a solution...\n\nThe question of how to track build products (such as pdf compiled from\nLaTeX) properly may be a more general one, more deserving a solution.\n\nMichael\n"},{"id":"281375","messageId":"xmqqr3f3lhvw.fsf@gitster.mtv.corp.google.com","threadId":"41669","inReplyTo":"CAA19uiSUV0C=WQAhgum9MM8r8NixMF8O0XOFxzywSJtBEcGNmQ@mail.gmail.com","subject":"Re: [ANNOUNCE] Git v2.8.0-rc2","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2016-03-21T20:01:55Z","receivedAt":"2016-03-21T20:01:55Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Michael J Gruber writes:\n\n> I think this is a general question about how to track build\n> products. The proper place may be in a tree that is referenced\n> from a note or so.\n\n> Maybe I shouldn't consider git.pot a build product - I don't know,\n> as I honestly don't know why we treat it the way we do.\n\nI think your LaTeX output analogy is interesting. When working with\nother people editing a single document, each person may update the\nbuild product (.dvi or .pdf or whatever) in his branch and when you\nmerge other people's work, this would create an unresolvable mess\nbut that is perfectly fine, because you wouldn't even attempt to\nmerge the build product. Instead, you would merge the source\nmaterial, run the formatter, and pretend as if its output is the\nresult of the merging of .dvi or .pdf or whatever.\n\nBut then we need to step back and consider the reason why we keep\nthe build product in the first place. Presumably that is to help\nthose who want to consume the build product without having the\ntoolchain to build from the source. If that is the case, perhaps it\nis also a valid workflow for these collaborating authors of a single\ndocument not to update the build product, if they know that nobody\ncares about how the final output looks like on their individual\nfork, until their work is merged to some \"mainline\".\n\nThe primary consumers of git.pot build product are the l10n teams,\nand I do not think that they want to (or it is practical to ask them\nto) work on translating new messages on individual topics code-side\npeople work on. So perhaps it is a valid workflow to leave git.pot\nbehind until i18n coordinator declares \"it is time to catch up\" and\nregenerates it at some \"snapshot\" time in the development cycle.\n"},{"id":"281425","messageId":"56F117C7.2030900@drmicha.warpmail.net","threadId":"41669","inReplyTo":"xmqqr3f3lhvw.fsf@gitster.mtv.corp.google.com","subject":"Re: [ANNOUNCE] Git v2.8.0-rc2","fromName":"Michael J Gruber","fromEmail":"git@drmicha.warpmail.net","sentAt":"2016-03-22T10:00:39Z","receivedAt":"2016-03-22T10:00:39Z","isPatch":false,"sender":{"key":"git@grubix.eu","avatar":"https://avatars.githubusercontent.com/u/233215?v=4"},"body":"Junio C Hamano venit, vidit, dixit 21.03.2016 21:01:\n> Michael J Gruber writes:\n> \n>> I think this is a general question about how to track build\n>> products. The proper place may be in a tree that is referenced\n>> from a note or so.\n> \n>> Maybe I shouldn't consider git.pot a build product - I don't know,\n>> as I honestly don't know why we treat it the way we do.\n> \n> I think your LaTeX output analogy is interesting. When working with\n> other people editing a single document, each person may update the\n> build product (.dvi or .pdf or whatever) in his branch and when you\n> merge other people's work, this would create an unresolvable mess\n> but that is perfectly fine, because you wouldn't even attempt to\n> merge the build product. Instead, you would merge the source\n> material, run the formatter, and pretend as if its output is the\n> result of the merging of .dvi or .pdf or whatever.\n\nI have to confess that that is actually my main use of git outside of\ngit.git, and that my \"solution\" is tracking .pdf etc. and using a\npre-commit (and pre-merge) hook that prevents me from commiting a state\nwhere \"make -q\" is not successful. Still not perfect, partly due to the\nfact that make is timestamp based, not checksum based. (\"Use scons\", I\nknow...)\n\n> But then we need to step back and consider the reason why we keep\n> the build product in the first place. Presumably that is to help\n> those who want to consume the build product without having the\n> toolchain to build from the source. \n\nThat is certainly a good use case - see, e.g., our documentation. On the\nother hand, \"make pot\" has very light toolchain requirements compared to\n\"make doc\".\n\nAnother use case is a \"poor man's release accountability\" - a build\nresult depends on the actual toolchain version also, not just the\nsource, and tracking a build stores a copy of a build result that you\nhanded out in class, deployed somewhere, ...\n\n> If that is the case, perhaps it\n> is also a valid workflow for these collaborating authors of a single\n> document not to update the build product, if they know that nobody\n> cares about how the final output looks like on their individual\n> fork, until their work is merged to some \"mainline\".\n\nWouldn't that reasoning imply that there's no point in tracking at all?\nThat is, in tracking in the main tree.\n\n> The primary consumers of git.pot build product are the l10n teams,\n> and I do not think that they want to (or it is practical to ask them\n> to) work on translating new messages on individual topics code-side\n> people work on. So perhaps it is a valid workflow to leave git.pot\n> behind until i18n coordinator declares \"it is time to catch up\" and\n> regenerates it at some \"snapshot\" time in the development cycle.\n\nYes, I just think that \"git.pot\" in the main tree sends a wrong signal\nbecause it is out of sync in the sense that it is not \"make pot\"-clean.\nThe main tree is the place for snapshots (especially not non-rolling\nones). Possible alternatives:\n\n- Do not track git.pot at all.\nIt can be recreated easily, and (unless I'm horribly mistaken) only l10n\npeople need it ; everyone else needs the .po files, or rather .mo.\n\n- Make po/ a submodule.\nSubmodules are exactly our way of stating \"this depends on a tree with a\nseparate history and/or different release cycle; commit bar in the super\nproject is 'based' on commit foo in the submodule (but not the other way\nround!)\n\n- Do it like the extra prebuild doc-branches.\n\n- Snapshot with tags (similar to github releases).\n\nWith current master at 047057b, the contained git.pot is actually not\n(guaranteed to be) the one containing the translatable strings from that\ntree. If you want to find out which one, you'd have to \"git log -1\npo/git.pot\" and rely on the fact that Jiang's tree actually does keep\ngit.pot in-sync.\n\nSomehow, submodules seem to best represent that one-way dependency. (I'm\nnot a submodule user at all.)\n\nMichael\n\nP.S.: git.pot is both a build product as well as a build/install\nrequirement, so my tex/pdf analogy does not apply fully.\n"},{"id":"281451","messageId":"xmqqoaa6if2a.fsf@gitster.mtv.corp.google.com","threadId":"41669","inReplyTo":"56F117C7.2030900@drmicha.warpmail.net","subject":"Re: [ANNOUNCE] Git v2.8.0-rc2","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2016-03-22T17:43:25Z","receivedAt":"2016-03-22T17:43:25Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"Michael J Gruber writes:\n\n> Junio C Hamano venit, vidit, dixit 21.03.2016 21:01:\n>\n>> If that is the case, perhaps it\n>> is also a valid workflow for these collaborating authors of a single\n>> document not to update the build product, if they know that nobody\n>> cares about how the final output looks like on their individual\n>> fork, until their work is merged to some \"mainline\".\n>\n> Wouldn't that reasoning imply that there's no point in tracking at all?\n> That is, in tracking in the main tree.\n\nNot really.\n\n >> ... this would create an unresolvable mess\n >> but that is perfectly fine, because you wouldn't even attempt to\n >> merge the build product. Instead, you would merge the source\n >> material, run the formatter, and pretend as if its output is the\n >> result of the merging of .dvi or .pdf or whatever.\n\ndeclares \"there is no point in tracking at all\" for them while\nworking on their document. These authors may realize that there is\nno point in tracking the build product in order for them to work\ntogether. And that may lead them to adopt a workflow where they\nleave the build product stale on their individual fork.\n\nIt is perfectly fine for them to have a (different) reason why they\nmay still want to keep the build product tracked and kept up-to-date\non the mainline (e.g. they want to give an easy access to non\nauthors the final document), even after they realize that tracking\nbuild product would not help _their_ work.\n"}]}