1 Git v2.14.3 Release Notes
2 =========================
7 * A helper function to read a single whole line into strbuf
8 mistakenly triggered OOM error at EOF under certain conditions,
11 * In addition to "cc: <a@dd.re.ss> # cruft", "cc: a@dd.re.ss # cruft"
12 was taught to "git send-email" as a valid way to tell it that it
13 needs to also send a carbon copy to <a@dd.re.ss> in the trailer
16 * Fix regression to "gitk --bisect" by a recent update.
18 * Unlike "git commit-tree < file", "git commit-tree -F file" did not
19 pass the contents of the file verbatim and instead completed an
20 incomplete line at the end, if exists. The latter has been updated
21 to match the behaviour of the former.
23 * "git archive", especially when used with pathspec, stored an empty
24 directory in its output, even though Git itself never does so.
27 * API error-proofing which happens to also squelch warnings from GCC.
29 * "git gc" tries to avoid running two instances at the same time by
30 reading and writing pid/host from and to a lock file; it used to
31 use an incorrect fscanf() format when reading, which has been
34 * The test linter has been taught that we do not like "echo -e".
36 * Code cmp.std.c nitpick.
38 * "git describe --match" learned to take multiple patterns in v2.13
39 series, but the feature ignored the patterns after the first one
40 and did not work at all. This has been fixed.
42 * "git cat-file --textconv" started segfaulting recently, which
45 * The built-in pattern to detect the "function header" for HTML did
46 not match <H1>..<H6> elements without any attributes, which has
49 * "git mailinfo" was loose in decoding quoted printable and produced
50 garbage when the two letters after the equal sign are not
51 hexadecimal. This has been fixed.
53 * The documentation for '-X<option>' for merges was misleadingly
54 written to suggest that "-s theirs" exists, which is not the case.
56 * Spell the name of our system as "Git" in the output from
59 * Fixes for a handful memory access issues identified by valgrind.
61 * Backports a moral equivalent of 2015 fix to the poll emulation from
62 the upstream gnulib to fix occasional breakages on HPE NonStop.
64 * In the "--format=..." option of the "git for-each-ref" command (and
65 its friends, i.e. the listing mode of "git branch/tag"), "%(atom:)"
66 (e.g. "%(refname:)", "%(body:)" used to error out. Instead, treat
67 them as if the colon and an empty string that follows it were not
70 * Users with "color.ui = always" in their configuration were broken
71 by a recent change that made plumbing commands to pay attention to
72 them as the patch created internally by "git add -p" were colored
73 (heh) and made unusable. This has been fixed.
75 * "git branch -M a b" while on a branch that is completely unrelated
76 to either branch a or branch b misbehaved when multiple worktree
77 was in use. This has been fixed.
79 * "git fast-export" with -M/-C option issued "copy" instruction on a
80 path that is simultaneously modified, which was incorrect.
82 * The checkpoint command "git fast-import" did not flush updates to
83 refs and marks unless at least one object was created since the
84 last checkpoint, which has been corrected, as these things can
85 happen without any new object getting created.
87 * The scripts to drive TravisCI has been reorganized and then an
88 optimization to avoid spending cycles on a branch whose tip is
89 tagged has been implemented.
91 * "git fetch <there> <src>:<dst>" allows an object name on the <src>
92 side when the other side accepts such a request since Git v2.5, but
93 the documentation was left stale.
95 * A regression in 2.11 that made the code to read the list of
96 alternate object stores overrun the end of the string has been
99 Also contains various documentation updates and code clean-ups.