2d30c169da · matheusbsbs/matheus-bagliano

@@ -1,10 +1,10 @@

1 To: git@vger.kernel.org

2 Bcc: lwn@lwn.net

3 -Subject: What's cooking in git.git (Jun 2016, #02; Mon, 6)

3 +Subject: What's cooking in git.git (Jun 2016, #03; Thu, 9)

4 X-master-at: 49fa3dc76179e04b0833542fa52d0f287a4955ac

5 X-next-at: 3cd93d08f085fd56ac81614bd2e920c3b50eacec

6

7 -What's cooking in git.git (Jun 2016, #02; Mon, 6)

7 +What's cooking in git.git (Jun 2016, #03; Thu, 9)

8 --------------------------------------------------

9

10 Here are the topics that have been cooking. Commits prefixed with

@@ -12,271 +12,45 @@ Here are the topics that have been cooking. Commits prefixed with

12 '+' are in 'next'. The ones marked with '.' do not appear in any of

13 the integration branches, but I am still holding onto them.

14

15 -The hopefully final release candidate 2.9-rc2 together with 2.8.4

16 -has been pushed out.

17 -

18 -By the way, the tip of 'pu' is not expected to pass its own test

19 -suite, due to multiple broken topics queued near its tip. Earlier

20 -there was a talk about automatically bisecting breakage on 'pu', but

21 -tonight's pushout is a good illustration why that would not be very

22 -useful. What would be useful is (1) enumeration of commits on the

23 -first parent chain on master..pu, (2) test of the second parent of

24 -each of the commits found in (1), and (3) bisection between the

25 -broken commits in (2) and 'master', which would highlight broken

26 -topics and the commit(s) that might be responsible for each broken

27 -topic.

15 +The tip of 'pu' is not expected to pass its own test suite, due to

16 +multiple broken topics queued near its tip. Earlier there was a talk

17 +about automatically bisecting breakage on 'pu', but tonight's

18 +pushout is a good illustration why that would not be very useful.

19 +What would be useful is (1) enumeration of commits on the first

20 +parent chain on master..pu, (2) test of the second parent of each of

21 +the commits found in (1), and (3) bisection between the broken

22 +commits in (2) and 'master', which would highlight broken topics and

23 +the commit(s) that might be responsible for each broken topic.

24

25 You can find the changes described here in the integration branches

26 of the repositories listed at

27

28 http://git-blame.blogspot.com/p/git-public-repositories.html

29

34 ---------------------------------------------------

35 -[Graduated to "master"]

36 -

37 -* bd/readme.markdown-more (2016-05-31) 1 commit

38 - (merged to 'next' on 2016-06-02 at 1de7554)

39 - + README.md: format CLI commands with code syntax

40 -

41 - The mark-up in the top-level README.md file has been updated to

42 - typeset CLI command names differently from the body text.

43 -

44 -

45 -* em/man-bold-literal (2016-05-31) 1 commit

46 - (merged to 'next' on 2016-06-02 at efb4156)

47 - + Documentation: bold literals in man

48 -

49 - The manpage output of our documentation did not render well in

50 - terminal; typeset literals in bold by default to make them stand

51 - out more.

52 -

53 -

54 -* js/perf-rebase-i (2016-05-31) 1 commit

55 - (merged to 'next' on 2016-05-31 at 39d8dc8)

56 - + perf: make the tests work without a worktree

57 -

58 - The one in 'master' has a brown-paper-bag bug that breaks the perf

59 - test when used inside a usual Git repository with a working tree.

60 -

61 -

62 -* js/rebase-i-dedup-call-to-rerere (2016-05-31) 1 commit

63 - (merged to 'next' on 2016-05-31 at db1f9e3)

64 - + rebase -i: remove an unnecessary 'rerere' invocation

65 -

66 - "git rebase -i", after it fails to auto-resolve the conflict, had

67 - an unnecessary call to "git rerere" from its very early days, which

68 - was spotted recently; the call has been removed.

69 -

70 -

71 -* mm/makefile-developer-can-be-in-config-mak (2016-06-01) 2 commits

72 - (merged to 'next' on 2016-06-02 at f3a0678)

73 - + Makefile: add $(DEVELOPER_CFLAGS) variable

74 - + Makefile: move 'ifdef DEVELOPER' after config.mak* inclusion

75 -

76 - "make DEVELOPER=1" worked as expected; setting DEVELOPER=1 in

77 - config.mak didn't.

78 -

79 -

80 -* mr/send-email-doc-gmail-2fa (2016-06-01) 1 commit

81 - (merged to 'next' on 2016-06-02 at cd2ade8)

82 - + Documentation/git-send-email: fix typo in gmail 2FA section

83 -

84 - Typofix.

85 -

86 -

87 -* pa/cherry-pick-doc-typo (2016-05-29) 1 commit

88 - (merged to 'next' on 2016-06-02 at e2c2457)

89 - + git-cherry-pick.txt: correct a small typo

90 -

91 - "git cherry-pick --help" had three instances of word "behavior",

92 - one of which was spelled "behaviour", which is updated to match the

93 - other two.

94 -

95 -

96 -* rs/apply-name-terminate (2016-05-29) 1 commit

97 - (merged to 'next' on 2016-06-02 at 8df9332)

98 - + apply: remove unused parameters from name_terminate()

99 -

100 - Code clean-up.

101 -

102 -

103 -* rs/patch-id-use-skip-prefix (2016-05-29) 1 commit

104 - (merged to 'next' on 2016-06-02 at c993de3)

105 - + patch-id: use starts_with() and skip_prefix()

106 -

107 - Code clean-up.

108 -

109 -

110 -* sb/submodule-helper-list-signal-unmatch-via-exit-status (2016-06-01) 1 commit

111 - (merged to 'next' on 2016-06-02 at 29064a2)

112 - + submodule--helper: offer a consistent API

113 - (this branch is used by sb/submodule-helper-relative-path.)

114 -

115 - The way how "submodule--helper list" signals unmatch error to its

116 - callers has been updated.

117 -

118 -

119 -* sb/submodule-helper-relative-path (2016-06-01) 1 commit

120 - (merged to 'next' on 2016-06-02 at 8a191e1)

121 - + submodule: remove bashism from shell script

122 - (this branch uses sb/submodule-helper-list-signal-unmatch-via-exit-status.)

123 -

124 - A bash-ism "local" has been removed from "git submodule" scripted

125 - Porcelain.

126 -

30 --------------------------------------------------

31 [New Topics]

32

130 -* wd/userdiff-css (2016-06-03) 1 commit

131 - (merged to 'next' on 2016-06-06 at 536102f)

132 - + userdiff: add built-in pattern for CSS

133 -

134 - Update the funcname definition to support css files.

135 -

136 - Will merge to 'master' after 2.9 final.

137 -

138 -

139 -* jc/attr-more (2016-06-03) 2 commits

140 - - attr.c: pass struct git_attr_check down the callchain

141 - - attr.c: add push_stack() helper

142 - (this branch uses jc/attr; is tangled with sb/pathspec-label and sb/submodule-default-paths.)

143 -

144 - The beginning of long and tortuous journey to clean-up attribute

145 - subsystem implementation.

146 -

147 -

148 -* jk/rev-list-count-with-bitmap (2016-06-03) 2 commits

149 - (merged to 'next' on 2016-06-06 at dd9b30f)

150 - + rev-list: disable bitmaps when "-n" is used with listing objects

151 - + rev-list: "adjust" results of "--count --use-bitmap-index -n"

152 -

153 - "git rev-list --count" whose walk-length is limited with "-n"

154 - option did not work well with the counting optimized to look at the

155 - bitmap index.

156 -

157 - Will merge to 'master' after 2.9 final.

158 -

159 -

160 -* mh/ref-iterators (2016-06-03) 13 commits

161 - (merged to 'next' on 2016-06-06 at c8e79dc)

162 - + for_each_reflog(): reimplement using iterators

163 - + dir_iterator: new API for iterating over a directory tree

164 - + for_each_reflog(): don't abort for bad references

165 - + do_for_each_ref(): reimplement using reference iteration

166 - + refs: introduce an iterator interface

167 - + ref_resolves_to_object(): new function

168 - + entry_resolves_to_object(): rename function from ref_resolves_to_object()

169 - + get_ref_cache(): only create an instance if there is a submodule

170 - + remote rm: handle symbolic refs correctly

171 - + delete_refs(): add a flags argument

172 - + refs: use name "prefix" consistently

173 - + do_for_each_ref(): move docstring to the header file

174 - + refs: remove unnecessary "extern" keywords

175 - (this branch uses mh/split-under-lock.)

176 -

177 - The API to iterate over all the refs (i.e. for_each_ref(), etc.)

178 - has been revamped.

33 +* jk/avoid-unbounded-alloca (2016-06-07) 1 commit

34 + - tree-diff: avoid alloca for large allocations

35

180 - Will merge to 'master' after 2.9 final.

181 -

182 -

183 -* ew/mboxrd-format-am (2016-06-06) 3 commits

184 - - am: support --patch-format=mboxrd

185 - - mailsplit: support unescaping mboxrd messages

186 - - pretty: support "mboxrd" output format

187 -

188 - Teach format-patch and mailsplit (hence "am") how a line that

189 - happens to begin with "From " in the e-mail message is quoted with

190 - ">", so that these lines can be restored to their original shape.

191 -

192 -

193 -* lf/receive-pack-auto-gc-to-client (2016-06-06) 1 commit

194 - - receive-pack: send auto-gc output over sideband 2

195 -

196 - Allow messages that are generated by auto gc during "git push" on

197 - the receiving end to be explicitly passed back to the sending end

198 - over sideband, so that they are shown with "remote: " prefix to

199 - avoid confusing the users.

200 -

201 -

202 -* mg/cherry-pick-multi-on-unborn (2016-06-06) 1 commit

203 - - cherry-pick: allow to pick to unborn branches

204 -

205 - "git cherry-pick A" worked on an unborn branch, but "git

206 - cherry-pick A..B" didn't.

207 -

208 -

209 -* mj/log-show-signature-conf (2016-06-06) 2 commits

210 - - log: "--no-show-signature" commmand-line option

211 - - log: add "log.showsignature" configuration variable

212 -

213 - "git log" learns log.showSignature configuration variable, and a

214 - command line option "--no-show-signature" to countermand it.

215 -

216 - The order of the commits in the topic need to be reversed.

217 -

218 -

219 -* nd/i-t-a-commitable (2016-06-06) 3 commits

220 - - commit: don't count i-t-a entries when checking if the new commit is empty

221 - - Resurrect "diff-lib.c: adjust position of i-t-a entries in diff"

222 - - diff.h: extend "flags" field to 64 bits because we're out of bits

223 -

224 -

225 -* sg/reflog-past-root (2016-06-06) 1 commit

226 - - reflog: continue walking the reflog past root commits

227 -

228 - "git reflog" stopped upon seeing an entry that denotes a branch

229 - creation event (aka "unborn"), which made it appear as if the

230 - reflog was truncated.

231 -

232 -

233 -* tb/complete-status (2016-06-06) 3 commits

234 - - completion: add git status

235 - - completion: add __git_get_option_value helper

236 - - completion: factor out untracked file modes into a variable

237 -

238 - The completion script (in contrib/) learned to complete "git

239 - status" options.

240 -

241 -

242 -* tr/doc-tt (2016-06-06) 3 commits

243 - - doc: change configuration variables format

244 - - doc: change environment variables format

245 - - doc: clearer rule about formatting literals

36 + Will merge to 'next'.

37

247 - The documentation set has been updated so that literal commands,

248 - configuration variables and environment variables are consistently

249 - typeset in fixed-width font and bold in manpages.

38

39 +* jk/send-pack-stdio (2016-06-08) 1 commit

40 + - send-pack: use buffered I/O to talk to pack-objects

41

252 -* vs/prompt-avoid-unset-variable (2016-06-06) 1 commit

253 - - git-prompt.sh: Don't error on null ${ZSH,BASH}_VERSION, $short_sha

42 + Code clean-up.

43

255 - The git-prompt scriptlet (in contrib/) was not friendly with those

256 - who uses "set -u", which has been fixed.

44 + Will merge to 'next'.

45

46

259 -* rj/compat-regex-size-max-fix (2016-06-06) 1 commit

260 - - regex: fix a SIZE_MAX macro redefinition warning

47 +* pb/commit-editmsg-path (2016-06-09) 1 commit

48 + - builtin/commit.c: memoize git-path for COMMIT_EDITMSG

49

262 - A compilation fix.

50 + Code clean-up.

51

52 Will merge to 'next'.

53

266 -

267 -* bc/cocci-object-id (2016-06-06) 8 commits

268 - - merge-recursive: convert merge_recursive_generic to object_id

269 - - merge-recursive: convert leaf functions to use struct object_id

270 - - merge-recursive: convert struct merge_file_info to object_id

271 - - merge-recursive: convert struct stage_data to use object_id

272 - - Rename struct diff_filespec's sha1_valid member.

273 - - Convert struct diff_filespec to struct object_id

274 - - Apply standard object_id Coccinelle transformations.

275 - - Add basic Coccinelle transforms.

276 -

277 - Move from "unsigned char [20]" to "struct object_id" continues,

278 - with help from an automated tool.

279 -

54 --------------------------------------------------

55 [Stalled]

56

@@ -495,16 +269,184 @@ of the repositories listed at

269 --------------------------------------------------

270 [Cooking]

271

498 -* et/add-chmod-x (2016-06-01) 2 commits

499 - - SQUASH??? add: add --chmod=+x / --chmod=-x options

272 +* wd/userdiff-css (2016-06-03) 1 commit

273 + (merged to 'next' on 2016-06-06 at 536102f)

274 + + userdiff: add built-in pattern for CSS

275 +

276 + Update the funcname definition to support css files.

277 +

278 + Will merge to 'master' after 2.9 final.

279 +

280 +

281 +* jc/attr-more (2016-06-08) 5 commits

282 + - attr.c: always pass check[] to collect_some_attrs()

283 + - attr.c: correct ugly hack for git_all_attrs()

284 + - fixup! d5ad6c13

285 + - attr.c: pass struct git_attr_check down the callchain

286 + - attr.c: add push_stack() helper

287 + (this branch uses jc/attr; is tangled with sb/pathspec-label and sb/submodule-default-paths.)

288 +

289 + The beginning of long and tortuous journey to clean-up attribute

290 + subsystem implementation.

291 +

292 +

293 +* jk/rev-list-count-with-bitmap (2016-06-03) 2 commits

294 + (merged to 'next' on 2016-06-06 at dd9b30f)

295 + + rev-list: disable bitmaps when "-n" is used with listing objects

296 + + rev-list: "adjust" results of "--count --use-bitmap-index -n"

297 +

298 + "git rev-list --count" whose walk-length is limited with "-n"

299 + option did not work well with the counting optimized to look at the

300 + bitmap index.

301 +

302 + Will merge to 'master' after 2.9 final.

303 +

304 +

305 +* mh/ref-iterators (2016-06-03) 13 commits

306 + (merged to 'next' on 2016-06-06 at c8e79dc)

307 + + for_each_reflog(): reimplement using iterators

308 + + dir_iterator: new API for iterating over a directory tree

309 + + for_each_reflog(): don't abort for bad references

310 + + do_for_each_ref(): reimplement using reference iteration

311 + + refs: introduce an iterator interface

312 + + ref_resolves_to_object(): new function

313 + + entry_resolves_to_object(): rename function from ref_resolves_to_object()

314 + + get_ref_cache(): only create an instance if there is a submodule

315 + + remote rm: handle symbolic refs correctly

316 + + delete_refs(): add a flags argument

317 + + refs: use name "prefix" consistently

318 + + do_for_each_ref(): move docstring to the header file

319 + + refs: remove unnecessary "extern" keywords

320 + (this branch is used by mh/ref-store; uses mh/split-under-lock; is tangled with mh/update-ref-errors.)

321 +

322 + The API to iterate over all the refs (i.e. for_each_ref(), etc.)

323 + has been revamped.

324 +

325 + Will merge to 'master' after 2.9 final.

326 +

327 +

328 +* ew/mboxrd-format-am (2016-06-06) 3 commits

329 + - am: support --patch-format=mboxrd

330 + - mailsplit: support unescaping mboxrd messages

331 + - pretty: support "mboxrd" output format

332 +

333 + Teach format-patch and mailsplit (hence "am") how a line that

334 + happens to begin with "From " in the e-mail message is quoted with

335 + ">", so that these lines can be restored to their original shape.

336 +

337 +

338 +* lf/receive-pack-auto-gc-to-client (2016-06-06) 1 commit

339 + - receive-pack: send auto-gc output over sideband 2

340 +

341 + Allow messages that are generated by auto gc during "git push" on

342 + the receiving end to be explicitly passed back to the sending end

343 + over sideband, so that they are shown with "remote: " prefix to

344 + avoid confusing the users.

345 +

346 +

347 +* mg/cherry-pick-multi-on-unborn (2016-06-06) 1 commit

348 + - cherry-pick: allow to pick to unborn branches

349 +

350 + "git cherry-pick A" worked on an unborn branch, but "git

351 + cherry-pick A..B" didn't.

352 +

353 +

354 +* mj/log-show-signature-conf (2016-06-06) 2 commits

355 + - log: "--no-show-signature" commmand-line option

356 + - log: add "log.showsignature" configuration variable

357 +

358 + "git log" learns log.showSignature configuration variable, and a

359 + command line option "--no-show-signature" to countermand it.

360 +

361 + The order of the commits in the topic need to be reversed.

362 +

363 +

364 +* nd/i-t-a-commitable (2016-06-06) 3 commits

365 + - commit: don't count i-t-a entries when checking if the new commit is empty

366 + - Resurrect "diff-lib.c: adjust position of i-t-a entries in diff"

367 + - diff.h: extend "flags" field to 64 bits because we're out of bits

368 +

369 + "rm .git/index && git add -N * && git commit" allows you to create

370 + an empty commit without --allow-empty; attempt to forbid it.

371 +

372 + Breaks many tests by completely butchering "git commit", it seems.

373 +

374 +

375 +* sg/reflog-past-root (2016-06-06) 1 commit

376 + - reflog: continue walking the reflog past root commits

377 +

378 + "git reflog" stopped upon seeing an entry that denotes a branch

379 + creation event (aka "unborn"), which made it appear as if the

380 + reflog was truncated.

381 +

382 +

383 +* tb/complete-status (2016-06-06) 3 commits

384 + - completion: add git status

385 + - completion: add __git_get_option_value helper

386 + - completion: factor out untracked file modes into a variable

387 +

388 + The completion script (in contrib/) learned to complete "git

389 + status" options.

390 +

391 + Breaks t9902.

392 +

393 +

394 +* tr/doc-tt (2016-06-08) 4 commits

395 + - doc: change configuration variables format

396 + - doc: more consistency in environment variables format

397 + - doc: change environment variables format

398 + - doc: clearer rule about formatting literals

399 +

400 + The documentation set has been updated so that literal commands,

401 + configuration variables and environment variables are consistently

402 + typeset in fixed-width font and bold in manpages.

403 +

404 + Will merge to 'next'.

405 +

406 +

407 +* vs/prompt-avoid-unset-variable (2016-06-06) 1 commit

408 + - git-prompt.sh: Don't error on null ${ZSH,BASH}_VERSION, $short_sha

409 +

410 + The git-prompt scriptlet (in contrib/) was not friendly with those

411 + who uses "set -u", which has been fixed.

412 +

413 +

414 +* rj/compat-regex-size-max-fix (2016-06-06) 1 commit

415 + - regex: fix a SIZE_MAX macro redefinition warning

416 +

417 + A compilation fix.

418 +

419 + Will merge to 'next'.

420 +

421 +

422 +* bc/cocci-object-id (2016-06-06) 8 commits

423 + - merge-recursive: convert merge_recursive_generic to object_id

424 + - merge-recursive: convert leaf functions to use struct object_id

425 + - merge-recursive: convert struct merge_file_info to object_id

426 + - merge-recursive: convert struct stage_data to use object_id

427 + - Rename struct diff_filespec's sha1_valid member.

428 + - Convert struct diff_filespec to struct object_id

429 + - Apply standard object_id Coccinelle transformations.

430 + - Add basic Coccinelle transforms.

431 +

432 + Move from "unsigned char [20]" to "struct object_id" continues,

433 + with help from an automated tool.

434 +

435 +

436 +* et/add-chmod-x (2016-06-07) 1 commit

437 - add: add --chmod=+x / --chmod=-x options

438

502 - Waiting for a reroll (or ack for proposed fixup).

439 + "git update-index --add --chmod=+x file" may be usable as an escape

440 + hatch, but not a friendly thing to force for people who do need to

441 + use it regularly. "git add --chmod=+x file" can be used instead.

442 +

443 + Will merge to 'next'.

444

445

446 * jc/t2300-setup (2016-06-01) 1 commit

447 (merged to 'next' on 2016-06-06 at 20f7f83)

448 + t2300: run git-sh-setup in an environment that better mimics the real life

449 + (this branch is used by va/i18n-even-more.)

450

451 A test fix.

452

@@ -570,14 +512,30 @@ of the repositories listed at

512 Will merge to 'master' after 2.9 final.

513

514

573 -* tb/convert-peek-in-index (2016-05-24) 2 commits

574 - - convert: ce_compare_data() checks for a sha1 of a path

515 +* tb/convert-peek-in-index (2016-06-07) 3 commits

516 + - correct ce_compare_data() in a middle of a merge

517 - read-cache: factor out get_sha1_from_index() helper

518 + - convert: unify the "auto" handling of CRLF

519

577 - The motivation is rather iffy.

520 + Needs review.

521

522

580 -* va/i18n-even-more (2016-05-26) 22 commits

523 +* va/i18n-even-more (2016-06-07) 38 commits

524 + - i18n: branch: mark comment when editing branch description for translation

525 + - i18n: unmark die messages for translation

526 + - i18n: submodule: escape shell variables inside eval_gettext

527 + - i18n: submodule: join strings marked for translation

528 + - i18n: init-db: join message pieces

529 + - i18n: remote: allow translations to reorder message

530 + - i18n: remote: mark URL fallback text for translation

531 + - i18n: standardise messages

532 + - i18n: sequencer: add period to error message

533 + - i18n: merge: change command option help to lowercase

534 + - i18n: merge: mark messages for translation

535 + - i18n: notes: mark options for translation

536 + - i18n: notes: mark strings for translation

537 + - i18n: transport-helper.c: change N_() call to _()

538 + - i18n: bisect: mark strings for translation

539 - t5523: use test_i18ngrep for negation

540 - t4153: fix negated test_i18ngrep call

541 - t9003: become resilient to GETTEXT_POISON

@@ -600,11 +558,14 @@ of the repositories listed at

558 - i18n: advice: internationalize message for conflicts

559 - i18n: advice: mark string about detached head for translation

560 - i18n: builtin/remote.c: fix mark for translation

561 + - Merge branch 'jc/t2300-setup' into HEAD

562 + (this branch uses jc/t2300-setup.)

563

564 More markings of messages for i18n, with updates to various tests

565 to pass GETTEXT_POISON tests.

566

607 - Is everybody happy with this version?

567 + One patch from the original submission dropped due to conflicts

568 + with other topics in flight.

569

570

571 * jg/dash-is-last-branch-in-worktree-add (2016-05-31) 1 commit

@@ -920,7 +881,7 @@ of the repositories listed at

881 + read_raw_ref(): don't get confused by an empty directory

882 + commit_ref(): if there is an empty dir in the way, delete it

883 + t1404: demonstrate a bug resolving references

923 - (this branch is used by mh/ref-iterators.)

884 + (this branch is used by mh/ref-iterators, mh/ref-store and mh/update-ref-errors.)

885

886 Further preparatory work on the refs API before the pluggable

887 backend series can land.

What's cooking (2016/06 #03)