Documentation / RelNotes / 1.7.11.txton commit Merge bundle error message fix in (210a75c)
   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 * Error messages given when @{u} is used for a branch without its
  12   upstream configured have been clatified.
  13
  14 * Even with "-q"uiet option, "checkout" used to report setting up
  15   tracking.  Also "branch" learned the "-q"uiet option to squelch
  16   informational message.
  17
  18 * The smart-http backend used to always override GIT_COMMITTER_*
  19   variables with REMOTE_USER and REMOTE_ADDR, but these variables are
  20   now preserved when set.
  21
  22 * "git am" learned the "--include" option, which is an opposite of
  23   existing the "--exclude" option.
  24
  25 * When "git am -3" needs to fall back to an application to a
  26   synthesized preimage followed by a 3-way merge, the paths that
  27   needed such treatment are now reported to the end user, so that the
  28   result in them can be eyeballed with extra care.
  29
  30 * The "fmt-merge-msg" command learns to list the primary contributors
  31   involved in the side topic you are merging.
  32
  33 * The cases "git push" fails due to non-ff can be broken into three
  34   categories; each case is given a separate advise message.
  35
  36 * "git push --recurse-submodules" learned to optionally look into the
  37   histories of submodules bound to the superproject and push them
  38   out.
  39
  40 * A 'snapshot' request to "gitweb" honors If-Modified-Since: header,
  41   based on the commit date.
  42
  43 * "gitweb" learned to highlight the patch it outputs even more.
  44
  45Foreign Interface
  46
  47 * "git svn" used to die with unwanted SIGPIPE when talking with HTTP
  48   server that uses keep-alive.
  49
  50 * "git p4" has been moved out of contrib/ area.
  51
  52Performance
  53
  54 * "git apply" had some memory leaks plugged.
  55
  56 * Setting up a revision traversal with many starting points was
  57   inefficient as these were placed in a date-order priority queue
  58   one-by-one.  Now they are collected in the queue unordered first,
  59   and sorted immediately before getting used.
  60
  61Internal Implementation (please report possible regressions)
  62
  63 * "git rev-parse --show-prefix" used to emit nothing when run at the
  64   top-level of the working tree, but now it gives a blank line.
  65
  66 * Minor memory leak during unpack_trees (hence "merge" and "checkout"
  67   to check out another branch) has been plugged.
  68
  69 * More lower-level commands learned to use the streaming API to read
  70   from the object store without keeping everything in core.
  71
  72 * Because "sh" on the user's PATH may be utterly broken on some
  73   systems, run-command API now uses SHELL_PATH, not /bin/sh, when
  74   spawning an external command (not applicable to Windows port).
  75
  76 * The API to iterate over refs/ hierarchy has been tweaked to allow
  77   walking only a subset of it more efficiently.
  78
  79Also contains minor documentation updates and code clean-ups.
  80
  81
  82Fixes since v1.7.10
  83-------------------
  84
  85Unless otherwise noted, all the fixes since v1.7.10 in the maintenance
  86releases are contained in this release (see release notes to them for
  87details).
  88
  89 * The parser in "fast-import" did not diagnose ":9" style references
  90   that is not followed by required SP/LF as an error.
  91   (merge 06454cb pw/fast-import-dataref-parsing later to maint).
  92
  93 * When "git fetch" encounters repositories with too many references,
  94   the command line of "fetch-pack" that is run by a helper
  95   e.g. remote-curl, may fail to hold all of them. Now such an
  96   internal invocation can feed the references through the standard
  97   input of "fetch-pack".
  98   (merge 7103d25 it/fetch-pack-many-refs later to maint).
  99
 100 * "git fetch" that recurses into submodules on demand did not check
 101   if it needs to go into submodules when non branches (most notably,
 102   tags) are fetched.
 103   (merge a6801ad jl/maint-submodule-recurse-fetch later to maint).
 104
 105 * "git blame" started missing quite a few changes from the origin
 106   since we stopped using the diff minimalization by default in v1.7.2
 107   era.
 108   (merge 059a500 jc/maint-blame-minimal later to maint).
 109
 110 * "log -p --graph" used with "--stat" had a few formatting error.
 111   (merge e2c5966 lp/maint-diff-three-dash-with-graph later to maint).
 112
 113 * Giving "--continue" to a conflicted "rebase -i" session skipped a
 114   commit that only results in changes to submodules.
 115   (merge a6754cd jk/rebase-i-submodule-conflict-only later to maint).
 116
 117 * When PATH contains an unreadable directory, alias expansion code
 118   did not kick in, and failed with an error that said "git-subcmd"
 119   was not found.
 120   (merge 38f865c jk/run-command-eacces later to maint).
 121
 122 * The 'push to upstream' implementation was broken in some corner
 123   cases. "git push $there" without refspec, when the current branch
 124   is set to push to a remote different from $there, used to push to
 125   $there using the upstream information to a remote unreleated to
 126   $there.
 127   (merge 135dade jc/push-upstream-sanity later to maint).
 128
 129 * "git clean -d -f" (not "-d -f -f") is supposed to protect nested
 130   working trees of independent git repositories that exist in the
 131   current project working tree from getting removed, but the
 132   protection applied only to such working trees that are at the
 133   top-level of the current project by mistake.
 134   (merge ae2f203 jc/maint-clean-nested-worktree-in-subdir later to maint).
 135
 136 * Rename detection logic used to match two empty files as renames
 137   during merge-recursive, leading unnatural mismerges.
 138   (merge 4f7cb99 jk/diff-no-rename-empty later to maint).
 139
 140 * An age-old corner case bug in combine diff (only triggered with -U0
 141   and the hunk at the beginning of the file needs to be shown) has
 142   been fixed.
 143   (merge e5e9b56 rs/combine-diff-zero-context-at-the-beginning later to maint).
 144
 145 * When "git commit --template F" errors out because the user did not
 146   touch the message, it claimed that it aborts due to "empty
 147   message", which was utterly wrong.
 148   (merge 1f08c2c jc/commit-unedited-template later to maint).
 149
 150 * "git add -p" is not designed to deal with unmerged paths but did
 151   not exclude them and tried to apply funny patches only to fail.
 152   (merge 4066bd6 jk/add-p-skip-conflicts later to maint).
 153
 154 * "git commit --author=$name" did not tell the name that was being
 155   recorded in the resulting commit to hooks, even though it does do
 156   so when the end user overrode the authorship via the
 157   "GIT_AUTHOR_NAME" environment variable.
 158   (merge 7dfe8ad jc/commit-hook-authorship later to maint).
 159
 160 * The regexp configured with diff.wordregex was incorrectly reused
 161   across files.
 162   (merge 6440d34 tr/maint-word-diff-regex-sticky later to maint).
 163
 164 * Running "notes merge --commit" failed to perform correctly when run
 165   from any directory inside $GIT_DIR/.  When "notes merge" stops with
 166   conflicts, $GIT_DIR/NOTES_MERGE_WORKTREE is the place a user edits
 167   to resolve it.
 168   (merge dabba59 jh/notes-merge-in-git-dir-worktree later to maint).