77ac1a0b4f · matheusbsbs/matheus-bagliano

@@ -1,18 +1,20 @@

1 To: git@vger.kernel.org

2 Bcc: lwn@lwn.net

3 -Subject: What's cooking in git.git (Mar 2015, #07; Fri, 20)

4 -X-master-at: e80e85a52adfda053cafc3b23058a86aff35404f

5 -X-next-at: d87118a61e74087359ac4d43ef91e5be61958443

3 +Subject: What's cooking in git.git (Mar 2015, #08; Mon, 23)

4 +X-master-at: f53fc38c083da1a1d14f54713ffd8fd0965a2194

5 +X-next-at: b1122c2910374ef5dde8cb25b25be0abb1538aba

6

7 -What's cooking in git.git (Mar 2015, #07; Fri, 20)

7 +What's cooking in git.git (Mar 2015, #08; Mon, 23)

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

9

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

11 '-' are only in 'pu' (proposed updates) while commits prefixed with

12 '+' are in 'next'.

13

14 -Now we are almost at the end of week #6, and hopefully we can do a

15 --rc0 late next week to start wrapping this cycle up.

14 +This cycle is turning out to be a "shoot for product excellence"

15 +release. About half of the commits that have been merged to the

16 +'master' so far have also been merged to 'maint' and v2.3.4 has been

17 +tagged.

18

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

20 of the repositories listed at

@@ -22,276 +24,133 @@ of the repositories listed at

24 --------------------------------------------------

25 [Graduated to "master"]

26

25 -* jc/decorate-leaky-separator-color (2015-03-04) 7 commits

26 - (merged to 'next' on 2015-03-13 at 38e7cd1)

27 - + log --decorate: do not leak "commit" color into the next item

28 - + Documentation/config.txt: simplify boolean description in the syntax section

29 - + Documentation/config.txt: describe 'color' value type in the "Values" section

30 - + Documentation/config.txt: have a separate "Values" section

31 - + Documentation/config.txt: describe the structure first and then meaning

32 - + Documentation/config.txt: explain multi-valued variables once

33 - + Documentation/config.txt: avoid unnecessary negation

34 - (this branch is used by mg/log-decorate-HEAD.)

35 -

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

37 - branch names.

38 -

39 -

40 -* km/bsd-shells (2015-03-10) 5 commits

41 - (merged to 'next' on 2015-03-13 at f839e4c)

42 - + t5528: do not fail with FreeBSD shell

43 - + help.c: use SHELL_PATH instead of hard-coded "/bin/sh"

44 - + git-compat-util.h: move SHELL_PATH default into header

45 - + git-instaweb: use @SHELL_PATH@ instead of /bin/sh

46 - + git-instaweb: allow running in a working tree subdirectory

47 -

48 - Portability fixes and workarounds for shell scripts have been added

49 - to help BSD-derived systems.

50 -

51 -

52 -* km/bsd-sysctl (2015-03-10) 2 commits

53 - (merged to 'next' on 2015-03-13 at c4557c9)

54 - + thread-utils.c: detect CPU count on older BSD-like systems

55 - + configure: support HAVE_BSD_SYSCTL option

56 -

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

58 -

59 -

60 -* km/imap-send-libcurl-options (2015-03-10) 1 commit

61 - (merged to 'next' on 2015-03-13 at 569d0c4)

62 - + imap-send: use cURL automatically when NO_OPENSSL defined

63 -

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

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

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

67 - configuration.

68 -

69 -

70 -* kn/git-cd-to-empty (2015-03-06) 1 commit

71 - (merged to 'next' on 2015-03-12 at 54dea03)

72 - + git: treat "git -C '<path>'" as a no-op when <path> is empty

73 -

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

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

76 -

77 -

78 -* mg/detached-head-report (2015-03-06) 2 commits

79 - (merged to 'next' on 2015-03-12 at 89443b5)

80 - + branch: name detached HEAD analogous to status

81 - + wt-status: refactor detached HEAD analysis

82 -

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

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

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

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

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

88 -

89 -

90 -* mg/log-decorate-HEAD (2015-03-10) 2 commits

91 - (merged to 'next' on 2015-03-13 at 96ce660)

92 - + log: decorate HEAD with branch name

93 - + Merge branch 'jc/decorate-leaky-separator-color' into HEAD

94 - (this branch uses jc/decorate-leaky-separator-color.)

95 -

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

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

27 +* jk/push-config (2015-03-14) 4 commits

28 + (merged to 'next' on 2015-03-16 at 6452570)

29 + + push: allow --follow-tags to be set by config push.followTags

30 + + cmd_push: pass "flags" pointer to config callback

31 + + cmd_push: set "atomic" bit directly

32 + + git_push_config: drop cargo-culted wt_status pointer

33

99 - This is a potentially backward-incompatible change.

34 + Restructure "git push" codepath to make it easier to add new

35 + configuration bits and then add push.followTags configuration that

36 + turns --follow-tags option on by default.

37

38

102 -* mg/verify-commit (2015-03-10) 1 commit

103 - (merged to 'next' on 2015-03-13 at 292197f)

104 - + t7510: do not fail when gpg warns about insecure memory

39 +* jk/smart-http-hide-refs (2015-03-12) 2 commits

40 + (merged to 'next' on 2015-03-16 at 530df4c)

41 + + upload-pack: do not check NULL return of lookup_unknown_object

42 + + upload-pack: fix transfer.hiderefs over smart-http

43

106 - Workarounds for certain build of GPG that triggered false breakage

107 - in a test.

44 + The transfer.hiderefs support did not quite work for smart-http

45 + transport.

46

47

110 -* nd/versioncmp-prereleases (2015-03-10) 2 commits

111 - (merged to 'next' on 2015-03-10 at 1df647c)

112 - + config.txt: update versioncmp.prereleaseSuffix

113 - (merged to 'next' on 2015-03-03 at 6ab29cf)

114 - + versionsort: support reorder prerelease suffixes

48 +* jk/tag-h-column-is-a-listing-option (2015-03-12) 1 commit

49 + (merged to 'next' on 2015-03-16 at 42b04c6)

50 + + tag: fix some mis-organized options in "-h" listing

51

116 - The versionsort.prerelease configuration variable can be used to

117 - specify that v1.0-pre1 comes before v1.0.

52 + "git tag -h" used to show the "--column" and "--sort" options

53 + that are about listing in a wrong section.

54

55

120 -* rs/daemon-hostname-in-strbuf (2015-03-09) 2 commits

121 - (merged to 'next' on 2015-03-12 at 7f61f07)

122 - + daemon: deglobalize hostname information

123 - + daemon: use strbuf for hostname info

56 +* jk/test-annoyances (2015-03-12) 5 commits

57 + (merged to 'next' on 2015-03-16 at 845b091)

58 + + t5551: make EXPENSIVE test cheaper

59 + + t5541: move run_with_cmdline_limit to test-lib.sh

60 + + t: pass GIT_TRACE through Apache

61 + + t: redirect stderr GIT_TRACE to descriptor 4

62 + + t: translate SIGINT to an exit

63

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

126 - request interpolation has been simplified.

64 + Test fixes.

65

66

129 -* rs/use-isxdigit (2015-03-10) 1 commit

130 - (merged to 'next' on 2015-03-13 at 39e99dd)

131 - + use isxdigit() for checking if a character is a hexadecimal digit

67 +* js/completion-ctags-pattern-substitution-fix (2015-03-14) 1 commit

68 + (merged to 'next' on 2015-03-17 at 4a68861)

69 + + contrib/completion: escape the forward slash in __git_match_ctag

70

133 - Code cleanup.

71 + The code that reads from the ctags file in the completion script

72 + (in contrib/) did not spell ${param/pattern/string} substitution

73 + correctly, which happened to work with bash but not with zsh.

74

75

136 -* sb/leaks (2015-03-10) 3 commits

137 - (merged to 'next' on 2015-03-13 at 50317d6)

138 - + builtin/help.c: fix memory leak

139 - + bundle.c: fix memory leak

140 - + connect.c: do not leak "conn" after showing diagnosis

76 +* nd/config-doc-camelCase (2015-03-13) 1 commit

77 + (merged to 'next' on 2015-03-16 at 0e3fedb)

78 + + *config.txt: stick to camelCase naming convention

79

142 - Code cleanup.

80 + Documentation updates.

81

82 --------------------------------------------------

83 [New Topics]

84

147 -* jk/cleanup-failed-clone (2015-03-19) 2 commits

148 - - clone: drop period from end of die_errno message

149 - - clone: initialize atexit cleanup handler earlier

85 +* jc/diff-no-index-d-f (2015-03-22) 1 commit

86 + - align D/F handling of "diff --no-index" with that of normal Git

87

151 - An failure early in the "git clone" that started creating the

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

153 - and files left without getting cleaned up.

88 + The usual "git diff" when seeing a file turning into a directory

89 + showed a patchset to remove the file and create all files in the

90 + directory, but "git diff --no-index" simply refused to work.

91

155 - Will merge to 'next'.

92

93 +* jk/run-command-capture (2015-03-22) 7 commits

94 + (merged to 'next' on 2015-03-23 at f6db88b)

95 + + run-command: forbid using run_command with piped output

96 + + trailer: use capture_command

97 + + submodule: use capture_command

98 + + wt-status: use capture_command

99 + + run-command: introduce capture_command helper

100 + + wt_status: fix signedness mismatch in strbuf_read call

101 + + wt-status: don't flush before running "submodule status"

102

158 -* jk/fetch-pack (2015-03-19) 4 commits

159 - - fetch-pack: remove dead assignment to ref->new_sha1

160 - - fetch_refs_via_pack: free extra copy of refs

161 - - filter_ref: make a copy of extra "sought" entries

162 - - filter_ref: avoid overwriting ref->old_sha1 with garbage

163 -

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

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

166 -

167 - Will merge to 'next'.

168 -

169 -

170 -* jk/prune-with-corrupt-refs (2015-03-20) 5 commits

171 - - refs.c: drop curate_packed_refs

172 - - repack: turn on "ref paranoia" when doing a destructive repack

173 - - prune: turn on ref_paranoia flag

174 - - refs: introduce a "ref paranoia" flag

175 - - t5312: test object deletion code paths in a corrupted repository

176 -

177 - "git prune" used to largely ignore broken refs when deciding which

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

179 - damage and make it a larger one.

180 -

181 - Will merge to 'next'.

182 -

183 -

184 -* jk/simplify-csum-file-sha1fd-check (2015-03-19) 1 commit

185 - (merged to 'next' on 2015-03-20 at 6f6d1c2)

186 - + sha1fd_check: die when we cannot open the file

187 -

188 - Code simplification.

103 + The run-command interface was easy to abuse and make a pipe for us

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

105 + attempt to read its output, which is a pattern that lead to a

106 + deadlock. Fix such uses by introducing a helper to do this

107 + correctly (i.e. we need to read first and then wait the process to

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

109 + helper.

110

111 Will merge to 'master'.

112

113

193 -* mh/numparse (2015-03-19) 14 commits

194 - - diff_opt_parse(): use convert_i() when handling --abbrev=<num>

195 - - diff_opt_parse(): use convert_i() when handling "-l<num>"

196 - - opt_arg(): simplify pointer handling

197 - - opt_arg(): report errors parsing option values

198 - - opt_arg(): use convert_i() in implementation

199 - - opt_arg(): val is always non-NULL

200 - - builtin_diff(): detect errors when parsing --unified argument

201 - - handle_revision_opt(): use convert_ui() when handling "--abbrev="

202 - - strtoul_ui(), strtol_i(): remove functions

203 - - handle_revision_opt(): use convert_i() when handling "-<digit>"

204 - - handle_revision_opt(): use skip_prefix() in many places

205 - - write_subdirectory(): use convert_ui() for parsing mode

206 - - cacheinfo_callback(): use convert_ui() when handling "--cacheinfo"

207 - - numparse: new module for parsing integral numbers

208 -

209 - Many codepaths use unchecked use of strtol() and friends (or even

210 - worse, atoi()). Introduce a set of wrappers that try to be more

211 - careful.

212 -

213 -

214 -* tf/gitweb-project-listing (2015-03-19) 5 commits

215 - - gitweb: make category headings into links when they are directories

216 - - gitweb: optionally set project category from its pathname

217 - - gitweb: add a link under the search box to clear a project filter

218 - - gitweb: if the PATH_INFO is incomplete, use it as a project_filter

219 - - gitweb: fix typo in man page

220 -

221 - Update gitweb to make it more pleasant to deal with a hierarchical

222 - forest of repositories.

223 -

224 - Any comments from those who use or have their own code in Gitweb?

114 +* tg/fix-check-order-with-split-index (2015-03-20) 1 commit

115 + (merged to 'next' on 2015-03-23 at c361f8e)

116 + + read-cache: fix reading of split index

117

226 -

227 -* ws/grep-quiet-no-pager (2015-03-19) 1 commit

228 - (merged to 'next' on 2015-03-20 at b3f5fe6)

229 - + grep: fix "--quiet" overwriting current output

230 -

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

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

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

119 + codepath to protect us against a broken reimplementation of Git

120 + that writes an invalid index with duplicated index entries, etc.

121

122 Will merge to 'master'.

123

124

237 -* kd/rev-list-bisect-first-parent (2015-03-19) 1 commit

238 - (merged to 'next' on 2015-03-20 at 5477bf5)

239 - + rev-list: refuse --first-parent combined with --bisect

125 +* ts/man-pdf (2015-03-20) 1 commit

126 + - Documentation: make 'make pdf' format manpages to PDF as well

127

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

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

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

244 - so it is also forbidden for now.

128 + For offline consumption of manual pages, a target to produce one

129 + pdf document per manual page was added to Documentation/Makefile.

130

246 - Will merge to 'master'.

131 + Its usefulness is unknown, given that this does not produce a

132 + single document with all the manual pages in it, which would be not

133 + much better than keeping a bunch of HTML manual pages we already

134 + produce and use them instead for offline consumption.

135

136

249 -* jk/test-chain-lint (2015-03-20) 27 commits

250 - - t9158, t9161: fix broken &&-chain in git-svn tests

251 - - t9104: fix test for following larger parents

252 - - t4104: drop hand-rolled error reporting

253 - - t0005: fix broken &&-chains

254 - - t7004: fix embedded single-quotes

255 - - t0050: appease --chain-lint

256 - - t9001: use test_when_finished

257 - - t4117: use modern test_* helpers

258 - - t6034: use modern test_* helpers

259 - - t1301: use modern test_* helpers

260 - - t0020: use modern test_* helpers

261 - - t6030: use modern test_* helpers

262 - - t9502: fix &&-chain breakage

263 - - t7201: fix &&-chain breakage

264 - - t3600: fix &&-chain breakage for setup commands

265 - - t: avoid using ":" for comments

266 - - t: wrap complicated expect_code users in a block

267 - - t: use test_expect_code instead of hand-rolled comparison

268 - - t: use test_might_fail for diff and grep

269 - - t: fix &&-chaining issues around setup which might fail

270 - - t: use test_must_fail instead of hand-rolled blocks

271 - - t: use verbose instead of hand-rolled errors

272 - - t: assume test_cmp produces verbose output

273 - - t: fix trivial &&-chain breakage

274 - - t: fix moderate &&-chain breakage

275 - - t: fix severe &&-chain breakage

276 - - t/test-lib: introduce --chain-lint option

277 -

278 - People often forget to chain the commands in their test together

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

280 - unnoticed. The new GIT_TEST_CHAIN_LINT mechanism allows you to

281 - catch such a mistake more easily.

282 -

283 - What I queued here has fix to the issue J6t found in 15/25 squashed

284 - in, and also has 26/25 and 27/25 follow-up fixes from Michael. If

285 - everybody involved is happy with it, then we can just proceed with

286 - this copy, otherwise I'll let Peff reroll. I am happy either way.

137 +* nd/diff-i-t-a (2015-03-23) 1 commit

138 + - diff-lib.c: adjust position of i-t-a entries in diff

139

140

289 -* tg/test-index-v4 (2015-03-20) 1 commit

290 - - t1700: make test pass with index-v4

141 +* sb/leaks (2015-03-23) 9 commits

142 + (merged to 'next' on 2015-03-23 at 5397daf)

143 + + read-cache: fix memleak

144 + + add_to_index(): free unused cache-entry

145 + + commit.c: fix a memory leak

146 + + http-push: remove unneeded cleanup

147 + + merge-recursive: fix memleaks

148 + + merge-blobs.c: fix a memleak

149 + + builtin/apply.c: fix a memleak

150 + + update-index: fix a memleak

151 + + read-cache: free cache entry in add_to_index in case of early return

152

292 - A test fix.

293 -

294 - Will merge to 'next'.

153 + Will merge to 'master'.

154

155 --------------------------------------------------

156 [Stalled]

@@ -490,13 +349,158 @@ of the repositories listed at

349 --------------------------------------------------

350 [Cooking]

351

493 -* js/completion-ctags-pattern-substitution-fix (2015-03-14) 1 commit

494 - (merged to 'next' on 2015-03-17 at 4a68861)

495 - + contrib/completion: escape the forward slash in __git_match_ctag

352 +* jk/cleanup-failed-clone (2015-03-19) 2 commits

353 + (merged to 'next' on 2015-03-23 at 1f26d93)

354 + + clone: drop period from end of die_errno message

355 + + clone: initialize atexit cleanup handler earlier

356 +

357 + An failure early in the "git clone" that started creating the

358 + working tree and repository could have resulted in some directories

359 + and files left without getting cleaned up.

360 +

361 + Will merge to 'master'.

362

497 - The code that reads from the ctags file in the completion script

498 - (in contrib/) did not spell ${param/pattern/string} substitution

499 - correctly, which happened to work with bash but not with zsh.

363 +

364 +* jk/fetch-pack (2015-03-19) 4 commits

365 + (merged to 'next' on 2015-03-23 at 4357f3d)

366 + + fetch-pack: remove dead assignment to ref->new_sha1

367 + + fetch_refs_via_pack: free extra copy of refs

368 + + filter_ref: make a copy of extra "sought" entries

369 + + filter_ref: avoid overwriting ref->old_sha1 with garbage

370 +

371 + "git fetch" that fetches a commit using the allow-tip-sha1-in-want

372 + extension could have failed to fetch all the requested refs.

373 +

374 + Will merge to 'master'.

375 +

376 +

377 +* jk/prune-with-corrupt-refs (2015-03-20) 5 commits

378 + (merged to 'next' on 2015-03-23 at 68af8a9)

379 + + refs.c: drop curate_packed_refs

380 + + repack: turn on "ref paranoia" when doing a destructive repack

381 + + prune: turn on ref_paranoia flag

382 + + refs: introduce a "ref paranoia" flag

383 + + t5312: test object deletion code paths in a corrupted repository

384 +

385 + "git prune" used to largely ignore broken refs when deciding which

386 + objects are still being used, which could spread an existing small

387 + damage and make it a larger one.

388 +

389 + Will merge to 'master'.

390 +

391 +

392 +* jk/simplify-csum-file-sha1fd-check (2015-03-19) 1 commit

393 + (merged to 'next' on 2015-03-20 at 6f6d1c2)

394 + + sha1fd_check: die when we cannot open the file

395 +

396 + Code simplification.

397 +

398 + Will merge to 'master'.

399 +

400 +

401 +* mh/numparse (2015-03-19) 14 commits

402 + - diff_opt_parse(): use convert_i() when handling --abbrev=<num>

403 + - diff_opt_parse(): use convert_i() when handling "-l<num>"

404 + - opt_arg(): simplify pointer handling

405 + - opt_arg(): report errors parsing option values

406 + - opt_arg(): use convert_i() in implementation

407 + - opt_arg(): val is always non-NULL

408 + - builtin_diff(): detect errors when parsing --unified argument

409 + - handle_revision_opt(): use convert_ui() when handling "--abbrev="

410 + - strtoul_ui(), strtol_i(): remove functions

411 + - handle_revision_opt(): use convert_i() when handling "-<digit>"

412 + - handle_revision_opt(): use skip_prefix() in many places

413 + - write_subdirectory(): use convert_ui() for parsing mode

414 + - cacheinfo_callback(): use convert_ui() when handling "--cacheinfo"

415 + - numparse: new module for parsing integral numbers

416 +

417 + Many codepaths use unchecked use of strtol() and friends (or even

418 + worse, atoi()). Introduce a set of wrappers that try to be more

419 + careful.

420 +

421 +

422 +* tf/gitweb-project-listing (2015-03-19) 5 commits

423 + - gitweb: make category headings into links when they are directories

424 + - gitweb: optionally set project category from its pathname

425 + - gitweb: add a link under the search box to clear a project filter

426 + - gitweb: if the PATH_INFO is incomplete, use it as a project_filter

427 + - gitweb: fix typo in man page

428 +

429 + Update gitweb to make it more pleasant to deal with a hierarchical

430 + forest of repositories.

431 +

432 + Any comments from those who use or have their own code in Gitweb?

433 +

434 +

435 +* ws/grep-quiet-no-pager (2015-03-19) 1 commit

436 + (merged to 'next' on 2015-03-20 at b3f5fe6)

437 + + grep: fix "--quiet" overwriting current output

438 +

439 + Even though "git grep --quiet" is run merely to ask for the exit

440 + status, we spawned the pager regardless. Stop doing that.

441 +

442 + Will merge to 'master'.

443 +

444 +

445 +* kd/rev-list-bisect-first-parent (2015-03-19) 1 commit

446 + (merged to 'next' on 2015-03-20 at 5477bf5)

447 + + rev-list: refuse --first-parent combined with --bisect

448 +

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

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

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

452 + so it is also forbidden for now.

453 +

454 + Will merge to 'master'.

455 +

456 +

457 +* jk/test-chain-lint (2015-03-22) 28 commits

458 + - t6039: fix broken && chain

459 + - t9158, t9161: fix broken &&-chain in git-svn tests

460 + - t9104: fix test for following larger parents

461 + - t4104: drop hand-rolled error reporting

462 + - t0005: fix broken &&-chains

463 + - t7004: fix embedded single-quotes

464 + - t0050: appease --chain-lint

465 + - t9001: use test_when_finished

466 + - t4117: use modern test_* helpers

467 + - t6034: use modern test_* helpers

468 + - t1301: use modern test_* helpers

469 + - t0020: use modern test_* helpers

470 + - t6030: use modern test_* helpers

471 + - t9502: fix &&-chain breakage

472 + - t7201: fix &&-chain breakage

473 + - t3600: fix &&-chain breakage for setup commands

474 + - t: avoid using ":" for comments

475 + - t: wrap complicated expect_code users in a block

476 + - t: use test_expect_code instead of hand-rolled comparison

477 + - t: use test_might_fail for diff and grep

478 + - t: fix &&-chaining issues around setup which might fail

479 + - t: use test_must_fail instead of hand-rolled blocks

480 + - t: use verbose instead of hand-rolled errors

481 + - t: assume test_cmp produces verbose output

482 + - t: fix trivial &&-chain breakage

483 + - t: fix moderate &&-chain breakage

484 + - t: fix severe &&-chain breakage

485 + - t/test-lib: introduce --chain-lint option

486 +

487 + People often forget to chain the commands in their test together

488 + with &&, leaving a failure from an earlier command in the test go

489 + unnoticed. The new GIT_TEST_CHAIN_LINT mechanism allows you to

490 + catch such a mistake more easily.

491 +

492 + What I queued here has fix to the issue J6t found in 15/25 squashed

493 + in, and also has 26/25 and 27/25 follow-up fixes from Michael, plus

494 + 28/25 follow-up from Torsten. If everybody involved is happy with

495 + it, we can just proceed with this copy, otherwise I'll let Peff

496 + reroll. I am happy either way.

497 +

498 +

499 +* tg/test-index-v4 (2015-03-20) 1 commit

500 + (merged to 'next' on 2015-03-23 at 6b2eb0a)

501 + + t1700: make test pass with index-v4

502 +

503 + A test fix.

504

505 Will merge to 'master'.

506

@@ -515,35 +519,12 @@ of the repositories listed at

519

520

521 * jc/submitting-patches-mention-send-email (2015-03-15) 1 commit

518 - - SubmittingPatches: encourage users to use format-patch and send-email

522 + (merged to 'next' on 2015-03-23 at a9581fd)

523 + + SubmittingPatches: encourage users to use format-patch and send-email

524

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

526 patches to this project.

527

523 - Will merge to 'next'.

524 -

525 -

526 -* jk/push-config (2015-03-14) 4 commits

527 - (merged to 'next' on 2015-03-16 at 6452570)

528 - + push: allow --follow-tags to be set by config push.followTags

529 - + cmd_push: pass "flags" pointer to config callback

530 - + cmd_push: set "atomic" bit directly

531 - + git_push_config: drop cargo-culted wt_status pointer

532 -

533 - Restructure "git push" codepath to make it easier to add new

534 - configuration bits and then add push.followTags configuration that

535 - turns --follow-tags option on by default.

536 -

537 - Will merge to 'master'.

538 -

539 -

540 -* jk/tag-h-column-is-a-listing-option (2015-03-12) 1 commit

541 - (merged to 'next' on 2015-03-16 at 42b04c6)

542 - + tag: fix some mis-organized options in "-h" listing

543 -

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

545 - that are about listing in a wrong section.

546 -

528 Will merge to 'master'.

529

530

@@ -574,7 +555,7 @@ of the repositories listed at

555 of hardcoded 20 as byte count and/or (2) use struct object_id

556 instead of unsigned char [20] for object names.

557

577 - Will cook in 'next'.

558 + Will merge to and cook in 'next'.

559

560

561 * ct/prompt-untracked-fix (2015-03-15) 1 commit

@@ -587,39 +568,6 @@ of the repositories listed at

568 Will merge to 'master'.

569

570

590 -* jk/smart-http-hide-refs (2015-03-12) 2 commits

591 - (merged to 'next' on 2015-03-16 at 530df4c)

592 - + upload-pack: do not check NULL return of lookup_unknown_object

593 - + upload-pack: fix transfer.hiderefs over smart-http

594 -

595 - The transfer.hiderefs support did not quite work for smart-http

596 - transport.

597 -

598 - Will merge to 'master'.

599 -

600 -

601 -* jk/test-annoyances (2015-03-12) 5 commits

602 - (merged to 'next' on 2015-03-16 at 845b091)

603 - + t5551: make EXPENSIVE test cheaper

604 - + t5541: move run_with_cmdline_limit to test-lib.sh

605 - + t: pass GIT_TRACE through Apache

606 - + t: redirect stderr GIT_TRACE to descriptor 4

607 - + t: translate SIGINT to an exit

608 -

609 - Test fixes.

610 -

611 - Will merge to 'master'.

612 -

613 -

614 -* nd/config-doc-camelCase (2015-03-13) 1 commit

615 - (merged to 'next' on 2015-03-16 at 0e3fedb)

616 - + *config.txt: stick to camelCase naming convention

617 -

618 - Documentation updates.

619 -

620 - Will merge to 'master'.

621 -

622 -

571 * dj/log-graph-with-no-walk (2015-03-19) 1 commit

572 (merged to 'next' on 2015-03-20 at cb636c0)

573 + revision: forbid combining --graph and --no-walk

What's cooking (2015/03 #08)