6eb1401375 · matheusbsbs/matheus-bagliano

@@ -6,19 +6,19 @@ Backward compatibility warning(s)

6

7 This release has a few changes in the user-visible output from

8 Porcelain commands. These are not meant to be parsed by scripts, but

9 -the users still may want to be aware of the changes:

9 +users still may want to be aware of the changes:

10

11 - * Output from "git log --decorate" (and "%d" format specifier used in

12 - the userformat "--format=<string>" parameter "git log" family of

13 - command takes) used to list "HEAD" just like other tips of branch

14 - names, separated with a comma in between. E.g.

11 + * The output from "git log --decorate" (and, more generally, the "%d"

12 + format specifier used in the "--format=<string>" parameter to the

13 + "git log" family of commands) has changed. It used to list "HEAD"

14 + just like other branches; e.g.,

15

16 $ git log --decorate -1 master

17 commit bdb0f6788fa5e3cacc4315e9ff318a27b2676ff4 (HEAD, master)

18 ...

19

20 - This release updates the output slightly when HEAD refers to the tip

21 - of a branch whose name is also shown in the output. The above is

20 + This release changes the output slightly when HEAD refers to a

21 + branch whose name is also shown in the output. The above is now

22 shown as:

23

24 $ git log --decorate -1 master

@@ -26,15 +26,15 @@ the users still may want to be aware of the changes:

26 ...

27

28 * The phrasing "git branch" uses to describe a detached HEAD has been

29 - updated to match that of "git status":

29 + updated to agree with the phrasing used by "git status":

30

31 - - When the HEAD is at the same commit as it was originally

31 + - When HEAD is at the same commit as when it was originally

32 detached, they now both show "detached at <commit object name>".

33

34 - - When the HEAD has moved since it was originally detached,

35 - they now both show "detached from <commit object name>".

34 + - When HEAD has moved since it was originally detached, they now

35 + both show "detached from <commit object name>".

36

37 - Earlier "git branch" always used "from"

37 + Previously, "git branch" always used "from".

38

39

40 Updates since v2.3

@@ -46,8 +46,9 @@ Ports

46 platforms with smaller SSIZE_MAX, leading to read(2)/write(2)

47 failures.

48

49 - * We did not check the curl library version before using

50 - CURLOPT_PROXYAUTH feature that may not exist.

49 + * We did not check the curl library version before using the

50 + CURLOPT_PROXYAUTH feature, which did not exist in older versions of

51 + the library.

52

53 * We now detect number of CPUs on older BSD-derived systems.

54

@@ -66,99 +67,105 @@ UI, Workflows & Features

67 * "git log --invert-grep --grep=WIP" will show only commits that do

68 not have the string "WIP" in their messages.

69

69 - * "git push" has been taught a "--atomic" option that makes push to

70 - update more than one ref an "all-or-none" affair.

70 + * "git push" has been taught an "--atomic" option that makes a push

71 + that updates more than one ref an "all-or-none" affair.

72

72 - * Extending the "push to deploy" added in 2.3, the behaviour of "git

73 - push" when updating the branch that is checked out can now be

74 - tweaked by push-to-checkout hook.

73 + * Extending the "push to deploy" feature that was added in 2.3, the

74 + behaviour of "git push" when updating the branch that is checked

75 + out can now be tweaked by a "push-to-checkout" hook.

76

76 - * Using environment variable LANGUAGE and friends on the client side,

77 - HTTP-based transports now send Accept-Language when making requests.

77 + * HTTP-based transports now send Accept-Language when making

78 + requests. The languages to accept are inferred from environment

79 + variables on the client side (LANGUAGE, etc).

80

81 * "git send-email" used to accept a mistaken "y" (or "yes") as an

80 - answer to "What encoding do you want to use [UTF-8]? " without

81 - questioning. Now it asks for confirmation when the answer looks

82 - too short to be a valid encoding name.

82 + answer to "What encoding do you want to use [UTF-8]?" without

83 + questioning. Now it asks for confirmation when the answer looks too

84 + short to be a valid encoding name.

85

86 * When "git apply --whitespace=fix" fixed whitespace errors in the

87 common context lines, the command reports that it did so.

88

87 - * "git status" now allows the "-v" to be given twice to show the

88 - differences that are left in the working tree not to be committed.

89 + * "git status" now allows the "-v" option to be given twice, in which

90 + case it also shows the differences in the working tree that are not

91 + staged to be committed.

92

90 - * "git cherry-pick" used to clean-up the log message even when it is

91 - merely replaying an existing commit. It now replays the message

92 - verbatim unless you are editing the message of resulting commits.

93 + * "git cherry-pick" used to clean up the log message even when it is

94 + merely replaying an existing commit. It now replays the message

95 + verbatim unless you are editing the message of the resulting

96 + commit.

97

98 * "git archive" can now be told to set the 'text' attribute in the

99 resulting zip archive.

100

97 - * Output from "git log --decorate" mentions HEAD when it points at a

98 - tip of an branch differently from a detached HEAD.

101 + * Output from "git log --decorate" now distinguishes between a

102 + detached HEAD vs. a HEAD that points at a branch.

103

100 - This is a potentially backward-incompatible change.

104 + This is a potentially backward-incompatible change; see above for

105 + more information.

106

102 - * "git branch" on a detached HEAD always said "(detached from xyz)",

103 - even when "git status" would report "detached at xyz". The HEAD is

104 - actually at xyz and haven't been moved since it was detached in

105 - such a case, but the user cannot read what the current value of

106 - HEAD is when "detached from" is used.

107 + * When HEAD was detached when at commit xyz and hasn't been moved

108 + since it was detached, "git status" would report "detached at xyz"

109 + whereas "git branch" would report "detached from xyz". Now the

110 + output of "git branch" agrees with that of "git status".

111

108 - * "git -C '' subcmd" refused to work in the current directory, unlike

109 - "cd ''" which silently behaves as a no-op.

112 + This is a potentially backward-incompatible change; see above for

113 + more information.

114 +

115 + * "git -C '' subcmd" now works in the current directory (analogously

116 + to "cd ''") rather than dying with an error message.

117 (merge 6a536e2 kn/git-cd-to-empty later to maint).

118

119 * The versionsort.prereleaseSuffix configuration variable can be used

113 - to specify that v1.0-pre1 comes before v1.0.

120 + to specify that, for example, v1.0-pre1 comes before v1.0.

121

122 * A new "push.followTags" configuration turns the "--follow-tags"

123 option on by default for the "git push" command.

124

118 - * "git log --graph --no-walk A B..." is a conflicting request that

119 - asks nonsense; no-walk tells us show discrete points in the

120 - history, while graph asks to draw connections between these

121 - discrete points. Forbid the combination.

125 + * "git log --graph --no-walk A B..." is a nonsensical combination of

126 + options: "--no-walk" requests discrete points in the history, while

127 + "--graph" asks to draw connections between these discrete points.

128 + Forbid the use of these options together.

129

130 * "git rev-list --bisect --first-parent" does not work (yet) and can

124 - even cause SEGV; forbid it. "git log --bisect --first-parent"

125 - would not be useful until "git bisect --first-parent" materializes,

126 - so it is also forbidden for now.

131 + even cause SEGV; forbid it. "git log --bisect --first-parent" would

132 + not be useful until "git bisect --first-parent" materializes, so

133 + also forbid it for now.

134

135

136 Performance, Internal Implementation, Development Support etc.

137

131 - * Implementation of N_() macro has been updated slightly to help us

138 + * Slightly change the implementation of the N_() macro to help us

139 detect mistakes.

140

134 - * Implementation of "reflog expire" has been restructured to fit the

135 - reflogs better with the recently updated ref API.

141 + * Restructure the implementation of "reflog expire" to fit better

142 + with the recently updated reference API.

143

137 - * The transport-helper did not give transport options such as

144 + * The transport-helper did not pass transport options such as

145 verbosity, progress, cloning, etc. to import and export based

146 helpers, like it did for fetch and push based helpers, robbing them

140 - the chance to honor the wish of the end-users better.

141 -

142 - * The tests that wanted to see that file becomes unreadable after

143 - running "chmod a-r file", and the tests that wanted to make sure it

144 - is not run as root, we used "can we write into the / directory?" as

145 - a cheap substitute, but on some platforms that is not a good

146 - heuristics. The tests and their prerequisites have been updated to

147 - check what they really require.

147 + of the chance to honor the wish of the end-users better.

148 +

149 + * The tests that wanted to see that a file becomes unreadable after

150 + running "chmod a-r file", and the tests that wanted to make sure

151 + that they are not run as root, used "can we write into the /

152 + directory?" as a cheap substitute. But on some platforms that is

153 + not a good heuristic. The tests and their prerequisites have been

154 + updated to check what they really require.

155 (merge f400e51 jk/sanity later to maint).

156

157 * Various issues around "reflog expire", e.g. using --updateref when

158 expiring a reflog for a symbolic reference, have been corrected

159 and/or made saner.

160

154 - * The strbuf API was explained between the API documentation and in

155 - the header file. Move missing bits to strbuf.h so that programmers

156 - can check only one place for all necessary information.

161 + * The documentation for the strbuf API had been split between the API

162 + documentation and the header file. Consolidate the documentation in

163 + strbuf.h.

164

165 * The error handling functions and conventions are now documented in

159 - the API manual.

166 + the API manual (in api-error-handling.txt).

167

161 - * Optimize attribute look-up, mostly useful in "git grep" on a

168 + * Optimize gitattribute look-up, mostly useful in "git grep" on a

169 project that does not use many attributes, by avoiding it when we

170 (should) know that the attributes are not defined in the first

171 place.

@@ -170,26 +177,25 @@ Performance, Internal Implementation, Development Support etc.

177 (merge 0b868f0 sb/hex-object-name-is-at-most-41-bytes-long later to maint).

178 (merge 5d30851 dp/remove-duplicated-header-inclusion later to maint).

179

173 - * Simplify the ref transaction API around how "the ref should be

174 - pointing at this object" is specified.

180 + * Simplify the ref transaction API for verifying that "the ref should

181 + be pointing at this object".

182

176 - * Code in "git daemon" to parse out and hold hostnames used in

177 - request interpolation has been simplified.

183 + * Simplify the code in "git daemon" that parses out and holds

184 + hostnames used in request interpolation.

185

179 - * "git push" codepath has been restructured to make it easier to add

180 - new configuration bits.

186 + * Restructure the "git push" codepath to make it easier to add new

187 + configuration bits.

188

182 - * The run-command interface was easy to abuse and make a pipe for us

183 - to read from the process, wait for the process to finish and then

184 - attempt to read its output, which is a pattern that lead to a

185 - deadlock. Fix such uses by introducing a helper to do this

186 - correctly (i.e. we need to read first and then wait the process to

187 - finish) and also add code to prevent such abuse in the run-command

188 - helper.

189 + * The run-command interface made it easy to make a pipe for us to

190 + read from a process, wait for the process to finish, and then

191 + attempt to read its output. But this pattern can lead to deadlock.

192 + So introduce a helper to do this correctly (i.e., first read, and

193 + then wait the process to finish) and also add code to prevent such

194 + abuse in the run-command helper.

195

196 * People often forget to chain the commands in their test together

191 - with &&, leaving a failure from an earlier command in the test go

192 - unnoticed. The new GIT_TEST_CHAIN_LINT mechanism allows you to

197 + with &&, letting a failure from an earlier command in the test go

198 + unnoticed. The new GIT_TEST_CHAIN_LINT mechanism allows you to

199 catch such a mistake more easily.

200

201

@@ -208,33 +214,34 @@ notes for details).

214 (merge a46442f jk/blame-commit-label later to maint).

215

216 * "git rerere" (invoked internally from many mergy operations) did

211 - not correctly signal errors when told to update the working tree

212 - files and failed to do so for whatever reason.

217 + not correctly signal errors when it attempted to update the working

218 + tree files but failed for whatever reason.

219 (merge 89ea903 jn/rerere-fail-on-auto-update-failure later to maint).

220

221 * Setting diff.submodule to 'log' made "git format-patch" produce

222 broken patches.

223 (merge 339de50 dk/format-patch-ignore-diff-submodule later to maint).

224

219 - * After attempting and failing a password-less authentication

220 - (e.g. Kerberos), libcURL refuses to fall back to password based

221 - Basic authentication without a bit of help/encouragement.

225 + * After attempting and failing a password-less authentication (e.g.,

226 + Kerberos), libcURL refuses to fall back to password-based Basic

227 + authentication without a bit of help/encouragement.

228 (merge 4dbe664 bc/http-fallback-to-password-after-krb-fails later to maint).

229

224 - * The "git push" documentation made the "--repo=<there>" option

230 + * The "git push" documentation for the "--repo=<there>" option was

231 easily misunderstood.

232 (merge 57b92a7 mg/push-repo-option-doc later to maint).

233

228 - * Code to read branch name from various files in .git/ directory

229 - would have misbehaved if the code to write them left an empty file.

234 + * Code to read a branch name from various files in the .git/

235 + directory would have overrun array limits if asked to read an empty

236 + file.

237 (merge 66ec904 jk/status-read-branch-name-fix later to maint).

238

232 - * A misspelled conditional that is always true has been fixed.

239 + * Remove a superfluous conditional that is always true.

240 (merge 94ee8e2 jk/remote-curl-an-array-in-struct-cannot-be-null later to maint).

241

235 - * The documentation incorrectly said that C(opy) and R(ename) are the

236 - only ones that can be followed by the score number in the output in

237 - the --raw format.

242 + * The "git diff --raw" documentation incorrectly implied that C(opy)

243 + and R(ename) are the only statuses that can be followed by a score

244 + number.

245 (merge ac1c2d9 jc/diff-format-doc later to maint).

246

247 * A broken pack .idx file in the receiving repository prevented the

@@ -251,44 +258,44 @@ notes for details).

258 to the "log" command.

259 (merge 3cab02d jc/doc-log-rev-list-options later to maint).

260

254 - * "git apply --whitespace=fix" used to under-allocate the memory when

255 - the fix resulted in a longer text than the original patch.

261 + * "git apply --whitespace=fix" used to under-allocate memory when the

262 + fix resulted in a longer text than the original patch.

263 (merge 407a792 jc/apply-ws-fix-expands later to maint).

264

265 * The interactive "show a list and let the user choose from it"

259 - interface "add -i" used showed and prompted to the user even when

260 - the candidate list was empty, against which the only "choice" the

261 - user could have made was to choose nothing.

266 + interface used by "git add -i" unnecessarily prompted the user even

267 + when the candidate list was empty, against which the only "choice"

268 + the user could have made was to choose nothing.

269 (merge a9c4641 ak/add-i-empty-candidates later to maint).

270

264 - * The todo list "git rebase -i" creates did not fully honor

271 + * The todo list created by "git rebase -i" did not fully honor

272 core.abbrev settings.

273 (merge edb72d5 ks/rebase-i-abbrev later to maint).

274

268 - * "git fetch" over a remote-helper that cannot respond to "list"

269 - command could not fetch from a symbolic reference e.g. HEAD.

275 + * "git fetch" over a remote-helper that cannot respond to the "list"

276 + command could not fetch from a symbolic reference (e.g., HEAD).

277 (merge 33cae54 mh/deref-symref-over-helper-transport later to maint).

278

279 * "git push --signed" gave an incorrectly worded error message when

280 the other side did not support the capability.

281

282 * The "git push --signed" protocol extension did not limit what the

276 - "nonce" that is a server-chosen string can contain or how long it

277 - can be, which was unnecessarily lax. Limit both the length and the

278 - alphabet to a reasonably small space that can still have enough

283 + "nonce" (a server-chosen string) could contain nor how long it

284 + could be, which was unnecessarily lax. Limit both the length and

285 + the alphabet to a reasonably small space that can still have enough

286 entropy.

287 (merge afcb6ee jc/push-cert later to maint).

288

282 - * The completion script (in contrib/) contaminated global namespace

283 - and clobbered on a shell variable $x.

289 + * The completion script (in contrib/) clobbered the shell variable $x

290 + in the global shell namespace.

291 (merge 852ff1c ma/bash-completion-leaking-x later to maint).

292

286 - * We didn't format an integer that wouldn't fit in "int" but in

287 - "uintmax_t" correctly.

293 + * We incorrectly formatted a "uintmax_t" integer that doesn't fit in

294 + "int".

295 (merge d306f3d jk/decimal-width-for-uintmax later to maint).

296

290 - * Reading configuration from a blob object, when it ends with a lone

291 - CR, use to confuse the configuration parser.

297 + * The configuration parser used to be confused when reading

298 + configuration from a blob object that ends with a lone CR.

299 (merge 1d0655c jk/config-no-ungetc-eof later to maint).

300

301 * The pack bitmap support did not build with older versions of GCC.

@@ -304,21 +311,21 @@ notes for details).

311 (merge 1f985d6 ch/new-gpg-drops-rfc-1991 later to maint).

312

313 * The credential helper for Windows (in contrib/) used to mishandle

307 - a user name with an at-sign in it.

314 + user names that contain an at-sign.

315 (merge 13d261e av/wincred-with-at-in-username-fix later to maint).

316

317 * "diff-highlight" (in contrib/) used to show byte-by-byte

311 - differences, which meant that multi-byte characters can be chopped

312 - in the middle. It learned to pay attention to character boundaries

313 - (assuming the UTF-8 payload).

318 + differences, which could cause multi-byte characters to be chopped

319 + in the middle. It learned to pay attention to character boundaries

320 + (assuming UTF-8).

321 (merge 8d00662 jk/colors later to maint).

322

316 - * Longstanding configuration variable naming rules has been added to

317 - the documentation.

323 + * Document longstanding configuration variable naming rules in

324 + CodingGuidelines.

325 (merge 35840a3 jc/conf-var-doc later to maint).

326

327 * An earlier workaround to squelch unhelpful deprecation warnings

321 - from the compiler on Mac OSX unnecessarily set minimum required

328 + from the compiler on OS X unnecessarily set a minimum required

329 version of the OS, which the user might want to raise (or lower)

330 for other reasons.

331 (merge 88c03eb es/squelch-openssl-warnings-on-macosx later to maint).

@@ -335,33 +342,36 @@ notes for details).

342 "path/to/submodule".

343 (merge 8196e72 ps/submodule-sanitize-path-upon-add later to maint).

344

338 - * "git merge-file" did not work correctly in a subdirectory.

345 + * "git merge-file" did not work correctly when invoked in a

346 + subdirectory.

347 (merge 204a8ff ab/merge-file-prefix later to maint).

348

341 - * "git blame" died, trying to free an uninitialized piece of memory.

349 + * "git blame" could die trying to free an uninitialized piece of

350 + memory.

351 (merge e600592 es/blame-commit-info-fix later to maint).

352

353 * "git fast-import" used to crash when it could not close and

345 - conclude the resulting packfile cleanly.

354 + finalize the resulting packfile cleanly.

355 (merge 5e915f3 jk/fast-import-die-nicely-fix later to maint).

356

348 - * "update-index --refresh" used to leak when an entry cannot be

349 - refreshed for whatever reason.

357 + * "update-index --refresh" used to leak memory when an entry could

358 + not be refreshed for whatever reason.

359 (merge bc1c2ca sb/plug-leak-in-make-cache-entry later to maint).

360

361 * The "interpolated-path" option of "git daemon" inserted any string

353 - client declared on the "host=" capability request without checking.

354 - Sanitize and limit %H and %CH to a saner and a valid DNS name.

362 + the client declared on the "host=" capability request without

363 + checking. Sanitize and limit %H and %CH to a saner and a valid DNS

364 + name.

365 (merge b485373 jk/daemon-interpolate later to maint).

366

357 - * "git daemon" looked up the hostname even when "%CH" and "%IP"

358 - interpolations are not requested, which was unnecessary.

367 + * "git daemon" unnecessarily looked up the hostname even when "%CH"

368 + and "%IP" interpolations were not requested.

369 (merge dc8edc8 rs/daemon-interpolate later to maint).

370

361 - * Even though we officially haven't dropped Perl 5.8 support, the

362 - Getopt::Long package that came with it does not support "--no-"

363 - prefix to negate a boolean option; manually add support to help

364 - people with older Getopt::Long package.

371 + * We relied on "--no-" prefix handling in Perl's Getopt::Long

372 + package, even though that support didn't exist in Perl 5.8 (which

373 + we still support). Manually add support to help people with older

374 + Getopt::Long packages.

375 (merge f471494 km/send-email-getopt-long-workarounds later to maint).

376

377 * "git apply" was not very careful about reading from, removing,

@@ -370,13 +380,13 @@ notes for details).

380 replacement for GNU patch).

381 (merge e0d201b jc/apply-beyond-symlink later to maint).

382

373 - * A breakage to git-svn around v2.2 era that triggers premature

374 - closing of FileHandle has been corrected.

383 + * Correct a breakage in git-svn, introduced around the v2.2 era, that

384 + can cause FileHandles to be closed prematurely.

385 (merge e426311 ew/svn-maint-fixes later to maint).

386

377 - * We did not parse username followed by literal IPv6 address in SSH

378 - transport URLs, e.g. ssh://user@[2001:db8::1]:22/repo.git

379 - correctly.

387 + * We did not parse usernames followed by literal IPv6 addresses

388 + correctly in SSH transport URLs; e.g.,

389 + ssh://user@[2001:db8::1]:22/repo.git.

390 (merge 6b6c5f7 tb/connect-ipv6-parse-fix later to maint).

391

392 * The configuration variable 'mailinfo.scissors' was hard to

@@ -387,28 +397,26 @@ notes for details).

397 submodule.*.update configuration was not clearly documented.

398 (merge 5c31acf ms/submodule-update-config-doc later to maint).

399

390 - * "git diff --shortstat --dirstat=changes" showed a dirstat based on

391 - lines that was never asked by the end user in addition to the

392 - dirstat that the user asked for.

400 + * "git diff --shortstat" used together with "--dirstat=changes" or

401 + "--dirstat=files" incorrectly output dirstat information twice.

402 (merge ab27389 mk/diff-shortstat-dirstat-fix later to maint).

403

395 - * "git remote add" mentioned "--tags" and "--no-tags" and was not

396 - clear that fetch from the remote in the future will use the default

397 - behaviour when neither is given to override it.

404 + * The manpage for "git remote add" mentioned "--tags" and "--no-tags"

405 + but did not explain what happens if neither option is provided.

406 (merge aaba0ab mg/doc-remote-tags-or-not later to maint).

407

400 - * Description given by "grep -h" for its --exclude-standard option

401 - was phrased poorly.

408 + * The description of "--exclude-standard option" in the output of

409 + "git grep -h" was phrased poorly.

410 (merge 77fdb8a nd/grep-exclude-standard-help-fix later to maint).

411

404 - * "git rebase -i" recently started to include the number of

405 - commits in the todo list to be processed, but on a platform

406 - that prepends leading whitespaces to "wc -l" output, the numbers

407 - are shown with extra whitespaces that aren't necessary.

412 + * "git rebase -i" recently started to include the number of commits

413 + in the todo list, but that output included extraneous whitespace on

414 + a platform that prepends leading whitespaces to its "wc -l" output.

415 (merge 2185d3b es/rebase-i-count-todo later to maint).

416

410 - * The borrowed code in kwset API did not follow our usual convention

411 - to use "unsigned char" to store values that range from 0-255.

417 + * The borrowed code in the kwset API did not follow our usual

418 + convention to use "unsigned char" to store values that range from

419 + 0-255.

420 (merge 189c860 bw/kwset-use-unsigned later to maint).

421

422 * A corrupt input to "git diff -M" used to cause it to segfault.

@@ -418,9 +426,8 @@ notes for details).

426 (merge 3f88c1b mg/verify-commit later to maint).

427

428 * "git imap-send" learned to optionally talk with an IMAP server via

421 - libcURL; because there is no other option when Git is built with

422 - NO_OPENSSL option, use that codepath by default under such

423 - configuration.

429 + libcURL. Because there is no other option when Git is built with

430 + the NO_OPENSSL option, use libcURL by default in that case.

431 (merge dcd01ea km/imap-send-libcurl-options later to maint).

432

433 * "git log --decorate" did not reset colors correctly around the

@@ -436,44 +443,45 @@ notes for details).

443 transport.

444 (merge 8ddf3ca jk/smart-http-hide-refs later to maint).

445

439 - * "git tag -h" used to show the "--column" and "--sort" options

440 - that are about listing in a wrong section.

446 + * In the "git tag -h" output, move the documentation for the

447 + "--column" and "--sort" options to the "Tag listing options"

448 + section.

449 (merge dd059c6 jk/tag-h-column-is-a-listing-option later to maint).

450

451 * "git prune" used to largely ignore broken refs when deciding which

444 - objects are still being used, which could spread an existing small

445 - damage and make it a larger one.

452 + objects are still being used, which could cause reference

453 + corruption to lead to object loss.

454 (merge ea56c4e jk/prune-with-corrupt-refs later to maint).

455

448 - * The split-index mode introduced at v2.3.0-rc0~41 was broken in the

456 + * The split-index mode introduced in v2.3.0-rc0~41 was broken in the

457 codepath to protect us against a broken reimplementation of Git

458 that writes an invalid index with duplicated index entries, etc.

459 (merge 03f15a7 tg/fix-check-order-with-split-index later to maint).

460

453 - * "git fetch" that fetches a commit using the allow-tip-sha1-in-want

454 - extension could have failed to fetch all the requested refs.

461 + * "git fetch", when fetching a commit using the

462 + allow-tip-sha1-in-want extension, could have failed to fetch all of

463 + the requested refs.

464 (merge 32d0462 jk/fetch-pack later to maint).

465

466 * An failure early in the "git clone" that started creating the

458 - working tree and repository could have resulted in some directories

459 - and files left without getting cleaned up.

467 + working tree and repository could have resulted in the failure to

468 + clean up some directories and files.

469 (merge 16eff6c jk/cleanup-failed-clone later to maint).

470

471 * Recommend format-patch and send-email for those who want to submit

472 patches to this project.

473 (merge b25c469 jc/submitting-patches-mention-send-email later to maint).

474

466 - * Even though "git grep --quiet" is run merely to ask for the exit

467 - status, we spawned the pager regardless. Stop doing that.

475 + * Do not spawn the pager when "git grep" is run with "--quiet".

476 (merge c2048f0 ws/grep-quiet-no-pager later to maint).

477

478 * The prompt script (in contrib/) did not show the untracked sign

479 when working in a subdirectory without any untracked files.

480 (merge 9bdc517 ct/prompt-untracked-fix later to maint).

481

474 - * An earlier update to the parser that disects a URL broke an

475 - address, followed by a colon, followed by an empty string (instead

476 - of the port number), e.g. ssh://example.com:/path/to/repo.

482 + * An earlier update to the URL parser broke an address that contains

483 + a colon but an empty string for the port number, like

484 + ssh://example.com:/path/to/repo.

485 (merge 6b6c5f7 tb/connect-ipv6-parse-fix later to maint).

486

487 * Code cleanups and documentation updates.

RelNotes: wordsmithing