1291402d18 · 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 (Apr 2016, #02; Mon, 4)

4 -X-master-at: 6a269e52a5d0c82567c5b39cc194cc90f9fa9156

5 -X-next-at: a49ec4a3520e76c257792c73d6347383649ee5c2

3 +Subject: What's cooking in git.git (Apr 2016, #03; Thu, 7)

4 +X-master-at: 72d917a7f9a5123ef7c0d6db306bbc0d385227e2

5 +X-next-at: 209d99dbd5bc1cff6fe9618b5ae6dc65a6d421d9

6

7 -What's cooking in git.git (Apr 2016, #02; Mon, 4)

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

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

9

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

@@ -12,10 +12,8 @@ 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 tip of 'next' has been rewound, the first maintenance release

16 -v2.8.1 that most everybody can safely ignore has been tagged, and

17 -the first batch of topics that have been cooking in 'next' during

18 -the pre-2.8 freeze period are now in 'master'.

15 +The 'master' branch now has the second batch of topics of this

16 +cycle.

17

18 There are a handful of topics that are stuck; they are marked as

19 "Needs review", "Needs an Ack", etc. in the following list.

@@ -25,393 +23,285 @@ of the repositories listed at

23

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

25

28 ---------------------------------------------------

29 -[New Topics]

30 -

31 -* es/format-patch-doc-hide-no-patch (2016-04-04) 1 commit

32 - - git-format-patch.txt: don't show -s as shorthand for multiple options

33 -

34 - "git format-patch --help" showed `-s` and `--no-patch` as if these

35 - are valid options to the command. We already hide `--patch` option

36 - from the documentation, because format-patch is about showing the

37 - diff, and the documentation now hides these options as well.

38 -

39 - Will merge to next.

40 -

41 -

42 -* jk/branch-shortening-funny-symrefs (2016-04-04) 1 commit

43 - - branch: fix shortening of non-remote symrefs

44 -

45 - A change back in version 2.7 to "git branch" broke display of a

46 - symbolic refs in a non-standard place in the refs/ hierarchy (we

47 - expect symbolic refs to appear in refs/remotes/*/HEAD to point at

48 - the primary branch the remote has, and as .git/HEAD to point at the

49 - branch we locally checked out).

50 -

51 - Will merge to next.

52 - and later down to maint-2.7.

53 -

26 --------------------------------------------------

27 [Graduated to "master"]

28

57 -* es/test-gpg-tags (2016-03-06) 4 commits

58 - (merged to 'next' on 2016-03-15 at 617119f)

59 - + t6302: skip only signed tags rather than all tests when GPG is missing

60 - + t6302: also test annotated in addition to signed tags

61 - + t6302: normalize names and descriptions of signed tags

62 - + lib-gpg: drop unnecessary "missing GPG" warning

63 -

64 - A test for tags has been restructured so that more parts of it can

65 - easily be run on a platform without a working GnuPG.

29 +* cc/doc-recommend-performance-trace-to-file (2016-03-07) 1 commit

30 + (merged to 'next' on 2016-04-04 at 26f94c0)

31 + + Documentation: talk about pager in api-trace.txt

32

33 + Originally merged to 'next' on 2016-03-23

34

68 -* gf/fetch-pack-direct-object-fetch (2016-03-05) 2 commits

69 - (merged to 'next' on 2016-03-06 at 5628860)

70 - + fetch-pack: update the documentation for "<refs>..." arguments

71 - (merged to 'next' on 2016-03-04 at 49199e0)

72 - + fetch-pack: fix object_id of exact sha1

35 + A minor documentation update.

36

74 - Fetching of history by naming a commit object name directly didn't

75 - work across remote-curl transport.

37

38 +* da/mergetool-delete-delete-conflict (2016-03-10) 2 commits

39 + (merged to 'next' on 2016-04-04 at 34e645f)

40 + + mergetool: honor tempfile configuration when resolving delete conflicts

41 + + mergetool: support delete/delete conflicts

42

78 -* jc/index-pack (2016-03-03) 2 commits

79 - (merged to 'next' on 2016-03-15 at 42efaa7)

80 - + index-pack: add a helper function to derive .idx/.keep filename

81 - + Merge branch 'jc/maint-index-pack-keep' into jc/index-pack

82 - (this branch is used by jc/bundle; uses jc/maint-index-pack-keep; is tangled with jc/index-pack-clone-bundle.)

43 + Originally merged to 'next' on 2016-03-15

44

84 - Code clean-up.

45 + "git mergetool" did not work well with conflicts that both sides

46 + deleted.

47

48

87 -* jc/maint-index-pack-keep (2016-03-03) 1 commit

88 - (merged to 'next' on 2016-03-04 at bc1d37a)

89 - + index-pack: correct --keep[=<msg>]

90 - (this branch is used by jc/bundle, jc/index-pack and jc/index-pack-clone-bundle.)

49 +* jk/credential-cache-comment-exit (2016-03-18) 1 commit

50 + (merged to 'next' on 2016-04-04 at 50427fe)

51 + + credential-cache--daemon: clarify "exit" action semantics

52

92 - "git index-pack --keep[=<msg>] pack-$name.pack" simply did not work.

53 + Originally merged to 'next' on 2016-03-23

54

55 + A code clarification.

56

95 -* jk/add-i-highlight (2016-02-28) 1 commit

96 - (merged to 'next' on 2016-03-04 at 4ac3aa1)

97 - + add--interactive: allow custom diff highlighting programs

57

99 - A new "interactive.diffFilter" configuration can be used to

100 - customize the diff shown in "git add -i" session.

58 +* jk/send-email-rtrim-mailrc-alias (2016-03-18) 1 commit

59 + (merged to 'next' on 2016-04-04 at 4d54956)

60 + + send-email: ignore trailing whitespace in mailrc alias file

61

62 + Originally merged to 'next' on 2016-03-23

63

103 -* jk/config-get-urlmatch (2016-02-28) 3 commits

104 - (merged to 'next' on 2016-03-04 at feeb192)

105 - + Documentation/git-config: fix --get-all description

106 - + Documentation/git-config: use bulleted list for exit codes

107 - + config: fail if --get-urlmatch finds no value

64 + "git send-email" had trouble parsing alias file in mailrc format

65 + when lines in it had trailing whitespaces on them.

66

109 - "git config --get-urlmatch", unlike other variants of the "git

110 - config --get" family, did not signal error with its exit status

111 - when there was no matching configuration.

67

68 +* jk/submodule-c-credential (2016-03-23) 7 commits

69 + (merged to 'next' on 2016-04-04 at 8de8e8c)

70 + + git_config_push_parameter: handle empty GIT_CONFIG_PARAMETERS

71 + + git: submodule honor -c credential.* from command line

72 + + quote: implement sq_quotef()

73 + + submodule: fix segmentation fault in submodule--helper clone

74 + + submodule: fix submodule--helper clone usage

75 + + submodule: check argc count for git submodule--helper clone

76 + + submodule: don't pass empty string arguments to submodule--helper clone

77

114 -* jk/credential-clear-config (2016-02-26) 1 commit

115 - (merged to 'next' on 2016-03-04 at f7b26b7)

116 - + credential: let empty credential specs reset helper list

78 + Originally merged to 'next' on 2016-03-23

79

118 - The credential.helper configuration variable is cumulative and

119 - there is no good way to override it from the command line. As

120 - a special case, giving an empty string as its value now serves

121 - as the signal to clear the values specified in various files.

80 + "git -c credential.<var>=<value> submodule" can now be used to

81 + propagate configuration variables related to credential helper

82 + down to the submodules.

83

84

124 -* jk/getwholeline-getdelim-empty (2016-03-05) 1 commit

125 - (merged to 'next' on 2016-03-15 at e70d4bd)

126 - + strbuf_getwholeline: NUL-terminate getdelim buffer on error

85 +* jk/test-httpd-config-nosystem (2016-03-18) 1 commit

86 + (merged to 'next' on 2016-04-04 at 5fa6274)

87 + + t/lib-httpd: pass through GIT_CONFIG_NOSYSTEM env

88

128 - strbuf_getwholeline() did not NUL-terminate the buffer on certain

129 - corner cases in its error codepath.

89 + Originally merged to 'next' on 2016-03-23

90

91 + The tests that involve running httpd leaked the system-wide

92 + configuration in /etc/gitconfig to the tested environment.

93

132 -* jk/rev-parse-local-env-vars (2016-02-29) 2 commits

133 - (merged to 'next' on 2016-03-04 at a0300d5)

134 - + rev-parse: let some options run outside repository

135 - + t1515: add tests for rev-parse out-of-repo helpers

94

137 - The "--local-env-vars" and "--resolve-git-dir" options of "git

138 - rev-parse" failed to work outside a repository when the command's

139 - option parsing was rewritten in 1.8.5 era.

95 +* jv/merge-nothing-into-void (2016-03-23) 1 commit

96 + (merged to 'next' on 2016-04-04 at aa37405)

97 + + merge: fix NULL pointer dereference when merging nothing into void

98

99 + Originally merged to 'next' on 2016-03-23

100

142 -* jk/startup-info (2016-03-07) 6 commits

143 - (merged to 'next' on 2016-03-15 at eb95e5f)

144 - + use setup_git_directory() in test-* programs

145 - + grep: turn off gitlink detection for --no-index

146 - + mailmap: do not resolve blobs in a non-repository

147 - + remote: don't resolve HEAD in non-repository

148 - + setup: set startup_info->have_repository more reliably

149 - + setup: make startup_info available everywhere

101 + "git merge FETCH_HEAD" dereferenced NULL pointer when merging

102 + nothing into an unborn history (which is arguably unusual usage,

103 + which perhaps was the reason why nobody noticed it).

104

151 - The startup_info data, which records if we are working inside a

152 - repository (among other things), are now uniformly available to Git

153 - subcommand implementations, and Git avoids attempting to touch

154 - references when we are not in a repository.

105

106 +* la/tag-force-signing-annotated-tags (2016-03-22) 1 commit

107 + (merged to 'next' on 2016-04-04 at a49ec4a)

108 + + tag: add the option to force signing of annotated tags

109

157 -* mm/diff-renames-default (2016-02-25) 5 commits

158 - (merged to 'next' on 2016-02-25 at 947c399)

159 - + diff: activate diff.renames by default

160 - + log: introduce init_log_defaults()

161 - + t: add tests for diff.renames (true/false/unset)

162 - + t4001-diff-rename: wrap file creations in a test

163 - + Documentation/diff-config: fix description of diff.renames

110 + Originally merged to 'next' on 2016-03-24

111

165 - The end-user facing Porcelain level commands like "diff" and "log"

166 - now enables the rename detection by default.

112 + "git tag" can create an annotated tag without explicitly given an

113 + "-a" (or "-s") option (i.e. when a tag message is given). A new

114 + configuration variable, tag.forceSignAnnotated, can be used to tell

115 + the command to create signed tag in such a situation.

116

117

169 -* mm/lockfile-error-message (2016-03-01) 2 commits

170 - (merged to 'next' on 2016-03-04 at 04ed7e6)

171 - + lockfile: improve error message when lockfile exists

172 - + lockfile: mark strings for translation

118 +* ls/p4-map-user (2016-03-15) 1 commit

119 + (merged to 'next' on 2016-04-04 at a56b011)

120 + + git-p4: map a P4 user to Git author name and email address

121

122 + Originally merged to 'next' on 2016-03-23

123

175 -* mm/readme-markdown (2016-04-01) 2 commits

176 - - git.spec: use README.md, not README

177 - (merged to 'next' on 2016-03-01 at 81f3858)

178 - + README.md: don't take 'commandname' literally

124 + "git p4" now allows P4 author names to be mapped to Git author

125 + names.

126

180 - The top-level README file has been updated to be more appropriate

181 - for the sign on the front door to welcome new acquaintances to Git

182 - by toning down inside jokes and making it into MarkDown.

127

128 +* pb/t7502-drop-dup (2016-03-11) 1 commit

129 + (merged to 'next' on 2016-04-04 at 4799cad)

130 + + t/t7502 : drop duplicate test

131 + (this branch is used by pb/commit-verbose-config.)

132

185 -* mp/upload-pack-use-embedded-args (2016-02-25) 1 commit

186 - (merged to 'next' on 2016-02-26 at f0a54e5)

187 - + upload-pack: use argv_array for pack_objects

133 + Originally merged to 'next' on 2016-03-15

134

189 - The embedded args argv-array in the child process is used to build

190 - the command line to run pack-objects instead of using a separate

191 - array of strings.

135 + Code clean-up.

136

137

194 -* rj/xdiff-prepare-plug-leak-on-error-codepath (2016-03-04) 2 commits

195 - (merged to 'next' on 2016-03-15 at f72755e)

196 - + xdiff/xprepare: fix a memory leak

197 - + xdiff/xprepare: use the XDF_DIFF_ALG() macro to access flag bits

138 +* sb/clone-t57-t56 (2016-03-16) 1 commit

139 + (merged to 'next' on 2016-04-04 at 5c20247)

140 + + clone tests: rename t57* => t56*

141

199 - A small memory leak in an error codepath has been plugged in xdiff

200 - code.

142 + Originally merged to 'next' on 2016-03-23

143

202 ---------------------------------------------------

203 -[Stalled]

144 + Rename bunch of tests on "git clone" for better organization.

145

205 -* ec/annotate-deleted (2015-11-20) 1 commit

206 - - annotate: skip checking working tree if a revision is provided

146

208 - Usability fix for annotate-specific "<file> <rev>" syntax with deleted

209 - files.

147 +* sb/rebase-x (2016-03-18) 2 commits

148 + (merged to 'next' on 2016-04-04 at feda620)

149 + + t3404: cleanup double empty lines between tests

150 + + rebase: decouple --exec from --interactive

151

211 - Waiting for review.

152 + Originally merged to 'next' on 2016-03-23

153

154 + "git rebase -x" can be used without passing "-i" option.

155

214 -* dg/subtree-rebase-test (2016-01-19) 1 commit

215 - - contrib/subtree: Add a test for subtree rebase that loses commits

156

217 - Reviewed up to v5.

218 - Will be rerolled.

219 - ($gmane/284426)

157 +* sb/submodule-parallel-update (2016-03-01) 10 commits

158 + (merged to 'next' on 2016-04-04 at a0aea8d)

159 + + clone: allow an explicit argument for parallel submodule clones

160 + + submodule update: expose parallelism to the user

161 + + submodule helper: remove double 'fatal: ' prefix

162 + + git submodule update: have a dedicated helper for cloning

163 + + run_processes_parallel: rename parameters for the callbacks

164 + + run_processes_parallel: treat output of children as byte array

165 + + submodule update: direct error message to stderr

166 + + fetching submodules: respect `submodule.fetchJobs` config option

167 + + submodule-config: drop check against NULL

168 + + submodule-config: keep update strategy around

169 + (this branch is used by sb/clone-shallow-passthru and sb/submodule-init.)

170

171 + Originally merged to 'next' on 2016-03-15

172

222 -* js/am-3-merge-recursive-direct (2015-10-12) 2 commits

223 - - am: make a direct call to merge_recursive

224 - - merge_recursive_options: introduce the "gently" flag

173 + A major part of "git submodule update" has been ported to C to take

174 + advantage of the recently added framework to run download tasks in

175 + parallel.

176

226 - The merge_recursive_generic() function has been made a bit safer to

227 - call from inside a process. "git am -3" was taught to make a direct

228 - call to the function when falling back to three-way merge.

177

230 - Being able to make a direct call would be good in general, but as a

231 - performance thing, the change needs to be backed up by numbers.

178 +* ss/commit-squash-msg (2016-03-21) 1 commit

179 + (merged to 'next' on 2016-04-04 at d389f19)

180 + + commit: do not lose SQUASH_MSG contents

181

233 - Needs review.

182 + Originally merged to 'next' on 2016-03-23

183

235 - I haven't gone through the "gently" change with fine toothed comb;

236 - I can see that the change avoids calling die(), but I haven't made

237 - sure that the program states (e.g. what's in the in-core index) are

238 - adjusted sensibly when it returns to the caller instead of dying,

239 - or the codepaths that used to die() are free of resource leaks.

240 - The original code certainly did not care the program states at the

241 - point of dying exactly because it knew it is going to exit, but now

242 - they have to care, and they need to be audited.

184 + When "git merge --squash" stopped due to conflict, the concluding

185 + "git commit" failed to read in the SQUASH_MSG that shows the log

186 + messages from all the squashed commits.

187

188

245 -* dk/gc-more-wo-pack (2016-01-13) 4 commits

246 - - gc: clean garbage .bitmap files from pack dir

247 - - t5304: ensure non-garbage files are not deleted

248 - - t5304: test .bitmap garbage files

249 - - prepare_packed_git(): find more garbage

189 +* ss/exc-flag-is-a-collection-of-bits (2016-03-01) 1 commit

190 + (merged to 'next' on 2016-04-04 at 9f0207e)

191 + + dir: store EXC_FLAG_* values in unsigned integers

192

251 - Follow-on to dk/gc-idx-wo-pack topic, to clean up stale

252 - .bitmap and .keep files.

193 + Originally merged to 'next' on 2016-03-04

194

254 - Waiting for a reroll.

255 - ($gmane/284368).

195 + Code clean-up.

196

197

258 -* jc/diff-b-m (2015-02-23) 5 commits

259 - . WIPWIP

260 - . WIP: diff-b-m

261 - - diffcore-rename: allow easier debugging

262 - - diffcore-rename.c: add locate_rename_src()

263 - - diffcore-break: allow debugging

198 +* ss/receive-pack-parse-options (2016-03-01) 1 commit

199 + (merged to 'next' on 2016-04-04 at fd6ab4c)

200 + + builtin/receive-pack.c: use parse_options API

201

265 - "git diff -B -M" produced incorrect patch when the postimage of a

266 - completely rewritten file is similar to the preimage of a removed

267 - file; such a resulting file must not be expressed as a rename from

268 - other place.

202 + Originally merged to 'next' on 2016-03-04

203

270 - The fix in this patch is broken, unfortunately.

271 - Will discard.

204 + The command line argument parser for "receive-pack" has been

205 + rewritten to use parse-options.

206

207 --------------------------------------------------

274 -[Cooking]

275 -

276 -* jn/mergetools-examdiff (2016-04-04) 2 commits

277 - - mergetools: add support for ExamDiff

278 - - mergetools: create mergetool_find_win32_cmd() helper function for winmerge

208 +[New Topics]

209

280 - "git mergetools" learned to drive ExamDiff.

210 +* jc/drop-git-spec-in (2016-04-06) 1 commit

211 + - Makefile: stop pretending to support rpmbuild

212

282 - Will merge to 'next'.

213 + As nobody maintains our in-tree git.spec.in and distros use their

214 + own spec file, we stopped pretending that we support "make rpm".

215

216

285 -* kn/for-each-tag-branch (2016-03-30) 1 commit

286 - - for-each-ref: fix description of '--contains' in manpage

217 +* jc/makefile-redirection-stderr (2016-04-05) 1 commit

218 + (merged to 'next' on 2016-04-06 at e3f2ded)

219 + + Makefile: fix misdirected redirections

220

288 - Docfix.

221 + A minor fix in the Makefile.

222

290 - Will merge to 'next'.

223 + Will merge to 'master'.

224

225

293 -* kn/ref-filter-branch-list (2016-03-30) 16 commits

294 - . branch: implement '--format' option

295 - . branch: use ref-filter printing APIs

296 - . branch, tag: use porcelain output

297 - . ref-filter: allow porcelain to translate messages in the output

298 - . ref-filter: add support for %(refname:dir) and %(refname:base)

299 - . ref-filter: introduce refname_atom_parser()

300 - . ref-filter: introduce symref_atom_parser()

301 - . ref-filter: make "%(symref)" atom work with the ':short' modifier

302 - . ref-filter: add support for %(upstream:track,nobracket)

303 - . ref-filter: make %(upstream:track) prints "[gone]" for invalid upstreams

304 - . ref-filter: introduce format_ref_array_item()

305 - . ref-filter: move get_head_description() from branch.c

306 - . ref-filter: modify "%(objectname:short)" to take length

307 - . ref-filter: implement %(if:equals=<string>) and %(if:notequals=<string>)

308 - . ref-filter: include reference to 'used_atom' within 'atom_value'

309 - . ref-filter: implement %(if), %(then), and %(else) atoms

226 +* js/mingw-tests-2.8 (2016-04-04) 1 commit

227 + (merged to 'next' on 2016-04-06 at f85a013)

228 + + Windows: shorten code by re-using convert_slashes()

229

311 - The code to list branches in "git branch" has been consolidated

312 - with the more generic ref-filter API.

230 + Code clean-up.

231

314 - Will be rerolled.

232 + Will merge to 'master'.

233

234

317 -* oa/doc-diff-check (2016-03-29) 1 commit

318 - - Documentation: git diff --check detects conflict markers

235 +* ar/diff-args-osx-precompose (2016-04-05) 1 commit

236 + - diff: run arguments through precompose_argv

237

320 - Docfix.

238 + Many commands normalize command line arguments from NFD to NFC

239 + variant of UTF-8 on OSX, but commands in the "diff" family did

240 + not, causing "git diff $path" to complain that no such path is

241 + known to Git. They have been taught to do the normalization.

242

322 - Will merge to 'next'.

243 + Will be rerolled?

244 + ($gmane/290724)

245

246

325 -* rz/worktree-no-checkout (2016-03-29) 1 commit

326 - - worktree: add: introduce --checkout option

247 +* ep/trace-doc-sample-fix (2016-04-05) 1 commit

248 + (merged to 'next' on 2016-04-06 at 0df7357)

249 + + api-trace.txt: fix typo

250

328 - "git worktree add" can be given "--no-checkout" option to only

329 - create an empty worktree without checking out the files.

251 + Fix a typo in an example in the trace API documentation.

252

331 - Will merge to 'next'.

253 + Will merge to 'master'.

254

255

334 -* sb/misc-cleanups (2016-04-01) 4 commits

335 - - credential-cache, send_request: close fd when done

336 - - bundle: don't leak an fd in case of early return

337 - - abbrev_sha1_in_line: don't leak memory

338 - - notes: don't leak memory in git_config_get_notes_strategy

256 +* ew/send-email-readable-message-id (2016-04-06) 1 commit

257 + - send-email: more meaningful Message-ID

258

340 - Assorted minor clean-ups.

259 + "git send-email" now uses a more readable timestamps when

260 + formulating a message ID.

261

262 Will merge to 'next'.

263

264

345 -* sb/submodule-helper-clone-regression-fix (2016-04-01) 6 commits

346 - - submodule--helper, module_clone: catch fprintf failure

347 - - submodule--helper: do not borrow absolute_path() result for too long

348 - - submodule--helper, module_clone: always operate on absolute paths

349 - - submodule--helper clone: create the submodule path just once

350 - - submodule--helper: fix potential NULL-dereference

351 - - recursive submodules: test for relative paths

352 -

353 - A partial rewrite of "git submodule" in the 2.7 timeframe changed

354 - the way the gitdir: pointer in the submodules point at the real

355 - repository location to use absolute paths by accident. This has

356 - been corrected.

265 +* mg/complete-cherry-mark-to-log (2016-04-05) 1 commit

266 + (merged to 'next' on 2016-04-06 at 3002be6)

267 + + completion: complete --cherry-mark for git log

268

358 - Any further comments? Otherwise will merge to 'next'.

269 + The completion scripts (in contrib/) did not include the

270 + "--cherry-mark" option when completing "git log <HT>".

271

272 + Will merge to 'master'.

273

361 -* sb/submodule-path-misc-bugs (2016-03-30) 6 commits

362 - - t7407: make expectation as clear as possible

363 - - submodule update: test recursive path reporting from subdirectory

364 - - submodule update: align reporting path for custom command execution

365 - - submodule status: correct path handling in recursive submodules

366 - - submodule update --init: correct path handling in recursive submodules

367 - - submodule foreach: correct path display in recursive submodules

274

369 - "git submodule" reports the paths of submodules the command

370 - recurses into, but this was incorrect when the command was not run

371 - from the root level of the superproject.

275 +* tb/blame-force-read-cache-to-workaround-safe-crlf (2016-04-05) 1 commit

276 + (merged to 'next' on 2016-04-06 at 263bba8)

277 + + correct blame for files commited with CRLF

278

373 - Any further comments? Otherwise will merge to 'next'.

279 + When running "git blame $path" with unnormalized data in the index

280 + for the path, the data in the working tree was blamed, even though

281 + "git add" would not have changed what is already in the index, due

282 + to "safe crlf" that disables the line-end conversion. It has been

283 + corrected.

284

285 + Will merge to 'master'.

286

376 -* sg/diff-multiple-identical-renames (2016-03-30) 1 commit

377 - - diffcore: fix iteration order of identical files during rename detection

287

379 - "git diff -M" used to work better when two originally identical

380 - files A and B got renamed to X/A and X/B by pairing A to X/A and B

381 - to X/B, but this was broken in 2.0 timeframe.

288 +* st/verify-tag (2016-04-06) 3 commits

289 + - verify-tag: change variable name for readability

290 + - t7030: test verifying multiple tags

291 + - builtin/verify-tag.c: ignore SIGPIPE in gpg-interface

292

383 - Will merge to 'next'.

293 + Only the first three patches in a six-patch series.

294

295

386 -* sk/send-pack-all-fix (2016-03-31) 1 commit

387 - - git-send-pack: fix --all option when used with directory

296 +* ew/send-email-drop-data-dumper (2016-04-06) 1 commit

297 + - send-email: do not load Data::Dumper

298

389 - "git send-pack --all <there>" was broken when its command line

390 - option parsing was written in 2.6 timeframe.

299 + Code clean-up.

300

301 Will merge to 'next'.

302

394 -

395 -* ss/msvc (2016-03-30) 2 commits

396 - - MSVC: use shipped headers instead of fallback definitions

397 - - MSVC: vsnprintf in Visual Studio 2015 doesn't need SNPRINTF_SIZE_CORR any more

398 -

399 - Will merge to 'next'.

400 -

401 -

402 -* xy/format-patch-base (2016-03-31) 4 commits

403 - - format-patch: introduce format.base configuration

404 - - format-patch: introduce --base=auto option

405 - - format-patch: add '--base' option to record base tree info

406 - - patch-ids: make commit_patch_id() a public helper function

407 -

408 - "git format-patch" learned a new "--base" option to record what

409 - (public, well-known) commit the original series was built on in

410 - its output.

411 -

412 - Will be rerolled.

413 - ($gmane/290365)

414 -

303 +--------------------------------------------------

304 +[Stalled]

305

306 * da/user-useconfigonly (2016-04-01) 2 commits

307 - ident: give "please tell me" message upon useConfigOnly error

@@ -429,602 +319,658 @@ of the repositories listed at

319 ($gmane/290340)

320

321

432 -* tb/safe-crlf-output-fix (2016-04-01) 7 commits

433 - . convert.c: more safer crlf handling with text attribute

434 - . correct blame for files commited with CRLF

435 - . convert: unify the "auto" handling of CRLF

436 - . t0027: test cases for combined attributes

437 - . convert: allow core.autocrlf=input and core.eol=crlf

438 - . convert.c: stream and early out

439 - . read-cache: factor out get_sha1_from_index() helper

440 -

441 - The "safe CRLF" facility disables line-end conversion from CRLF to

442 - LF when checking in if the blob registered to the index already

443 - contains CR, but some codepaths like "git blame" did not know this,

444 - and instead assumed that only the configuration and attribute

445 - settings determined how the data from the working tree is converted.

446 -

447 - Will be rerolled.

448 - ($gmane/290637)

449 -

450 -

451 -* ak/use-hashmap-iter-first-in-submodule-config (2016-03-23) 1 commit

452 - - submodule-config: use hashmap_iter_first()

453 -

454 - Minor code cleanup.

455 -

456 - Will merge to 'next'.

457 -

458 -

459 -* ky/branch-d-worktree (2016-03-29) 1 commit

460 - - branch -d: refuse deleting a branch which is currently checked out

461 -

462 - When "git worktree" feature is in use, "git branch -d" allowed

463 - deletion of a branch that is checked out in another worktree

464 -

465 - Will merge to 'next'.

322 +* sb/clone-shallow-passthru (2016-03-23) 3 commits

323 + - clone: add t5614 to test cloning submodules with shallowness involved

324 + - submodule clone: pass along `local` option

325 + - clone: add `--shallow-submodules` flag

326

327 + "git clone" learned "--shallow-submodules" option.

328

468 -* ky/branch-m-worktree (2016-04-04) 2 commits

469 - - branch -m: update all per-worktree HEADs

470 - - refs: add a new function set_worktree_head_symref

329 + Needs review.

330

472 - When "git worktree" feature is in use, "git branch -m" renamed a

473 - branch that is checked out in another worktree without adjusting

474 - the HEAD symbolic ref for the worktree.

331

476 - Will merge to 'next'.

332 +* sg/completion-updates (2016-02-28) 21 commits

333 + - completion: cache the path to the repository

334 + - completion: extract repository discovery from __gitdir()

335 + - completion: don't guard git executions with __gitdir()

336 + - completion: consolidate silencing errors from git commands

337 + - completion: don't use __gitdir() for git commands

338 + - completion: respect 'git -C <path>'

339 + - completion: fix completion after 'git -C <path>'

340 + - completion: don't offer commands when 'git --opt' needs an argument

341 + - rev-parse: add '--absolute-git-dir' option

342 + - completion: list short refs from a remote given as a URL

343 + - completion: don't list 'HEAD' when trying refs completion outside of a repo

344 + - completion: list refs from remote when remote's name matches a directory

345 + - completion: respect 'git --git-dir=<path>' when listing remote refs

346 + - completion: fix most spots not respecting 'git --git-dir=<path>'

347 + - completion: ensure that the repository path given on the command line exists

348 + - completion tests: add tests for the __git_refs() helper function

349 + - completion tests: check __gitdir()'s output in the error cases

350 + - completion tests: consolidate getting path of current working directory

351 + - completion tests: make the $cur variable local to the test helper functions

352 + - completion tests: don't add test cruft to the test repository

353 + - completion: improve __git_refs()'s in-code documentation

354

355 + Will be rerolled.

356 + ($gmane/287839)

357

479 -* nd/apply-doc (2016-03-24) 2 commits

480 - - git-apply.txt: mention the behavior inside a subdir

481 - - git-apply.txt: remove a space

358

483 - Will merge to 'next'.

359 +* nd/shallow-deepen (2016-02-23) 25 commits

360 + - fetch, upload-pack: --deepen=N extends shallow boundary by N commits

361 + - upload-pack: add get_reachable_list()

362 + - upload-pack: split check_unreachable() in two, prep for get_reachable_list()

363 + - t5500, t5539: tests for shallow depth excluding a ref

364 + - clone: define shallow clone boundary with --shallow-exclude

365 + - fetch: define shallow boundary with --shallow-exclude

366 + - upload-pack: support define shallow boundary by excluding revisions

367 + - refs: add expand_ref()

368 + - t5500, t5539: tests for shallow depth since a specific date

369 + - clone: define shallow clone boundary based on time with --shallow-since

370 + - fetch: define shallow boundary with --shallow-since

371 + - upload-pack: add deepen-since to cut shallow repos based on time

372 + - shallow.c: implement a generic shallow boundary finder based on rev-list

373 + - fetch-pack: use a separate flag for fetch in deepening mode

374 + - fetch-pack: use a common function for verbose printing

375 + - fetch-pack: use skip_prefix() instead of starts_with()

376 + - upload-pack: move rev-list code out of check_non_tip()

377 + - upload-pack: tighten number parsing at "deepen" lines

378 + - upload-pack: use skip_prefix() instead of starts_with()

379 + - upload-pack: move "unshallow" sending code out of deepen()

380 + - upload-pack: remove unused variable "backup"

381 + - upload-pack: move "shallow" sending code out of deepen()

382 + - upload-pack: move shallow deepen code out of receive_needs()

383 + - transport-helper.c: refactor set_helper_option()

384 + - remote-curl.c: convert fetch_git() to use argv_array

385

386 + The existing "git fetch --depth=<n>" option was hard to use

387 + correctly when making the history of an existing shallow clone

388 + deeper. A new option, "--deepen=<n>", has been added to make this

389 + easier to use. "git clone" also learned "--shallow-since=<date>"

390 + and "--shallow-exclude=<tag>" options to make it easier to specify

391 + "I am interested only in the recent N months worth of history" and

392 + "Give me only the history since that version".

393

486 -* nd/apply-report-skip (2016-03-24) 1 commit

487 - - apply: report patch skipping in verbose mode

394 + Needs review.

395

489 - "git apply -v" learned to report paths in the patch that were

490 - skipped via --include/--exclude mechanism or being outside the

491 - current working directory.

396

493 - Will merge to 'next'.

397 +* sb/submodule-init (2016-03-15) 2 commits

398 + . submodule: port init from shell to C

399 + . submodule: port resolve_relative_url from shell to C

400

401 + Update of "git submodule" to move pieces of logic to C continues.

402

496 -* pb/opt-cmdmode-doc (2016-03-25) 1 commit

497 - - api-parse-options.txt: document OPT_CMDMODE()

403 + Needs review.

404 + ($gmane/288824)

405

499 - Minor API documentation update.

406

501 - Will merge to 'next'.

407 +* az/p4-bare-no-rebase (2016-02-19) 1 commit

408 + - git-p4.py: Don't try to rebase on submit from bare repository

409

410 + "git p4 submit" attempts to do a rebase, which would fail if done

411 + in a bare repository. Not doing this rebase would paper over the

412 + failure, which is what this patch does, but it is unclear what the

413 + side effect of not rebasing is.

414

504 -* rt/completion-help (2016-03-24) 2 commits

505 - - completion: add 'revisions' and 'everyday' to 'git help'

506 - - completion: add option '--guides' to 'git help'

415 + Needs a better explanation.

416

508 - Shell completion (in contrib/) updates.

417

510 - Will merge to 'next'.

418 +* ss/commit-dry-run-resolve-merge-to-no-op (2016-02-17) 1 commit

419 + - wt-status.c: set commitable bit if there is a meaningful merge.

420

421 + "git commit --dry-run" reported "No, no, you cannot commit." in one

422 + case where "git commit" would have allowed you to commit, and this

423 + improves it a little bit ("git commit --dry-run --short" still does

424 + not give you the correct answer, for example).

425

513 -* rt/rebase-i-shorten-stop-report (2016-03-28) 1 commit

514 - - rebase-i: print an abbreviated hash when stop for editing

426

516 - The commit object name reported when "rebase -i" stops has been

517 - shortened.

427 +* nd/icase (2016-02-15) 12 commits

428 + - grep.c: reuse "icase" variable

429 + - diffcore-pickaxe: support case insensitive match on non-ascii

430 + - diffcore-pickaxe: "share" regex error handling code

431 + - grep/pcre: support utf-8

432 + - gettext: add is_utf8_locale()

433 + - grep/pcre: prepare locale-dependent tables for icase matching

434 + - grep/icase: avoid kwsset when -F is specified

435 + - grep/icase: avoid kwsset on literal non-ascii strings

436 + - test-regex: expose full regcomp() to the command line

437 + - test-regex: isolate the bug test code

438 + - grep: break down an "if" stmt in preparation for next changes

439 + - grep: allow -F -i combination

440

519 - Will merge to 'next'.

441 + "git grep -i" has been taught to fold case in non-ascii locales.

442

443 + Needs review.

444 + ($gmane/286137)

445

522 -* jk/check-repository-format (2016-03-11) 10 commits

523 - - verify_repository_format: mark messages for translation

524 - - setup: drop repository_format_version global

525 - - setup: unify repository version callbacks

526 - - init: use setup.c's repo version verification

527 - - setup: refactor repo format reading and verification

528 - - config: drop git_config_early

529 - - check_repository_format_gently: stop using git_config_early

530 - - lazily load core.sharedrepository

531 - - wrap shared_repository global in get/set accessors

532 - - setup: document check_repository_format()

446

534 - The repository set-up sequence has been streamlined (the biggest

535 - change is that there is no longer git_config_early()), so that we

536 - do not attempt to look into refs/* when we know we do not have a

537 - Git repository.

447 +* dt/refs-backend-lmdb (2016-02-25) 45 commits

448 + . SQUASH??? Minimum compilation band-aid

449 + . tests: add ref-storage argument

450 + . refs: tests for lmdb backend

451 + . refs: add LMDB refs storage backend

452 + . refs: break out resolve_ref_unsafe_submodule

453 + . config: read ref storage config on startup

454 + . refs: register ref storage backends

455 + . svn: learn ref-storage argument

456 + . clone: allow ref storage backend to be set for clone

457 + . refs: check submodules' ref storage config

458 + . init: allow alternate ref strorage to be set for new repos

459 + . refs: always handle non-normal refs in files backend

460 + . refs: resolve symbolic refs first

461 + . refs: on symref reflog expire, lock symref not referrent

462 + . refs: don't dereference on rename

463 + . refs: allow log-only updates

464 + . refs: move duplicate check to common code

465 + . refs: make lock generic

466 + . refs: handle non-normal ref renames

467 + . refs: add method to rename refs

468 + . refs: add methods to init refs db

469 + . refs: add method for delete_refs

470 + . refs: add method for initial ref transaction commit

471 + . refs: add methods for reflog

472 + . refs: add do_for_each_per_worktree_ref

473 + . refs: reduce the visibility of do_for_each_ref()

474 + . refs: add method for do_for_each_ref

475 + . refs: add methods for misc ref operations

476 + . refs: add a backend method structure with transaction functions

477 + . refs: move resolve_ref_unsafe into common code

478 + . files-backend: break out ref reading

479 + . refs: move for_each_*ref* functions into common code

480 + . refs: move head_ref{,_submodule} to the common code

481 + . Merge branch 'sb/submodule-parallel-update' into dt/refs-backend-lmdb

482 + . clone: allow an explicit argument for parallel submodule clones

483 + . submodule update: expose parallelism to the user

484 + . git submodule update: have a dedicated helper for cloning

485 + . run_processes_parallel: correctly terminate callbacks with an LF

486 + . run_processes_parallel: rename parameters for the callbacks

487 + . run-command: expose default_{start_failure, task_finished}

488 + . run_processes_parallel: treat output of children as byte array

489 + . submodule update: direct error message to stderr

490 + . fetching submodules: respect `submodule.fetchJobs` config option

491 + . submodule-config: drop check against NULL

492 + . submodule-config: keep update strategy around

493

539 - Will merge to 'next'.

494 + A reroll exists, but it seems that will further be rerolled.

495

496

542 -* mj/pull-rebase-autostash (2016-04-04) 9 commits

543 - - t5520: test --[no-]autostash with pull.rebase=true

544 - - t5520: reduce commom lines of code

545 - - t5520: factor out common "failing autostash" code

546 - - t5520: factor out common "successful autostash" code

547 - - t5520: use better test to check stderr output

548 - - t5520: ensure consistent test conditions

549 - - t5520: use consistent capitalization in test titles

550 - - pull --rebase: add --[no-]autostash flag

551 - - git-pull.c: introduce git_pull_config()

497 +* ec/annotate-deleted (2015-11-20) 1 commit

498 + - annotate: skip checking working tree if a revision is provided

499

553 - "git pull --rebase" learned "--[no-]autostash" option, so that

554 - the rebase.autostash configuration variable set to true can be

555 - overridden from the command line.

500 + Usability fix for annotate-specific "<file> <rev>" syntax with deleted

501 + files.

502

557 - Will merge to 'next'.

503 + Waiting for review.

504

505

560 -* pb/commit-verbose-config (2016-03-14) 1 commit

561 - - commit: add a commit.verbose config variable

562 - (this branch uses pb/t7502-drop-dup.)

506 +* dg/subtree-rebase-test (2016-01-19) 1 commit

507 + - contrib/subtree: Add a test for subtree rebase that loses commits

508

564 - "git commit" learned to pay attention to "commit.verbose"

565 - configuration variable and act as if "--verbose" option was

566 - given from the command line.

509 + Reviewed up to v5.

510 + Will be rerolled.

511 + ($gmane/284426)

512

568 - Will merge to 'next'.

513

514 +* js/am-3-merge-recursive-direct (2015-10-12) 2 commits

515 + - am: make a direct call to merge_recursive

516 + - merge_recursive_options: introduce the "gently" flag

517

571 -* pb/t7502-drop-dup (2016-03-11) 1 commit

572 - (merged to 'next' on 2016-04-04 at 4799cad)

573 - + t/t7502 : drop duplicate test

574 - (this branch is used by pb/commit-verbose-config.)

518 + The merge_recursive_generic() function has been made a bit safer to

519 + call from inside a process. "git am -3" was taught to make a direct

520 + call to the function when falling back to three-way merge.

521

576 - Originally merged to 'next' on 2016-03-15

522 + Being able to make a direct call would be good in general, but as a

523 + performance thing, the change needs to be backed up by numbers.

524

578 - Code clean-up.

525 + Needs review.

526

580 - Will merge to 'master' after 2.8 final.

527 + I haven't gone through the "gently" change with fine toothed comb;

528 + I can see that the change avoids calling die(), but I haven't made

529 + sure that the program states (e.g. what's in the in-core index) are

530 + adjusted sensibly when it returns to the caller instead of dying,

531 + or the codepaths that used to die() are free of resource leaks.

532 + The original code certainly did not care the program states at the

533 + point of dying exactly because it knew it is going to exit, but now

534 + they have to care, and they need to be audited.

535

536

583 -* ss/commit-squash-msg (2016-03-21) 1 commit

584 - (merged to 'next' on 2016-04-04 at d389f19)

585 - + commit: do not lose SQUASH_MSG contents

537 +* dk/gc-more-wo-pack (2016-01-13) 4 commits

538 + - gc: clean garbage .bitmap files from pack dir

539 + - t5304: ensure non-garbage files are not deleted

540 + - t5304: test .bitmap garbage files

541 + - prepare_packed_git(): find more garbage

542

587 - Originally merged to 'next' on 2016-03-23

543 + Follow-on to dk/gc-idx-wo-pack topic, to clean up stale

544 + .bitmap and .keep files.

545

589 - When "git merge --squash" stopped due to conflict, the concluding

590 - "git commit" failed to read in the SQUASH_MSG that shows the log

591 - messages from all the squashed commits.

546 + Waiting for a reroll.

547 + ($gmane/284368).

548

593 - Will merge to 'master' after 2.8 final.

549

550 +* jc/diff-b-m (2015-02-23) 5 commits

551 + . WIPWIP

552 + . WIP: diff-b-m

553 + - diffcore-rename: allow easier debugging

554 + - diffcore-rename.c: add locate_rename_src()

555 + - diffcore-break: allow debugging

556

596 -* ls/p4-map-user (2016-03-15) 1 commit

597 - (merged to 'next' on 2016-04-04 at a56b011)

598 - + git-p4: map a P4 user to Git author name and email address

557 + "git diff -B -M" produced incorrect patch when the postimage of a

558 + completely rewritten file is similar to the preimage of a removed

559 + file; such a resulting file must not be expressed as a rename from

560 + other place.

561

600 - Originally merged to 'next' on 2016-03-23

562 + The fix in this patch is broken, unfortunately.

563 + Will discard.

564

602 - "git p4" now allows P4 author names to be mapped to Git author

603 - names.

565 +--------------------------------------------------

566 +[Cooking]

567

605 - Will merge to 'master' after 2.8 final.

568 +* es/format-patch-doc-hide-no-patch (2016-04-04) 1 commit

569 + (merged to 'next' on 2016-04-06 at 25d79bb)

570 + + git-format-patch.txt: don't show -s as shorthand for multiple options

571

572 + "git format-patch --help" showed `-s` and `--no-patch` as if these

573 + are valid options to the command. We already hide `--patch` option

574 + from the documentation, because format-patch is about showing the

575 + diff, and the documentation now hides these options as well.

576

608 -* jc/merge-refuse-new-root (2016-03-23) 1 commit

609 - (merged to 'next' on 2016-04-04 at cd70fd6)

610 - + merge: refuse to create too cool a merge by default

577 + Will merge to 'master'.

578

612 - Originally merged to 'next' on 2016-03-23

579

614 - "git merge" used to allow merging two branches that have no common

615 - base by default, which led to a brand new history of an existing

616 - project created and then get pulled by an unsuspecting maintainer,

617 - which allowed an unnecessary parallel history merged into the

618 - existing project. The command has been taught not to allow this by

619 - default, with an escape hatch "--allow-unrelated-histories" option

620 - to be used in a rare event that merges histories of two projects

621 - that started their lives independently.

580 +* jk/branch-shortening-funny-symrefs (2016-04-04) 1 commit

581 + (merged to 'next' on 2016-04-06 at 1a3f8be)

582 + + branch: fix shortening of non-remote symrefs

583

623 - Will merge to 'master' after 2.8 final.

584 + A change back in version 2.7 to "git branch" broke display of a

585 + symbolic ref in a non-standard place in the refs/ hierarchy (we

586 + expect symbolic refs to appear in refs/remotes/*/HEAD to point at

587 + the primary branch the remote has, and as .git/HEAD to point at the

588 + branch we locally checked out).

589

590 + Will merge to 'next' and later down to maint-2.7.

591

626 -* jk/credential-cache-comment-exit (2016-03-18) 1 commit

627 - (merged to 'next' on 2016-04-04 at 50427fe)

628 - + credential-cache--daemon: clarify "exit" action semantics

592

630 - Originally merged to 'next' on 2016-03-23

593 +* jn/mergetools-examdiff (2016-04-04) 2 commits

594 + (merged to 'next' on 2016-04-06 at 819e858)

595 + + mergetools: add support for ExamDiff

596 + + mergetools: create mergetool_find_win32_cmd() helper function for winmerge

597

632 - A code clarification.

598 + "git mergetools" learned to drive ExamDiff.

599

634 - Will merge to 'master' after 2.8 final.

600 + Will merge to 'master'.

601

602

637 -* jk/send-email-rtrim-mailrc-alias (2016-03-18) 1 commit

638 - (merged to 'next' on 2016-04-04 at 4d54956)

639 - + send-email: ignore trailing whitespace in mailrc alias file

603 +* kn/for-each-tag-branch (2016-03-30) 1 commit

604 + (merged to 'next' on 2016-04-06 at 4595ad3)

605 + + for-each-ref: fix description of '--contains' in manpage

606

641 - Originally merged to 'next' on 2016-03-23

607 + A minor documentation update.

608

643 - "git send-email" had trouble parsing alias file in mailrc format

644 - when lines in it had trailing whitespaces on them.

609 + Will merge to 'master'.

610

646 - Will merge to 'master' after 2.8 final.

611

612 +* kn/ref-filter-branch-list (2016-03-30) 16 commits

613 + . branch: implement '--format' option

614 + . branch: use ref-filter printing APIs

615 + . branch, tag: use porcelain output

616 + . ref-filter: allow porcelain to translate messages in the output

617 + . ref-filter: add support for %(refname:dir) and %(refname:base)

618 + . ref-filter: introduce refname_atom_parser()

619 + . ref-filter: introduce symref_atom_parser()

620 + . ref-filter: make "%(symref)" atom work with the ':short' modifier

621 + . ref-filter: add support for %(upstream:track,nobracket)

622 + . ref-filter: make %(upstream:track) prints "[gone]" for invalid upstreams

623 + . ref-filter: introduce format_ref_array_item()

624 + . ref-filter: move get_head_description() from branch.c

625 + . ref-filter: modify "%(objectname:short)" to take length

626 + . ref-filter: implement %(if:equals=<string>) and %(if:notequals=<string>)

627 + . ref-filter: include reference to 'used_atom' within 'atom_value'

628 + . ref-filter: implement %(if), %(then), and %(else) atoms

629

649 -* jk/test-httpd-config-nosystem (2016-03-18) 1 commit

650 - (merged to 'next' on 2016-04-04 at 5fa6274)

651 - + t/lib-httpd: pass through GIT_CONFIG_NOSYSTEM env

630 + The code to list branches in "git branch" has been consolidated

631 + with the more generic ref-filter API.

632

653 - Originally merged to 'next' on 2016-03-23

633 + Will be rerolled.

634

655 - The tests that involve running httpd leaked the system-wide

656 - configuration in /etc/gitconfig to the tested environment.

635

658 - Will merge to 'master' after 2.8 final.

636 +* oa/doc-diff-check (2016-03-29) 1 commit

637 + (merged to 'next' on 2016-04-06 at e3d6e8d)

638 + + Documentation: git diff --check detects conflict markers

639

640 + A minor documentation update.

641

661 -* lt/pretty-expand-tabs (2016-03-30) 3 commits

662 - - pretty: allow tweaking tabwidth in --expand-tabs

663 - - pretty: enable --expand-tabs by default for selected pretty formats

664 - - pretty: expand tabs in indented logs to make things line up properly

642 + Will merge to 'master'.

643

666 - Needs tests.

644

645 +* rz/worktree-no-checkout (2016-03-29) 1 commit

646 + (merged to 'next' on 2016-04-06 at e725216)

647 + + worktree: add: introduce --checkout option

648

669 -* sb/clone-shallow-passthru (2016-03-23) 3 commits

670 - - clone: add t5614 to test cloning submodules with shallowness involved

671 - - submodule clone: pass along `local` option

672 - - clone: add `--shallow-submodules` flag

673 - (this branch uses sb/submodule-parallel-update; is tangled with sb/submodule-init.)

649 + "git worktree add" can be given "--no-checkout" option to only

650 + create an empty worktree without checking out the files.

651

675 - "git clone" learned "--shallow-submodules" option.

652 + Will merge to 'master'.

653

677 - Needs review.

654

655 +* sb/misc-cleanups (2016-04-01) 4 commits

656 + (merged to 'next' on 2016-04-06 at 4e63691)

657 + + credential-cache, send_request: close fd when done

658 + + bundle: don't leak an fd in case of early return

659 + + abbrev_sha1_in_line: don't leak memory

660 + + notes: don't leak memory in git_config_get_notes_strategy

661

680 -* sb/clone-t57-t56 (2016-03-16) 1 commit

681 - (merged to 'next' on 2016-04-04 at 5c20247)

682 - + clone tests: rename t57* => t56*

662 + Assorted minor clean-ups.

663

684 - Originally merged to 'next' on 2016-03-23

664 + Will merge to 'master'.

665

686 - Rename bunch of tests on "git clone" for better organization.

666

688 - Will merge to 'master' after 2.8 final.

667 +* sb/submodule-helper-clone-regression-fix (2016-04-01) 6 commits

668 + - submodule--helper, module_clone: catch fprintf failure

669 + - submodule--helper: do not borrow absolute_path() result for too long

670 + - submodule--helper, module_clone: always operate on absolute paths

671 + - submodule--helper clone: create the submodule path just once

672 + - submodule--helper: fix potential NULL-dereference

673 + - recursive submodules: test for relative paths

674

675 + A partial rewrite of "git submodule" in the 2.7 timeframe changed

676 + the way the gitdir: pointer in the submodules point at the real

677 + repository location to use absolute paths by accident. This has

678 + been corrected.

679

691 -* sb/rebase-x (2016-03-18) 2 commits

692 - (merged to 'next' on 2016-04-04 at feda620)

693 - + t3404: cleanup double empty lines between tests

694 - + rebase: decouple --exec from --interactive

680 + Any further comments? Otherwise will merge to 'next'.

681

696 - Originally merged to 'next' on 2016-03-23

682

698 - "git rebase -x" can be used without passing "-i" option.

683 +* sb/submodule-path-misc-bugs (2016-03-30) 6 commits

684 + - t7407: make expectation as clear as possible

685 + - submodule update: test recursive path reporting from subdirectory

686 + - submodule update: align reporting path for custom command execution

687 + - submodule status: correct path handling in recursive submodules

688 + - submodule update --init: correct path handling in recursive submodules

689 + - submodule foreach: correct path display in recursive submodules

690

700 - Will merge to 'master' after 2.8 final.

691 + "git submodule" reports the paths of submodules the command

692 + recurses into, but this was incorrect when the command was not run

693 + from the root level of the superproject.

694

695 + Any further comments? Otherwise will merge to 'next'.

696

703 -* cc/apply (2016-04-01) 4 commits

704 - - builtin/apply: free patch when parse_chunk() fails

705 - - builtin/apply: handle parse_binary() failure

706 - - apply: remove unused call to free() in gitdiff_{old,new}name()

707 - - builtin/apply: get rid of useless 'name' variable

697

709 - Minor code clean-up.

698 +* sg/diff-multiple-identical-renames (2016-03-30) 1 commit

699 + (merged to 'next' on 2016-04-06 at ac19e48)

700 + + diffcore: fix iteration order of identical files during rename detection

701

711 - Will merge to 'next'.

702 + "git diff -M" used to work better when two originally identical

703 + files A and B got renamed to X/A and X/B by pairing A to X/A and B

704 + to X/B, but this was broken in the 2.0 timeframe.

705

706 + Will merge to 'master'.

707

714 -* dt/index-helper (2016-03-23) 18 commits

715 - - SQUASH - minimum compilation fix

716 - - read-cache: config for waiting for index-helper

717 - - index-helper: optionally automatically run

718 - - index-helper: autorun mode

719 - - index-helper: don't run if already running

720 - - index-helper: kill mode

721 - - unpack-trees: preserve index extensions

722 - - update-index: enable/disable watchman support

723 - - index-helper: use watchman to avoid refreshing index with lstat()

724 - - Add watchman support to reduce index refresh cost

725 - - read-cache: invalidate untracked cache data when reading WAMA

726 - - read-cache: add watchman 'WAMA' extension

727 - - index-helper: add --detach

728 - - daemonize(): set a flag before exiting the main process

729 - - index-helper: add --strict

730 - - index-helper: new daemon for caching index and related stuff

731 - - read-cache: allow to keep mmap'd memory after reading

732 - - read-cache.c: fix constness of verify_hdr()

708

709 +* sk/send-pack-all-fix (2016-03-31) 1 commit

710 + (merged to 'next' on 2016-04-06 at 31e1e1b)

711 + + git-send-pack: fix --all option when used with directory

712

735 -* jv/merge-nothing-into-void (2016-03-23) 1 commit

736 - (merged to 'next' on 2016-04-04 at aa37405)

737 - + merge: fix NULL pointer dereference when merging nothing into void

713 + "git send-pack --all <there>" was broken when its command line

714 + option parsing was written in the 2.6 timeframe.

715

739 - Originally merged to 'next' on 2016-03-23

716 + Will merge to 'master'.

717

741 - "git merge FETCH_HEAD" dereferenced NULL pointer when merging

742 - nothing into an unborn history (which is arguably unusual usage,

743 - which perhaps was the reason why nobody noticed it).

718

745 - Will merge to 'master' after 2.8 final.

719 +* ss/msvc (2016-03-30) 2 commits

720 + (merged to 'next' on 2016-04-06 at 4b53bce)

721 + + MSVC: use shipped headers instead of fallback definitions

722 + + MSVC: vsnprintf in Visual Studio 2015 doesn't need SNPRINTF_SIZE_CORR any more

723

724 + Build updates for MSVC.

725

748 -* la/tag-force-signing-annotated-tags (2016-03-22) 1 commit

749 - (merged to 'next' on 2016-04-04 at a49ec4a)

750 - + tag: add the option to force signing of annotated tags

726 + Will merge to 'master'.

727

752 - Originally merged to 'next' on 2016-03-24

728

754 - "git tag" can create an annotated tag without explicitly given an

755 - "-a" (or "-s") option (i.e. when a tag message is given). A new

756 - configuration variable, tag.forceSignAnnotated, can be used to tell

757 - the command to create signed tag in such a situation.

729 +* xy/format-patch-base (2016-03-31) 4 commits

730 + - format-patch: introduce format.base configuration

731 + - format-patch: introduce --base=auto option

732 + - format-patch: add '--base' option to record base tree info

733 + - patch-ids: make commit_patch_id() a public helper function

734

759 - Will merge to 'master' after 2.8 final.

735 + "git format-patch" learned a new "--base" option to record what

736 + (public, well-known) commit the original series was built on in

737 + its output.

738

739 + Will be rerolled.

740 + ($gmane/290365)

741

762 -* cc/doc-recommend-performance-trace-to-file (2016-03-07) 1 commit

763 - (merged to 'next' on 2016-04-04 at 26f94c0)

764 - + Documentation: talk about pager in api-trace.txt

742

766 - Originally merged to 'next' on 2016-03-23

743 +* tb/safe-crlf-output-fix (2016-04-01) 7 commits

744 + . convert.c: more safer crlf handling with text attribute

745 + . correct blame for files commited with CRLF

746 + . convert: unify the "auto" handling of CRLF

747 + . t0027: test cases for combined attributes

748 + . convert: allow core.autocrlf=input and core.eol=crlf

749 + . convert.c: stream and early out

750 + . read-cache: factor out get_sha1_from_index() helper

751

768 - A minor documentation update.

752 + The "safe CRLF" facility disables line-end conversion from CRLF to

753 + LF when checking in if the blob registered to the index already

754 + contains CR, but some codepaths like "git blame" did not know this,

755 + and instead assumed that only the configuration and attribute

756 + settings determined how the data from the working tree is converted.

757

770 - Will merge to 'master' after 2.8 final.

758 + Will be rerolled.

759 + ($gmane/290637)

760

761

773 -* da/mergetool-delete-delete-conflict (2016-03-10) 2 commits

774 - (merged to 'next' on 2016-04-04 at 34e645f)

775 - + mergetool: honor tempfile configuration when resolving delete conflicts

776 - + mergetool: support delete/delete conflicts

762 +* ak/use-hashmap-iter-first-in-submodule-config (2016-03-23) 1 commit

763 + (merged to 'next' on 2016-04-06 at 2aab890)

764 + + submodule-config: use hashmap_iter_first()

765

778 - Originally merged to 'next' on 2016-03-15

766 + Minor code cleanup.

767

780 - "git mergetool" did not work well with conflicts that both sides

781 - deleted.

768 + Will merge to 'master'.

769

783 - Will merge to 'master' after 2.8 final.

770

771 +* ky/branch-d-worktree (2016-03-29) 1 commit

772 + (merged to 'next' on 2016-04-06 at 00f9bff)

773 + + branch -d: refuse deleting a branch which is currently checked out

774

786 -* sg/completion-updates (2016-02-28) 21 commits

787 - - completion: cache the path to the repository

788 - - completion: extract repository discovery from __gitdir()

789 - - completion: don't guard git executions with __gitdir()

790 - - completion: consolidate silencing errors from git commands

791 - - completion: don't use __gitdir() for git commands

792 - - completion: respect 'git -C <path>'

793 - - completion: fix completion after 'git -C <path>'

794 - - completion: don't offer commands when 'git --opt' needs an argument

795 - - rev-parse: add '--absolute-git-dir' option

796 - - completion: list short refs from a remote given as a URL

797 - - completion: don't list 'HEAD' when trying refs completion outside of a repo

798 - - completion: list refs from remote when remote's name matches a directory

799 - - completion: respect 'git --git-dir=<path>' when listing remote refs

800 - - completion: fix most spots not respecting 'git --git-dir=<path>'

801 - - completion: ensure that the repository path given on the command line exists

802 - - completion tests: add tests for the __git_refs() helper function

803 - - completion tests: check __gitdir()'s output in the error cases

804 - - completion tests: consolidate getting path of current working directory

805 - - completion tests: make the $cur variable local to the test helper functions

806 - - completion tests: don't add test cruft to the test repository

807 - - completion: improve __git_refs()'s in-code documentation

775 + When "git worktree" feature is in use, "git branch -d" allowed

776 + deletion of a branch that is checked out in another worktree

777

809 - Will be rerolled.

810 - ($gmane/287839)

778 + Will merge to 'master'.

779

780

813 -* ss/exc-flag-is-a-collection-of-bits (2016-03-01) 1 commit

814 - (merged to 'next' on 2016-04-04 at 9f0207e)

815 - + dir: store EXC_FLAG_* values in unsigned integers

781 +* ky/branch-m-worktree (2016-04-04) 2 commits

782 + (merged to 'next' on 2016-04-06 at e7b285c)

783 + + branch -m: update all per-worktree HEADs

784 + + refs: add a new function set_worktree_head_symref

785

817 - Originally merged to 'next' on 2016-03-04

786 + When "git worktree" feature is in use, "git branch -m" renamed a

787 + branch that is checked out in another worktree without adjusting

788 + the HEAD symbolic ref for the worktree.

789

819 - Code clean-up.

790 + Will merge to 'master'.

791

821 - Will merge to 'master' after 2.8 final.

792

793 +* nd/apply-doc (2016-03-24) 2 commits

794 + (merged to 'next' on 2016-04-06 at f9bd355)

795 + + git-apply.txt: mention the behavior inside a subdir

796 + + git-apply.txt: remove a space

797

824 -* jc/bundle (2016-03-03) 6 commits

825 - - index-pack: --clone-bundle option

826 - - Merge branch 'jc/index-pack' into jc/bundle

827 - - bundle v3: the beginning

828 - - bundle: keep a copy of bundle file name in the in-core bundle header

829 - - bundle: plug resource leak

830 - - bundle doc: 'verify' is not about verifying the bundle

798 + A minor documentation update.

799

832 - The beginning of "split bundle", which could be one of the

833 - ingredients to allow "git clone" traffic off of the core server

834 - network to CDN.

800 + Will merge to 'master'.

801

802

837 -* ss/receive-pack-parse-options (2016-03-01) 1 commit

838 - (merged to 'next' on 2016-04-04 at fd6ab4c)

839 - + builtin/receive-pack.c: use parse_options API

803 +* nd/apply-report-skip (2016-03-24) 1 commit

804 + (merged to 'next' on 2016-04-06 at ae2c824)

805 + + apply: report patch skipping in verbose mode

806

841 - Originally merged to 'next' on 2016-03-04

807 + "git apply -v" learned to report paths in the patch that were

808 + skipped via --include/--exclude mechanism or being outside the

809 + current working directory.

810

843 - The command line argument parser for "receive-pack" has been

844 - rewritten to use parse-options.

811 + Will merge to 'master'.

812

846 - Will merge to 'master' after 2.8 final.

813

814 +* pb/opt-cmdmode-doc (2016-03-25) 1 commit

815 + (merged to 'next' on 2016-04-06 at a5f3835)

816 + + api-parse-options.txt: document OPT_CMDMODE()

817

849 -* jk/submodule-c-credential (2016-03-23) 7 commits

850 - (merged to 'next' on 2016-04-04 at 8de8e8c)

851 - + git_config_push_parameter: handle empty GIT_CONFIG_PARAMETERS

852 - + git: submodule honor -c credential.* from command line

853 - + quote: implement sq_quotef()

854 - + submodule: fix segmentation fault in submodule--helper clone

855 - + submodule: fix submodule--helper clone usage

856 - + submodule: check argc count for git submodule--helper clone

857 - + submodule: don't pass empty string arguments to submodule--helper clone

818 + Minor API documentation update.

819

859 - Originally merged to 'next' on 2016-03-23

820 + Will merge to 'master'.

821

861 - "git -c credential.<var>=<value> submodule" can now be used to

862 - propagate configuration variables related to credential helper

863 - down to the submodules.

822

865 - Will merge to 'master' after 2.8 final.

823 +* rt/completion-help (2016-03-24) 2 commits

824 + (merged to 'next' on 2016-04-06 at 8c3ee08)

825 + + completion: add 'revisions' and 'everyday' to 'git help'

826 + + completion: add option '--guides' to 'git help'

827

828 + Shell completion (in contrib/) updates.

829

868 -* nd/shallow-deepen (2016-02-23) 25 commits

869 - - fetch, upload-pack: --deepen=N extends shallow boundary by N commits

870 - - upload-pack: add get_reachable_list()

871 - - upload-pack: split check_unreachable() in two, prep for get_reachable_list()

872 - - t5500, t5539: tests for shallow depth excluding a ref

873 - - clone: define shallow clone boundary with --shallow-exclude

874 - - fetch: define shallow boundary with --shallow-exclude

875 - - upload-pack: support define shallow boundary by excluding revisions

876 - - refs: add expand_ref()

877 - - t5500, t5539: tests for shallow depth since a specific date

878 - - clone: define shallow clone boundary based on time with --shallow-since

879 - - fetch: define shallow boundary with --shallow-since

880 - - upload-pack: add deepen-since to cut shallow repos based on time

881 - - shallow.c: implement a generic shallow boundary finder based on rev-list

882 - - fetch-pack: use a separate flag for fetch in deepening mode

883 - - fetch-pack: use a common function for verbose printing

884 - - fetch-pack: use skip_prefix() instead of starts_with()

885 - - upload-pack: move rev-list code out of check_non_tip()

886 - - upload-pack: tighten number parsing at "deepen" lines

887 - - upload-pack: use skip_prefix() instead of starts_with()

888 - - upload-pack: move "unshallow" sending code out of deepen()

889 - - upload-pack: remove unused variable "backup"

890 - - upload-pack: move "shallow" sending code out of deepen()

891 - - upload-pack: move shallow deepen code out of receive_needs()

892 - - transport-helper.c: refactor set_helper_option()

893 - - remote-curl.c: convert fetch_git() to use argv_array

830 + Will merge to 'master'.

831

895 - The existing "git fetch --depth=<n>" option was hard to use

896 - correctly when making the history of an existing shallow clone

897 - deeper. A new option, "--deepen=<n>", has been added to make this

898 - easier to use. "git clone" also learned "--shallow-since=<date>"

899 - and "--shallow-exclude=<tag>" options to make it easier to specify

900 - "I am interested only in the recent N months worth of history" and

901 - "Give me only the history since that version".

832

903 - Needs review.

833 +* rt/rebase-i-shorten-stop-report (2016-03-28) 1 commit

834 + (merged to 'next' on 2016-04-06 at 7a766b7)

835 + + rebase-i: print an abbreviated hash when stop for editing

836

837 + The commit object name reported when "rebase -i" stops has been

838 + shortened.

839

906 -* sb/submodule-init (2016-03-15) 2 commits

907 - . submodule: port init from shell to C

908 - . submodule: port resolve_relative_url from shell to C

909 - (this branch uses sb/submodule-parallel-update; is tangled with sb/clone-shallow-passthru.)

840 + Will merge to 'master'.

841

911 - Update of "git submodule" to move pieces of logic to C continues.

842

913 - Needs review.

914 - ($gmane/288824)

843 +* jk/check-repository-format (2016-03-11) 10 commits

844 + (merged to 'next' on 2016-04-06 at a0dada0)

845 + + verify_repository_format: mark messages for translation

846 + + setup: drop repository_format_version global

847 + + setup: unify repository version callbacks

848 + + init: use setup.c's repo version verification

849 + + setup: refactor repo format reading and verification

850 + + config: drop git_config_early

851 + + check_repository_format_gently: stop using git_config_early

852 + + lazily load core.sharedrepository

853 + + wrap shared_repository global in get/set accessors

854 + + setup: document check_repository_format()

855

856 + The repository set-up sequence has been streamlined (the biggest

857 + change is that there is no longer git_config_early()), so that we

858 + do not attempt to look into refs/* when we know we do not have a

859 + Git repository.

860

917 -* az/p4-bare-no-rebase (2016-02-19) 1 commit

918 - - git-p4.py: Don't try to rebase on submit from bare repository

861 + Will merge to 'master'.

862

920 - "git p4 submit" attempts to do a rebase, which would fail if done

921 - in a bare repository. Not doing this rebase would paper over the

922 - failure, which is what this patch does, but it is unclear what the

923 - side effect of not rebasing is.

863

925 - Needs a better explanation.

864 +* mj/pull-rebase-autostash (2016-04-04) 9 commits

865 + (merged to 'next' on 2016-04-06 at b4e4f31)

866 + + t5520: test --[no-]autostash with pull.rebase=true

867 + + t5520: reduce commom lines of code

868 + + t5520: factor out common "failing autostash" code

869 + + t5520: factor out common "successful autostash" code

870 + + t5520: use better test to check stderr output

871 + + t5520: ensure consistent test conditions

872 + + t5520: use consistent capitalization in test titles

873 + + pull --rebase: add --[no-]autostash flag

874 + + git-pull.c: introduce git_pull_config()

875

876 + "git pull --rebase" learned "--[no-]autostash" option, so that

877 + the rebase.autostash configuration variable set to true can be

878 + overridden from the command line.

879

928 -* ss/commit-dry-run-resolve-merge-to-no-op (2016-02-17) 1 commit

929 - - wt-status.c: set commitable bit if there is a meaningful merge.

880 + Will merge to 'master'.

881

931 - "git commit --dry-run" reported "No, no, you cannot commit." in one

932 - case where "git commit" would have allowed you to commit, and this

933 - improves it a little bit ("git commit --dry-run --short" still does

934 - not give you the correct answer, for example).

882

883 +* pb/commit-verbose-config (2016-03-14) 1 commit

884 + (merged to 'next' on 2016-04-06 at e5c744f)

885 + + commit: add a commit.verbose config variable

886

937 -* nd/icase (2016-02-15) 12 commits

938 - - grep.c: reuse "icase" variable

939 - - diffcore-pickaxe: support case insensitive match on non-ascii

940 - - diffcore-pickaxe: "share" regex error handling code

941 - - grep/pcre: support utf-8

942 - - gettext: add is_utf8_locale()

943 - - grep/pcre: prepare locale-dependent tables for icase matching

944 - - grep/icase: avoid kwsset when -F is specified

945 - - grep/icase: avoid kwsset on literal non-ascii strings

946 - - test-regex: expose full regcomp() to the command line

947 - - test-regex: isolate the bug test code

948 - - grep: break down an "if" stmt in preparation for next changes

949 - - grep: allow -F -i combination

887 + "git commit" learned to pay attention to "commit.verbose"

888 + configuration variable and act as if "--verbose" option was

889 + given from the command line.

890

951 - "git grep -i" has been taught to fold case in non-ascii locales.

891 + Will merge to 'master'.

892

953 - Needs review.

954 - ($gmane/286137)

893

894 +* jc/merge-refuse-new-root (2016-03-23) 1 commit

895 + (merged to 'next' on 2016-04-04 at cd70fd6)

896 + + merge: refuse to create too cool a merge by default

897

957 -* sb/submodule-parallel-update (2016-03-01) 10 commits

958 - (merged to 'next' on 2016-04-04 at a0aea8d)

959 - + clone: allow an explicit argument for parallel submodule clones

960 - + submodule update: expose parallelism to the user

961 - + submodule helper: remove double 'fatal: ' prefix

962 - + git submodule update: have a dedicated helper for cloning

963 - + run_processes_parallel: rename parameters for the callbacks

964 - + run_processes_parallel: treat output of children as byte array

965 - + submodule update: direct error message to stderr

966 - + fetching submodules: respect `submodule.fetchJobs` config option

967 - + submodule-config: drop check against NULL

968 - + submodule-config: keep update strategy around

969 - (this branch is used by sb/clone-shallow-passthru and sb/submodule-init.)

898 + Originally merged to 'next' on 2016-03-23

899

971 - Originally merged to 'next' on 2016-03-15

900 + "git merge" used to allow merging two branches that have no common

901 + base by default, which led to a brand new history of an existing

902 + project created and then get pulled by an unsuspecting maintainer,

903 + which allowed an unnecessary parallel history merged into the

904 + existing project. The command has been taught not to allow this by

905 + default, with an escape hatch "--allow-unrelated-histories" option

906 + to be used in a rare event that merges histories of two projects

907 + that started their lives independently.

908

973 - A major part of "git submodule update" has been ported to C to take

974 - advantage of the recently added framework to run download tasks in

975 - parallel.

909 + Will merge to 'master'.

910

977 - Will merge to 'master' after 2.8 final.

911

912 +* lt/pretty-expand-tabs (2016-04-04) 4 commits

913 + (merged to 'next' on 2016-04-06 at 186ac2a)

914 + + pretty: test --expand-tabs

915 + + pretty: allow tweaking tabwidth in --expand-tabs

916 + + pretty: enable --expand-tabs by default for selected pretty formats

917 + + pretty: expand tabs in indented logs to make things line up properly

918

980 -* dt/refs-backend-lmdb (2016-02-25) 45 commits

981 - . SQUASH??? Minimum compilation band-aid

982 - . tests: add ref-storage argument

983 - . refs: tests for lmdb backend

984 - . refs: add LMDB refs storage backend

985 - . refs: break out resolve_ref_unsafe_submodule

986 - . config: read ref storage config on startup

987 - . refs: register ref storage backends

988 - . svn: learn ref-storage argument

989 - . clone: allow ref storage backend to be set for clone

990 - . refs: check submodules' ref storage config

991 - . init: allow alternate ref strorage to be set for new repos

992 - . refs: always handle non-normal refs in files backend

993 - . refs: resolve symbolic refs first

994 - . refs: on symref reflog expire, lock symref not referrent

995 - . refs: don't dereference on rename

996 - . refs: allow log-only updates

997 - . refs: move duplicate check to common code

998 - . refs: make lock generic

999 - . refs: handle non-normal ref renames

1000 - . refs: add method to rename refs

1001 - . refs: add methods to init refs db

1002 - . refs: add method for delete_refs

1003 - . refs: add method for initial ref transaction commit

1004 - . refs: add methods for reflog

1005 - . refs: add do_for_each_per_worktree_ref

1006 - . refs: reduce the visibility of do_for_each_ref()

1007 - . refs: add method for do_for_each_ref

1008 - . refs: add methods for misc ref operations

1009 - . refs: add a backend method structure with transaction functions

1010 - . refs: move resolve_ref_unsafe into common code

1011 - . files-backend: break out ref reading

1012 - . refs: move for_each_*ref* functions into common code

1013 - . refs: move head_ref{,_submodule} to the common code

1014 - . Merge branch 'sb/submodule-parallel-update' into dt/refs-backend-lmdb

1015 - . clone: allow an explicit argument for parallel submodule clones

1016 - . submodule update: expose parallelism to the user

1017 - . git submodule update: have a dedicated helper for cloning

1018 - . run_processes_parallel: correctly terminate callbacks with an LF

1019 - . run_processes_parallel: rename parameters for the callbacks

1020 - . run-command: expose default_{start_failure, task_finished}

1021 - . run_processes_parallel: treat output of children as byte array

1022 - . submodule update: direct error message to stderr

1023 - . fetching submodules: respect `submodule.fetchJobs` config option

1024 - . submodule-config: drop check against NULL

1025 - . submodule-config: keep update strategy around

919 + When "git log" shows the log message indented by 4-spaces, the

920 + remainder of a line after a HT does not align in the way the author

921 + originally intended. The command now expands tabs by default in

922 + such a case, and allows the users to override it with a new option,

923 + '--no-expand-tabs'.

924

1027 - A reroll exists, but it seems that will further be rerolled.

925 + Will merge to 'master'.

926 +

927 +

928 +* cc/apply (2016-04-01) 4 commits

929 + (merged to 'next' on 2016-04-06 at 2e23c44)

930 + + builtin/apply: free patch when parse_chunk() fails

931 + + builtin/apply: handle parse_binary() failure

932 + + apply: remove unused call to free() in gitdiff_{old,new}name()

933 + + builtin/apply: get rid of useless 'name' variable

934 +

935 + Minor code clean-up.

936 +

937 + Will merge to 'master'.

938 +

939 +

940 +* dt/index-helper (2016-03-23) 18 commits

941 + - SQUASH - minimum compilation fix

942 + - read-cache: config for waiting for index-helper

943 + - index-helper: optionally automatically run

944 + - index-helper: autorun mode

945 + - index-helper: don't run if already running

946 + - index-helper: kill mode

947 + - unpack-trees: preserve index extensions

948 + - update-index: enable/disable watchman support

949 + - index-helper: use watchman to avoid refreshing index with lstat()

950 + - Add watchman support to reduce index refresh cost

951 + - read-cache: invalidate untracked cache data when reading WAMA

952 + - read-cache: add watchman 'WAMA' extension

953 + - index-helper: add --detach

954 + - daemonize(): set a flag before exiting the main process

955 + - index-helper: add --strict

956 + - index-helper: new daemon for caching index and related stuff

957 + - read-cache: allow to keep mmap'd memory after reading

958 + - read-cache.c: fix constness of verify_hdr()

959 +

960 + Needs review.

961 +

962 +

963 +* jc/bundle (2016-03-03) 6 commits

964 + - index-pack: --clone-bundle option

965 + - Merge branch 'jc/index-pack' into jc/bundle

966 + - bundle v3: the beginning

967 + - bundle: keep a copy of bundle file name in the in-core bundle header

968 + - bundle: plug resource leak

969 + - bundle doc: 'verify' is not about verifying the bundle

970 +

971 + The beginning of "split bundle", which could be one of the

972 + ingredients to allow "git clone" traffic off of the core server

973 + network to CDN.

974

975

976 * ad/cygwin-wants-rename (2015-08-07) 1 commit

@@ -1034,7 +980,7 @@ of the repositories listed at

980 ($gmane/275680).

981

982

1037 -* jc/rerere-multi (2016-03-28) 11 commits

983 +* jc/rerere-multi (2016-04-06) 11 commits

984 - rerere: adjust 'forget' to multi-variant world order

985 - rerere: split code to call ll_merge() further

986 - rerere: move code related to "forget" together

@@ -1051,14 +997,16 @@ of the repositories listed at

997 signature that have to be resolved in different ways, but there was

998 no way to record these separate resolutions.

999

1054 - Need to send out the final round of review as this should be now complete.

1000 + Will merge to 'next'.

1001

1002

1003 * jc/merge-drop-old-syntax (2015-04-29) 1 commit

1004 - merge: drop 'git merge <message> HEAD <commit>' syntax

1005

1006 Stop supporting "git merge <message> HEAD <commit>" syntax that has

1061 - been deprecated since October 2007. It has been reported that

1062 - git-gui still uses the deprecated syntax, which needs to be fixed

1063 - before this final step can proceed.

1007 + been deprecated since October 2007, and issues a deprecation

1008 + warning message since v2.5.0.

1009 +

1010 + It has been reported that git-gui still uses the deprecated syntax,

1011 + which needs to be fixed before this final step can proceed.

1012 ($gmane/282594)

What's cooking (2016/04 #03)