| blob | cfe94b99182d86d23e2e14e952e01549a16c5a5a |
1 Git 2.11 Release Notes
2 ======================
4 Updates since v2.10
5 -------------------
7 UI, Workflows & Features
9 * "git format-patch --cover-letter HEAD^" to format a single patch
10 with a separate cover letter now numbers the output as [PATCH 0/1]
11 and [PATCH 1/1] by default.
13 * An incoming "git push" that attempts to push too many bytes can now
14 be rejected by setting a new configuration variable at the receiving
15 end.
17 * "git nosuchcommand --help" said "No manual entry for gitnosuchcommand",
18 which was not intuitive, given that "git nosuchcommand" said "git:
19 'nosuchcommand' is not a git command".
21 * "git clone --resurse-submodules --reference $path $URL" is a way to
22 reduce network transfer cost by borrowing objects in an existing
23 $path repository when cloning the superproject from $URL; it
24 learned to also peek into $path for presense of corresponding
25 repositories of submodules and borrow objects from there when able.
27 * The "git diff --submodule={short,log}" mechanism has been enhanced
28 to allow "--submodule=diff" to show the patch between the submodule
29 commits bound to the superproject.
31 * Even though "git hash-objects", which is a tool to take an
32 on-filesystem data stream and put it into the Git object store,
33 allowed to perform the "outside-world-to-Git" conversions (e.g.
34 end-of-line conversions and application of the clean-filter), and
35 it had the feature on by default from very early days, its reverse
36 operation "git cat-file", which takes an object from the Git object
37 store and externalize for the consumption by the outside world,
38 lacked an equivalent mechanism to run the "Git-to-outside-world"
39 conversion. The command learned the "--filters" option to do so.
41 * Output from "git diff" can be made easier to read by selecting
42 which lines are common and which lines are added/deleted
43 intelligently when the lines before and after the changed section
44 are the same. A command line option is added to help with the
45 experiment to find a good heuristics.
47 * In some projects, it is common to use "[RFC PATCH]" as the subject
48 prefix for a patch meant for discussion rather than application. A
49 new option "--rfc" was a short-hand for "--subject-prefix=RFC PATCH"
50 to help the participants of such projects.
52 * "git add --chmod=+x <pathspec>" added recently only toggled the
53 executable bit for paths that are either new or modified. This has
54 been corrected to flip the executable bit for all paths that match
55 the given pathspec.
57 * When "git format-patch --stdout" output is placed as an in-body
58 header and it uses the RFC2822 header folding, "git am" failed to
59 put the header line back into a single logical line. The
60 underlying "git mailinfo" was taught to handle this properly.
63 Performance, Internal Implementation, Development Support etc.
65 * The delta-base-cache mechanism has been a key to the performance in
66 a repository with a tightly packed packfile, but it did not scale
67 well even with a larger value of core.deltaBaseCacheLimit.
69 * Enhance "git status --porcelain" output by collecting more data on
70 the state of the index and the working tree files, which may
71 further be used to teach git-prompt (in contrib/) to make fewer
72 calls to git.
74 * Extract a small helper out of the function that reads the authors
75 script file "git am" internally uses.
76 (merge a77598e jc/am-read-author-file later to maint).
78 * Lifts calls to exit(2) and die() higher in the callchain in
79 sequencer.c files so that more helper functions in it can be used
80 by callers that want to handle error conditions themselves.
82 * "git am" has been taught to make an internal call to "git apply"'s
83 innards without spawning the latter as a separate process.
85 * The ref-store abstraction was introduced to the refs API so that we
86 can plug in different backends to store references.
88 * The "unsigned char sha1[20]" to "struct object_id" conversion
89 continues. Notable changes in this round includes that ce->sha1,
90 i.e. the object name recorded in the cache_entry, turns into an
91 object_id.
93 * JGit can show a fake ref "capabilities^{}" to "git fetch" when it
94 does not advertise any refs, but "git fetch" was not prepared to
95 see such an advertisement. When the other side disconnects without
96 giving any ref advertisement, we used to say "there may not be a
97 repository at that URL", but we may have seen other advertisement
98 like "shallow" and ".have" in which case we definitely know that a
99 repository is there. The code to detect this case has also been
100 updated.
102 * Some codepaths in "git pack-objects" were not ready to use an
103 existing pack bitmap; now they are and as the result they have
104 become faster.
107 Also contains various documentation updates and code clean-ups.
110 Fixes since v2.10
111 -----------------
113 Unless otherwise noted, all the fixes since v2.9 in the maintenance
114 track are contained in this release (see the maintenance releases'
115 notes for details).
117 * Clarify various ways to specify the "revision ranges" in the
118 documentation.
120 * "diff-highlight" script (in contrib/) learned to work better with
121 "git log -p --graph" output.
123 * The test framework left the number of tests and success/failure
124 count in the t/test-results directory, keyed by the name of the
125 test script plus the process ID. The latter however turned out not
126 to serve any useful purpose. The process ID part of the filename
127 has been removed.
129 * Having a submodule whose ".git" repository is somehow corrupt
130 caused a few commands that recurse into submodules loop forever.
132 * "git symbolic-ref -d HEAD" happily removes the symbolic ref, but
133 the resulting repository becomes an invalid one. Teach the command
134 to forbid removal of HEAD.
136 * A test spawned a short-lived background process, which sometimes
137 prevented the test directory from getting removed at the end of the
138 script on some platforms.
140 * Update a few tests that used to use GIT_CURL_VERBOSE to use the
141 newer GIT_TRACE_CURL.
143 * "git pack-objects --include-tag" was taught that when we know that
144 we are sending an object C, we want a tag B that directly points at
145 C but also a tag A that points at the tag B. We used to miss the
146 intermediate tag B in some cases.
147 (merge b773dde jk/pack-tag-of-tag later to maint).
149 * Update Japanese translation for "git-gui".
150 (merge 02748bc sy/git-gui-i18n-ja later to maint).
152 * "git fetch http::/site/path" did not die correctly and segfaulted
153 instead.
154 (merge d63ed6e jk/fix-remote-curl-url-wo-proto later to maint).
156 * "git commit-tree" stopped reading commit.gpgsign configuration
157 variable that was meant for Porcelain "git commit" in Git 2.9; we
158 forgot to update "git gui" to look at the configuration to match
159 this change.
160 (merge f14a310 js/git-gui-commit-gpgsign later to maint).
162 * "git add --chmod=+x" added recently lacked documentation, which has
163 been corrected.
164 (merge 7ef7903 et/add-chmod-x later to maint).
166 * "git log --cherry-pick" used to include merge commits as candidates
167 to be matched up with other commits, resulting a lot of wasted time.
168 The patch-id generation logic has been updated to ignore merges to
169 avoid the wastage.
170 (merge 7c81040 jk/patch-ids-no-merges later to maint).
172 * The http transport (with curl-multi option, which is the default
173 these days) failed to remove curl-easy handle from a curlm session,
174 which led to unnecessary API failures.
175 (merge 2abc848 ew/http-do-not-forget-to-call-curl-multi-remove-handle later to maint).
177 * There were numerous corner cases in which the configuration files
178 are read and used or not read at all depending on the directory a
179 Git command was run, leading to inconsistent behaviour. The code
180 to set-up repository access at the beginning of a Git process has
181 been updated to fix them.
182 (merge 4d0efa1 jk/setup-sequence-update later to maint).
184 * "git diff -W" output needs to extend the context backward to
185 include the header line of the current function and also forward to
186 include the body of the entire current function up to the header
187 line of the next one. This process may have to merge to adjacent
188 hunks, but the code forgot to do so in some cases.
189 (merge 45d2f75 rs/xdiff-merge-overlapping-hunks-for-W-context later to maint).
191 * Performance tests done via "t/perf" did not use the same set of
192 build configuration if the user relied on autoconf generated
193 configuration.
194 (merge cd5c281 ks/perf-build-with-autoconf later to maint).
196 * "git format-patch --base=..." feature that was recently added
197 showed the base commit information after "-- " e-mail signature
198 line, which turned out to be inconvenient. The base information
199 has been moved above the signature line.
200 (merge 480871e jt/format-patch-base-info-above-sig later to maint).
202 * More i18n.
203 (merge 43073f8 va/i18n later to maint).
205 * Even when "git pull --rebase=preserve" (and the underlying "git
206 rebase --preserve") can complete without creating any new commit
207 (i.e. fast-forwards), it still insisted on having a usable ident
208 information (read: user.email is set correctly), which was less
209 than nice. As the underlying commands used inside "git rebase"
210 would fail with a more meaningful error message and advice text
211 when the bogus ident matters, this extra check was removed.
212 (merge 1e461c4 jk/rebase-i-drop-ident-check later to maint).
214 * "git gc --aggressive" used to limit the delta-chain length to 250,
215 which is way too deep for gaining additional space savings and is
216 detrimental for runtime performance. The limit has been reduced to
217 50.
218 (merge 07e7dbf jk/reduce-gc-aggressive-depth later to maint).
220 * Documentation for individual configuration variables to control use
221 of color (like `color.grep`) said that their default value is
222 'false', instead of saying their default is taken from `color.ui`.
223 When we updated the default value for color.ui from 'false' to
224 'auto' quite a while ago, all of them broke. This has been
225 corrected.
226 (merge 14d16e2 mm/config-color-ui-default-to-auto later to maint).
228 * The pretty-format specifier "%C(auto)" used by the "log" family of
229 commands to enable coloring of the output is taught to also issue a
230 color-reset sequence to the output.
231 (merge c99ad27 rs/c-auto-resets-attributes later to maint).
233 * A shell script example in check-ref-format documentation has been
234 fixed.
235 (merge 92dece7 ep/doc-check-ref-format-example later to maint).
237 * "git checkout <word>" does not follow the usual disambiguation
238 rules when the <word> can be both a rev and a path, to allow
239 checking out a branch 'foo' in a project that happens to have a
240 file 'foo' in the working tree without having to disambiguate.
241 This was poorly documented and the check was incorrect when the
242 command was run from a subdirectory.
243 (merge b829b94 nd/checkout-disambiguation later to maint).
245 * Some codepaths in "git diff" used regexec(3) on a buffer that was
246 mmap(2)ed, which may not have a terminating NUL, leading to a read
247 beyond the end of the mapped region. This was fixed by introducing
248 a regexec_buf() helper that takes a <ptr,len> pair with REG_STARTEND
249 extension.
250 (merge b7d36ff js/regexec-buf later to maint).
252 * The procedure to build Git on Mac OS X for Travis CI hardcoded the
253 internal directory structure we assumed HomeBrew uses, which was a
254 no-no. The procedure has been updated to ask HomeBrew things we
255 need to know to fix this.
256 (merge f86f49b ls/travis-homebrew-path-fix later to maint).
258 * When "git rebase -i" is given a broken instruction, it told the
259 user to fix it with "--edit-todo", but didn't say what the step
260 after that was (i.e. "--continue").
261 (merge 37875b4 rt/rebase-i-broken-insn-advise later to maint).
263 * Documentation around tools to import from CVS was fairly outdated.
264 (merge 106b672 jk/doc-cvs-update later to maint).
266 * "git clone --recurse-submodules" lost the progress eye-candy in
267 recent update, which has been corrected.
269 * A low-level function verify_packfile() was meant to show errors
270 that were detected without dying itself, but under some conditions
271 it didn't and died instead, which has been fixed.
272 (merge a9445d859e jk/verify-packfile-gently later to maint).
274 * When "git fetch" tries to find where the history of the repository
275 it runs in has diverged from what the other side has, it has a
276 mechanism to avoid digging too deep into irrelevant side branches.
277 This however did not work well over the "smart-http" transport due
278 to a design bug, which has been fixed.
279 (merge 06b3d386e0 jt/fetch-pack-in-vain-count-with-stateless later to maint).
281 * In the codepath that comes up with the hostname to be used in an
282 e-mail when the user didn't tell us, we looked at ai_canonname
283 field in struct addrinfo without making sure it is not NULL first.
284 (merge c375a7efa3 jk/ident-ai-canonname-could-be-null later to maint).
286 * "git worktree", even though it used the default_abbrev setting that
287 ought to be affected by core.abbrev configuration variable, ignored
288 the variable setting. The command has been taught to read the
289 default set of configuration variables to correct this.
290 (merge d49028e6e7 jc/worktree-config later to maint).
292 * Other minor doc, test and build updates and code cleanups.
293 (merge e78d57e bw/pathspec-remove-unused-extern-decl later to maint).
294 (merge ce25e4c rs/checkout-some-states-are-const later to maint).
295 (merge a8342a4 rs/strbuf-remove-fix later to maint).
296 (merge b56aa5b rs/unpack-trees-reduce-file-scope-global later to maint).
297 (merge 5efc60c mr/vcs-svn-printf-ulong later to maint).
298 (merge a22ae75 rs/cocci later to maint).