| blob | e4ca72e50e16bd4c5a04e31cc368d8e4ce4c766a |
1 Git 2.14 Release Notes
2 ======================
4 Backward compatibility notes.
6 * Use of an empty string as a pathspec element that is used for
7 'everything matches' is still warned and Git asks users to use a
8 more explicit '.' for that instead. The hope is that existing
9 users will not mind this change, and eventually the warning can be
10 turned into a hard error, upgrading the deprecation into removal of
11 this (mis)feature. That is not scheduled to happen in the upcoming
12 release (yet).
14 * Git now avoids blindly falling back to ".git" when the setup
15 sequence said we are _not_ in Git repository. A corner case that
16 happens to work right now may be broken by a call to die("BUG").
17 We've tried hard to locate such cases and fixed them, but there
18 might still be cases that need to be addressed--bug reports are
19 greatly appreciated.
21 * The experiment to improve the hunk-boundary selection of textual
22 diff output has finished, and the "indent heuristics" has now
23 become the default.
26 Updates since v2.13
27 -------------------
29 UI, Workflows & Features
31 * The colors in which "git status --short --branch" showed the names
32 of the current branch and its remote-tracking branch are now
33 configurable.
35 * "git clone" learned the "--no-tags" option not to fetch all tags
36 initially, and also set up the tagopt not to follow any tags in
37 subsequent fetches.
39 * "git archive --format=zip" learned to use zip64 extension when
40 necessary to go beyond the 4GB limit.
41 (merge 867e40ff3a rs/large-zip later to maint).
43 * "git reset" learned "--recurse-submodules" option.
45 * "git diff --submodule=diff" now recurses into nested submodules.
46 (merge 5a5221427c jk/diff-submodule-diff-inline later to maint).
48 * "git repack" learned to accept the --threads=<n> option and pass it
49 to pack-objects.
51 * "git send-email" learned to run sendemail-validate hook to inspect
52 and reject a message before sending it out.
53 (merge 177409e589 jt/send-email-validate-hook later to maint).
55 * There is no good reason why "git fetch $there $sha1" should fail
56 when the $sha1 names an object at the tip of an advertised ref,
57 even when the other side hasn't enabled allowTipSHA1InWant.
59 * The recently introduced "[includeIf "gitdir:$dir"] path=..."
60 mechansim has further been taught to take symlinks into account.
61 The directory "$dir" specified in "gitdir:$dir" may be a symlink to
62 a real location, not something that $(getcwd) may return. In such
63 a case, a realpath of "$dir" is compared with the real path of the
64 current repository to determine if the contents from the named path
65 should be included.
67 * Make the "indent" heuristics the default in "diff" and diff.indentHeuristics
68 configuration variable an escape hatch for those who do no want it.
71 Performance, Internal Implementation, Development Support etc.
73 * The default packed-git limit value has been raised on larger
74 platforms to save "git fetch" from a (recoverable) failure while
75 "gc" is running in parallel.
77 * Code to update the cache-tree has been tightened so that we won't
78 accidentally write out any 0{40} entry in the tree object.
79 (merge a96d3cc3f6 jk/no-null-sha1-in-cache-tree later to maint).
81 * Attempt to allow us notice "fishy" situation where we fail to
82 remove the temporary directory used during the test.
84 * Travis CI gained a task to format the documentation with both
85 AsciiDoc and AsciiDoctor.
87 * Some platforms have ulong that is smaller than time_t, and our
88 historical use of ulong for timestamp would mean they cannot
89 represent some timestamp that the platform allows. Invent a
90 separate and dedicated timestamp_t (so that we can distingiuish
91 timestamps and a vanilla ulongs, which along is already a good
92 move), and then declare uintmax_t is the type to be used as the
93 timestamp_t.
95 * We can trigger Windows auto-build tester (credits: Dscho &
96 Microsoft) from our existing Travis CI tester now.
98 * Conversion from uchar[20] to struct object_id continues.
100 * Simplify parse_pathspec() codepath and stop it from looking at the
101 default in-core index.
102 (merge 08de9151a8 bw/pathspec-sans-the-index later to maint).
104 * Add perf-test for wildmatch.
105 (merge 62ca75a6b9 ab/perf-wildmatch later to maint).
107 * Code from "conversion using external process" codepath has been
108 extracted to a separate sub-process.[ch] module.
109 (merge 4f2a2e9f0e bp/sub-process-convert-filter later to maint).
111 * When "git checkout", "git merge", etc. manipulates the in-core
112 index, various pieces of information in the index extensions are
113 discarded from the original state, as it is usually not the case
114 that they are kept up-to-date and in-sync with the operation on the
115 main index. The untracked cache extension is copied across these
116 operations now, which would speed up "git status" (as long as the
117 cache is properly invalidated).
119 * The internal implementation of "git grep" has seen some clean-up.
120 (merge 8df4c2953f ab/grep-preparatory-cleanup later to maint).
122 * Update the C style recommendation for notes for translators, as
123 recent versions of gettext tools can work with our style of
124 multi-line comments.
126 * The implementation of "ref" API around the "packed refs" have been
127 cleaned up, in preparation for further changes.
129 * The internal logic used in "git blame" has been libified to make it
130 easier to use by cgit.
133 Also contains various documentation updates and code clean-ups.
136 Fixes since v2.13
137 -----------------
139 Unless otherwise noted, all the fixes since v2.13 in the maintenance
140 track are contained in this release (see the maintenance releases'
141 notes for details).
143 * "git gc" did not interact well with "git worktree"-managed
144 per-worktree refs.
146 * "git cherry-pick" and other uses of the sequencer machinery
147 mishandled a trailer block whose last line is an incomplete line.
148 This has been fixed so that an additional sign-off etc. are added
149 after completing the existing incomplete line.
151 * The codepath in "git am" that is used when running "git rebase"
152 leaked memory held for the log message of the commits being rebased.
154 * "git clone --config var=val" is a way to populate the
155 per-repository configuration file of the new repository, but it did
156 not work well when val is an empty string. This has been fixed.
158 * Setting "log.decorate=false" in the configuration file did not take
159 effect in v2.13, which has been corrected.
161 * A few codepaths in "checkout" and "am" working on an unborn branch
162 tried to access an uninitialized piece of memory.
164 * The Web interface to gmane news archive is long gone, even though
165 the articles are still accessible via NTTP. Replace the links with
166 ones to public-inbox.org. Because their message identification is
167 based on the actual message-id, it is likely that it will be easier
168 to migrate away from it if/when necessary.
170 * The receive-pack program now makes sure that the push certificate
171 records the same set of push options used for pushing.
173 * Tests have been updated to pass under GETTEXT_POISON (a mechanism
174 to ensure that output strings that should not be translated are
175 not translated by mistake), and TravisCI is told to run them.
177 * "git checkout --recurse-submodules" did not quite work with a
178 submodule that itself has submodules.
179 (merge 218c883783 sb/checkout-recurse-submodules later to maint).
181 * Plug some leaks and updates internal API used to implement the
182 split index feature to make it easier to avoid such a leak in the
183 future.
184 (merge de950c5773 nd/split-index-unshare later to maint).
186 * "pack-objects" can stream a slice of an existing packfile out when
187 the pack bitmap can tell that the reachable objects are all needed
188 in the output, without inspecting individual objects. This
189 strategy however would not work well when "--local" and other
190 options are in use, and need to be disabled.
192 * Fix memory leaks pointed out by Coverity (and people).
194 * "git read-tree -m" (no tree-ish) gave a nonsense suggestion "use
195 --empty if you want to clear the index". With "-m", such a request
196 will still fail anyway, as you'd need to name at least one tree-ish
197 to be merged.
199 * Make sure our tests would pass when the sources are checked out
200 with "platform native" line ending convention by default on
201 Windows. Some "text" files out tests use and the test scripts
202 themselves that are meant to be run with /bin/sh, ought to be
203 checked out with eol=LF even on Windows.
205 * Introduce the BUG() macro to improve die("BUG: ...").
206 (merge 3d7dd2d3b6 jk/bug-to-abort later to maint).
208 * Clarify documentation for include.path and includeIf.<condition>.path
209 configuration variables.
211 * Git sometimes gives an advice in a rhetorical question that does
212 not require an answer, which can confuse new users and non native
213 speakers. Attempt to rephrase them.
215 * A few http:// links that are redirected to https:// in the
216 documentation have been updated to https:// links.
218 * "git for-each-ref --format=..." with %(HEAD) in the format used to
219 resolve the HEAD symref as many times as it had processed refs,
220 which was wasteful, and "git branch" shared the same problem.
222 * Regression fix to topic recently merged to 'master'.
224 * The shell completion script (in contrib/) learned "git stash" has
225 a new "push" subcommand.
227 * "git interpret-trailers", when used as GIT_EDITOR for "git commit
228 -v", looked for and appended to a trailer block at the very end,
229 i.e. at the end of the "diff" output. The command has been
230 corrected to pay attention to the cut-mark line "commit -v" adds to
231 the buffer---the real trailer block should appear just before it.
233 * A test allowed both "git push" and "git receive-pack" on the other
234 end write their traces into the same file. This is OK on platforms
235 that allows atomically appending to a file opened with O_APPEND,
236 but on other platforms led to a mangled output, causing
237 intermittent test failures. This has been fixed by disabling
238 traces from "receive-pack" in the test.
240 * Tag objects, which are not reachable from any ref, that point at
241 missing objects were mishandled by "git gc" and friends (they
242 should silently be ignored instead)
244 * "git describe --contains" penalized light-weight tags so much that
245 they were almost never considered. Instead, give them about the
246 same chance to be considered as an annotated tag that is the same
247 age as the underlying commit would.
248 (merge ef1e74065c jc/name-rev-lw-tag later to maint).
250 * The "run-command" API implementation has been made more robust
251 against dead-locking in a threaded environment.
252 (merge e3f43ce765 bw/forking-and-threading later to maint).
254 * A recent update to t5545-push-options.sh started skipping all the
255 tests in the script when a web server testing is disabled or
256 unavailable, not just the ones that require a web server. Non HTTP
257 tests have been salvaged to always run in this script.
258 (merge 2e397e4ddf jc/skip-test-in-the-middle later to maint).
260 * "git send-email" now uses Net::SMTP::SSL, which is obsolete, only
261 when needed. Recent versions of Net::SMTP can do TLS natively.
262 (merge bfbfc9a953 dk/send-email-avoid-net-smtp-ssl-when-able later to maint).
264 * "foo\bar\baz" in "git fetch foo\bar\baz", even though there is no
265 slashes in it, cannot be a nickname for a remote on Windows, as
266 that is likely to be a pathname on a local filesystem.
268 * "git clean -d" used to clean directories that has ignored files,
269 even though the command should not lose ignored ones without "-x".
270 "git status --ignored" did not list ignored and untracked files
271 without "-uall". These have been corrected.
272 (merge 6b1db43109 sl/clean-d-ignored-fix later to maint).
274 * The result from "git diff" that compares two blobs, e.g. "git diff
275 $commit1:$path $commit2:$path", used to be shown with the full
276 object name as given on the command line, but it is more natural to
277 use the $path in the output and use it to look up .gitattributes.
278 (merge 30d005c020 jk/diff-blob later to maint).
280 * The "collision detecting" SHA-1 implementation shipped with 2.13
281 was quite broken on some big-endian platforms and/or platforms that
282 do not like unaligned fetches. Update to the upstream code which
283 has already fixed these issues.
285 * "git am -h" triggered a BUG().
287 * The interaction of "url.*.insteadOf" and custom URL scheme's
288 whitelisting is now documented better.
290 * The timestamp of the index file is now taken after the file is
291 closed, to help Windows, on which a stale timestamp is reported by
292 fstat() on a file that is opened for writing and data was written
293 but not yet closed.
294 (merge 9f41c7a6b3 jh/close-index-before-stat later to maint).
296 * "git pull --rebase --autostash" didn't auto-stash when the local history
297 fast-forwards to the upstream.
298 (merge f15e7cf5cc tb/pull-ff-rebase-autostash later to maint).
300 * Other minor doc, test and build updates and code cleanups.
301 (merge c5a9157393 jh/memihash-opt later to maint).
302 (merge 44e2ff09ce ab/t3070-test-dedup later to maint).
303 (merge 9ee4aa95db rf/completion-config-commit later to maint).
304 (merge ef4fe5617e jk/connect-symref-info-leak-fix later to maint).
305 (merge a56eea28c4 jk/drop-free-refspecs later to maint).
306 (merge 0c79cee697 ad/pull-remote-doc later to maint).