| blob | 15ad460b594b5b1243cfb1f611f45ab5962a0d8f |
1 Git v1.8.3 Release Notes
2 ========================
4 Backward compatibility notes (for Git 2.0)
5 ------------------------------------------
7 When "git push [$there]" does not say what to push, we have used the
8 traditional "matching" semantics so far (all your branches were sent
9 to the remote as long as there already are branches of the same name
10 over there). In Git 2.0, the default will change to the "simple"
11 semantics that pushes the current branch to the branch with the same
12 name, only when the current branch is set to integrate with that
13 remote branch. There is a user preference configuration variable
14 "push.default" to change this. If you are an old-timer who is used
15 to the "matching" semantics, you can set it to "matching" to keep the
16 traditional behaviour. If you want to live in the future early,
17 you can set it to "simple" today without waiting for Git 2.0.
19 When "git add -u" and "git add -A", that does not specify what paths
20 to add on the command line is run from inside a subdirectory, these
21 commands will operate on the entire tree in Git 2.0 for consistency
22 with "git commit -a" and other commands. Because there will be no
23 mechanism to make "git add -u" behave as if "git add -u .", it is
24 important for those who are used to "git add -u" (without pathspec)
25 updating the index only for paths in the current subdirectory to start
26 training their fingers to explicitly say "git add -u ." when they mean
27 it before Git 2.0 comes. A warning is issued when these commands are
28 run without a pathspec and when you have local changes outside the
29 current directory, because the behaviour in Git 2.0 will be different
30 from today's version in such a situation.
33 Updates since v1.8.2
34 --------------------
36 Foreign interface
38 * remote-hg helper (in contrib/) has been updated.
41 UI, Workflows & Features
43 * "git branch --vv" learned to paint the name of the branch it
44 integrates with in a different color (color.branch.upstream,
45 which defaults to blue).
47 * In a sparsely populated working tree, "git checkout <pathspec>" no
48 longer unmarks paths that match the given pathspec that were
49 originally ignored with "--sparse" (use --ignore-skip-worktree-bits
50 option to resurrect these paths out of the index if you really want
51 to).
53 * "git bisect" leaves the final outcome as a comment in its bisect
54 log file.
56 * "git clone --reference" can now refer to a gitfile "textual symlink"
57 that points at the real location of the repository.
59 * "git count-objects" learned "--human-readable" aka "-H" option to
60 show various large numbers in Ki/Mi/GiB scaled as necessary.
62 * "git cherry-pick $blob" and "git cherry-pick $tree" are nonsense,
63 and a more readable error message e.g. "can't cherry-pick a tree"
64 is given (we used to say "expected exactly one commit").
66 * The "--annotate" option to "git send-email" can be turned on (or
67 off) by default with sendemail.annotate configuration variable (you
68 can use --no-annotate from the command line to override it).
70 * The "--cover-letter" option to "git format-patch" can be turned on
71 (or off) by default with format.coverLetter configuration
72 variable. By setting it to 'auto', you can turn it on only for a
73 series with two or more patches.
75 * The bash completion support (in contrib/) learned that cherry-pick
76 takes a few more options than it already knew about.
78 * "git help" learned "-g" option to show the list of guides just like
79 list of commands are given with "-a".
81 * A triangular "pull from one place, push to another place" workflow
82 is supported better by new remote.pushdefault (overrides the
83 "origin" thing) and branch.*.pushremote (overrides the
84 branch.*.remote) configuration variables.
86 * "git status" learned to report that you are in the middle of a
87 revert session, just like it does for a cherry-pick and a bisect
88 session.
90 * The handling by "git branch --set-upstream-to" against various forms
91 of erroneous inputs was suboptimal and has been improved.
93 * When the interactive access to git-shell is not enabled, it issues
94 a message meant to help the system administrator to enable it.
95 An explicit way to help the end users who connect to the service by
96 issuing custom messages to refuse such an access has been added.
98 * In addition to the case where the user edits the log message with
99 the "e)dit" option of "am -i", replace the "Applying: this patch"
100 message with the final log message contents after applymsg hook
101 munges it.
103 * "git status" suggests users to look into using --untracked=no option
104 when it takes too long.
106 * "git status" shows a bit more information to "git status" during a
107 rebase/bisect session.
109 * "git fetch" learned to fetch a commit at the tip of an unadvertised
110 ref by specifying a raw object name from the command line when the
111 server side supports this feature.
113 * Output from "git log --graph" works better with submodule log
114 output now.
116 * "git count-objects -v" learned to report leftover temporary
117 packfiles and other garbage in the object store.
119 * A new read-only credential helper (in contrib/) to interact with
120 the .netrc/.authinfo files has been added.
122 * "git send-email" can be used with the credential helper system.
124 * There was no Porcelain way to say "I no longer am interested in
125 this submodule", once you express your interest in a submodule with
126 "submodule init". "submodule deinit" is the way to do so.
128 * "git pull --rebase" learned to pass "-v/-q" options to underlying
129 "git rebase".
131 * The new "--follow-tags" option tells "git push" to push relevant
132 annotated tags when pushing branches out.
134 * "git merge" and "git pull" can optionally be told to inspect and
135 reject when merging a commit that does not carry a trusted GPG
136 signature.
138 * "git mergetool" now feeds files to the "p4merge" backend in the
139 order that matches the p4 convention, where "theirs" is usually
140 shown on the left side, which is the opposite from other backend
141 expects.
143 * "show/log" now honors gpg.program configuration just like other
144 parts of the code that use GnuPG.
146 * "git log" that shows the difference between the parent and the
147 child has been optimized somewhat.
149 * "git difftool" allows the user to write into the temporary files
150 being shown; if the user makes changes to the working tree at the
151 same time, one of the changes has to be lost in such a case, but it
152 tells the user what happened and refrains from overwriting the copy
153 in the working tree.
155 * There was no good way to ask "I have a random string that came from
156 outside world. I want to turn it into a 40-hex object name while
157 making sure such an object exists". A new peeling suffix ^{object}
158 can be used for that purpose, together with "rev-parse --verify".
161 Performance, Internal Implementation, etc.
163 * Updates for building under msvc.
165 * The stack footprint of some codepaths that access an object from a
166 pack has been shrunk.
168 * The logic to coalesce the same lines removed from the parents in
169 the output from "diff -c/--cc" has been updated, but with an O(n^2)
170 complexity, so this might turn out to be undesirable.
172 * The code to enforce permission bits on files in $GIT_DIR/ for
173 shared repositories have been simplified.
175 * A few codepaths knew how much data they need to put in the
176 hashtables they use upfront, but still started from a small table
177 repeatedly growing and rehashing.
179 * The API to walk reflog entries from the latest to older, which was
180 necessary for operations such as "git checkout -", was cumbersome
181 to use correctly and also inefficient.
183 * Codepaths that inspect log-message-to-be and decide when to add a
184 new Signed-off-by line in various commands have been consolidated.
186 * The pkt-line API, implementation and its callers have been cleaned
187 up to make them more robust.
189 * Cygwin port has a faster-but-lying lstat(2) emulation whose
190 incorrectness does not matter in practice except for a few
191 codepaths, and setting permission bits to directories is a codepath
192 that needs to use a more correct one.
194 * "git checkout" had repeated pathspec matches on the same paths,
195 which have been consolidated. Also a bug in "git checkout dir/"
196 that is started from an unmerged index has been fixed.
198 * A few bugfixes to "git rerere" working on corner case merge
199 conflicts have been applied.
202 Also contains various documentation updates and code clean-ups.
205 Fixes since v1.8.2
206 ------------------
208 Unless otherwise noted, all the fixes since v1.8.2 in the maintenance
209 track are contained in this release (see release notes to them for
210 details).
212 * A commit object whose author or committer ident are malformed
213 crashed some code that trusted that a name, an email and an
214 timestamp can always be found in it.
215 (merge de5abe9 jk/chopped-ident later to maint).
217 * When "upload-pack" fails while generating a pack in response to
218 "git fetch" (or "git clone"), the receiving side mistakenly said
219 there was a programming error to trigger the die handler
220 recursively.
221 (merge 1ece66b jk/a-thread-only-dies-once later to maint).
223 * "rev-list --stdin" and friends kept bogus pointers into input
224 buffer around as human readble object names. This was not a huge
225 problem but was exposed by a new change that uses these names in
226 error output.
227 (merge 70d26c6 tr/copy-revisions-from-stdin later to maint).
229 * Smart-capable HTTP servers were not restricted via the
230 GIT_NAMESPACE mechanism when talking with commit-walker clients,
231 like they do when talking with smart HTTP clients.
232 (merge 6130f86 jk/http-dumb-namespaces later to maint).
234 * "git merge-tree" did not omit a merge result that is identical to
235 "our" side in certain cases.
236 (merge aacecc3 jk/merge-tree-added-identically later to maint).
238 * Perl scripts like "git-svn" closed (not redirecting to /dev/null)
239 the standard error stream, which is not a very smart thing to do.
240 Later open may return file descriptor #2 for unrelated purpose, and
241 error reporting code may write into them.
242 (merge a749c0b tr/perl-keep-stderr-open later to maint).
244 * "git show-branch" was not prepared to show a very long run of
245 ancestor operators e.g. foobar^2~2^2^2^2...^2~4 correctly.
246 (merge aaa07e3 jk/show-branch-strbuf later to maint).
248 * "git diff --diff-algorithm algo" is also understood as "git diff
249 --diff-algorithm=algo".
250 (merge 0895c6d jk/diff-algo-finishing-touches later to maint).
252 * The new core.commentchar configuration was not applied to a few
253 places.
254 (merge 89c3bbd rt/commentchar-fmt-merge-msg later to maint).
256 * "git bundle" did not like a bundle created using a commit without
257 any message as its one of the prerequistes.
258 (merge 5446e33 lf/bundle-with-tip-wo-message later to maint).
260 * "git log -S/-G" started paying attention to textconv filter, but
261 there was no way to disable this. Make it honor --no-textconv
262 option.
263 (merge 61690bf sr/log-SG-no-textconv later to maint).
265 * When used with "-d temporary-directory" option, "git filter-branch"
266 failed to come back to the original working tree to perform the
267 final clean-up procedure.
268 (merge 9727601 jk/filter-branch-come-back-to-original later to maint).
270 * "git merge $(git rev-parse v1.8.2)" behaved quite differently from
271 "git merge v1.8.2", as if v1.8.2 were written as v1.8.2^0 and did
272 not pay much attention to the annotated tag payload. Make the code
273 notice the type of the tag object, in addition to the dwim_ref()
274 based classification the current code uses (i.e. the name appears
275 in refs/tags/) to decide when to special case merging of tags.
276 (merge a38d3d7 jc/merge-tag-object later to maint).
278 * Fix 1.8.1.x regression that stopped matching "dir" (without
279 trailing slash) to a directory "dir".
280 (merge efa5f82 jc/directory-attrs-regression-fix later to maint-1.8.1).
282 * "git apply --whitespace=fix" was not prepared to see a line getting
283 longer after fixing whitespaces (e.g. tab-in-indent aka Python).
284 (merge 329b26e jc/apply-ws-fix-tab-in-indent later to maint-1.8.1).
286 * The prompt string generator (in contrib/completion/) did not notice
287 when we are in a middle of a "git revert" session.
288 (merge 3ee4452 rr/prompt-revert-head later to maint).
290 * "submodule summary --summary-limit" option did not support
291 "--option=value" form.
292 (merge 862ae6c rs/submodule-summary-limit later to maint).
294 * "index-pack --fix-thin" used an uninitialized value to compute
295 delta depths of objects it appends to the resulting pack.
296 (merge 57165db jk/index-pack-correct-depth-fix later to maint).
298 * "index-pack --verify-stat" used a few counters outside protection
299 of mutex, possibly showing incorrect numbers.
300 (merge 8f82aad nd/index-pack-threaded-fixes later to maint).
302 * The code to keep track of what directory names are known to Git on
303 platforms with case insensitive filesystems can get confused upon a
304 hash collision between these pathnames and looped forever.
306 * Annotated tags outside refs/tags/ hierarchy were not advertised
307 correctly to the ls-remote and fetch with recent version of Git.
309 * Recent optimization broke shallow clones.
310 (merge f59de5d jk/peel-ref later to maint).
312 * "git cmd -- ':(top'" was not diagnosed as an invalid syntax, and
313 instead the parser kept reading beyond the end of the string.
315 * "git tag -f <tag>" always said "Updated tag '<tag>'" even when
316 creating a new tag (i.e. not overwriting nor updating).
318 * "git p4" did not behave well when the path to the root of the P4
319 client was not its real path.
320 (merge bbd8486 pw/p4-symlinked-root later to maint).
322 * "git archive" reports a failure when asked to create an archive out
323 of an empty tree. It would be more intuitive to give an empty
324 archive back in such a case.
326 * When "format-patch" quoted a non-ascii strings on the header files,
327 it incorrectly applied rfc2047 and chopped a single character in
328 the middle of it.
330 * An aliased command spawned from a bare repository that does not say
331 it is bare with "core.bare = yes" is treated as non-bare by mistake.
333 * In "git reflog expire", REACHABLE bit was not cleared from the
334 correct objects.
336 * The logic used by "git diff -M --stat" to shorten the names of
337 files before and after a rename did not work correctly when the
338 common prefix and suffix between the two filenames overlapped.
340 * The "--match=<pattern>" option of "git describe", when used with
341 "--all" to allow refs that are not annotated tags to be used as a
342 base of description, did not restrict the output from the command
343 to those that match the given pattern.
345 * Clarify in the documentation "what" gets pushed to "where" when the
346 command line to "git push" does not say these explicitly.
348 * The "--color=<when>" argument to the commands in the diff family
349 was described poorly.
351 * The arguments given to pre-rebase hook were not documented.
353 * The v4 index format was not documented.
355 * The "--match=<pattern>" argument "git describe" takes uses glob
356 pattern but it wasn't obvious from the documentation.
358 * Some sources failed to compile on systems that lack NI_MAXHOST in
359 their system header (e.g. z/OS).
361 * Add an example use of "--env-filter" in "filter-branch"
362 documentation.
364 * "git bundle verify" did not say "records a complete history" for a
365 bundle that does not have any prerequisites.
367 * In the v1.8.0 era, we changed symbols that do not have to be global
368 to file scope static, but a few functions in graph.c were used by
369 CGit from sideways bypassing the entry points of the API the
370 in-tree users use.
372 * "git update-index -h" did not do the usual "-h(elp)" thing.
374 * "git index-pack" had a buffer-overflow while preparing an
375 informational message when the translated version of it was too
376 long.
378 * 'git commit -m "$msg"' used to add an extra newline even when
379 $msg already ended with one.
381 * The SSL peer verification done by "git imap-send" did not ask for
382 Server Name Indication (RFC 4366), failing to connect SSL/TLS
383 sites that serve multiple hostnames on a single IP.
385 * perl/Git.pm::cat_blob slurped everything in core only to write it
386 out to a file descriptor, which was not a very smart thing to do.
388 * "git branch" did not bother to check nonsense command line
389 parameters and issue errors in many cases.
391 * Verification of signed tags were not done correctly when not in C
392 or en/US locale.
394 * Some platforms and users spell UTF-8 differently; retry with the
395 most official "UTF-8" when the system does not understand the
396 user-supplied encoding name that are the common alternative
397 spellings of UTF-8.
399 * When export-subst is used, "zip" output recorded incorrect
400 size of the file.
402 * "git am $maildir/" applied messages in an unexpected order; sort
403 filenames read from the maildir/ in a way that is more likely to
404 sort messages in the order the writing MUA meant to, by sorting
405 numeric segment in numeric order and non-numeric segment in
406 alphabetical order.
408 * "git submodule update", when recursed into sub-submodules, did not
409 accumulate the prefix paths.