1Git v1.7.11 Release Notes 2========================= 3 4Updates since v1.7.10 5--------------------- 6 7UI, Workflows & Features 8 9 * A third-party tool "git subtree" is distributed in contrib/ 10 11 * Even with "-q"uiet option, "checkout" used to report setting up 12 tracking. Also "branch" learned the "-q"uiet option to squelch 13 informational message. 14 15 * The smart-http backend used to always override GIT_COMMITTER_* 16 variables with REMOTE_USER and REMOTE_ADDR, but these variables are 17 now preserved when set. 18 19 * "git am" learned the "--include" option, which is an opposite of 20 existing the "--exclude" option. 21 22 * When "git am -3" needs to fall back to an application to a 23 synthesized preimage followed by a 3-way merge, the paths that 24 needed such treatment are now reported to the end user, so that the 25 result in them can be eyeballed with extra care. 26 27 * The "fmt-merge-msg" command learns to list the primary contributors 28 involved in the side topic you are merging. 29 30 * The cases "git push" fails due to non-ff can be broken into three 31 categories; each case is given a separate advise message. 32 33 * A 'snapshot' request to "gitweb" honors If-Modified-Since: header, 34 based on the commit date. 35 36Foreign Interface 37 38 * "git p4" has been moved out of contrib/ area. 39 40Performance 41 42 * "git apply" had some memory leaks plugged. 43 44 * Setting up a revision traversal with many starting points was 45 inefficient as these were placed in a date-order priority queue 46 one-by-one. Now they are collected in the queue unordered first, 47 and sorted immediately before getting used. 48 49Internal Implementation (please report possible regressions) 50 51 * "git rev-parse --show-prefix" used to emit nothing when run at the 52 top-level of the working tree, but now it gives a blank line. 53 54 * Minor memory leak during unpack_trees (hence "merge" and "checkout" 55 to check out another branch) has been plugged. 56 57 * More lower-level commands learned to use the streaming API to read 58 from the object store without keeping everything in core. 59 60 * Because "sh" on the user's PATH may be utterly broken on some 61 systems, run-command API now uses SHELL_PATH, not /bin/sh, when 62 spawning an external command (not applicable to Windows port). 63 64Also contains minor documentation updates and code clean-ups. 65 66 67Fixes since v1.7.10 68------------------- 69 70Unless otherwise noted, all the fixes since v1.7.10 in the maintenance 71releases are contained in this release (see release notes to them for 72details). 73 74 * "git fetch" that recurses into submodules on demand did not check 75 if it needs to go into submodules when non branches (most notably, 76 tags) are fetched. 77 (merge a6801ad jl/maint-submodule-recurse-fetch later to maint). 78 79 * "git blame" started missing quite a few changes from the origin 80 since we stopped using the diff minimalization by default in v1.7.2 81 era. 82 (merge 059a500 jc/maint-blame-minimal later to maint). 83 84 * "log -p --graph" used with "--stat" had a few formatting error. 85 (merge e2c5966 lp/maint-diff-three-dash-with-graph later to maint). 86 87 * Giving "--continue" to a conflicted "rebase -i" session skipped a 88 commit that only results in changes to submodules. 89 (merge a6754cd jk/rebase-i-submodule-conflict-only later to maint). 90 91 * When PATH contains an unreadable directory, alias expansion code 92 did not kick in, and failed with an error that said "git-subcmd" 93 was not found. 94 (merge 38f865c jk/run-command-eacces later to maint). 95 96 * The 'push to upstream' implementation was broken in some corner 97 cases. "git push $there" without refspec, when the current branch 98 is set to push to a remote different from $there, used to push to 99 $there using the upstream information to a remote unreleated to 100 $there. 101 (merge 135dade jc/push-upstream-sanity later to maint). 102 103 * "git clean -d -f" (not "-d -f -f") is supposed to protect nested 104 working trees of independent git repositories that exist in the 105 current project working tree from getting removed, but the 106 protection applied only to such working trees that are at the 107 top-level of the current project by mistake. 108 (merge ae2f203 jc/maint-clean-nested-worktree-in-subdir later to maint). 109 110 * Rename detection logic used to match two empty files as renames 111 during merge-recursive, leading unnatural mismerges. 112 (merge 4f7cb99 jk/diff-no-rename-empty later to maint). 113 114 * An age-old corner case bug in combine diff (only triggered with -U0 115 and the hunk at the beginning of the file needs to be shown) has 116 been fixed. 117 (merge e5e9b56 rs/combine-diff-zero-context-at-the-beginning later to maint). 118 119 * When "git commit --template F" errors out because the user did not 120 touch the message, it claimed that it aborts due to "empty 121 message", which was utterly wrong. 122 (merge 1f08c2c jc/commit-unedited-template later to maint). 123 124 * "git add -p" is not designed to deal with unmerged paths but did 125 not exclude them and tried to apply funny patches only to fail. 126 (merge 4066bd6 jk/add-p-skip-conflicts later to maint). 127 128 * "git commit --author=$name" did not tell the name that was being 129 recorded in the resulting commit to hooks, even though it does do 130 so when the end user overrode the authorship via the 131 "GIT_AUTHOR_NAME" environment variable. 132 (merge 7dfe8ad jc/commit-hook-authorship later to maint). 133 134 * The regexp configured with diff.wordregex was incorrectly reused 135 across files. 136 (merge 6440d34 tr/maint-word-diff-regex-sticky later to maint). 137 138 * Running "notes merge --commit" failed to perform correctly when run 139 from any directory inside $GIT_DIR/. When "notes merge" stops with 140 conflicts, $GIT_DIR/NOTES_MERGE_WORKTREE is the place a user edits 141 to resolve it. 142 (merge dabba59 jh/notes-merge-in-git-dir-worktree later to maint).