7e8a7f23fe · 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 (Jul 2016, #02; Wed, 6)

3 +Subject: What's cooking in git.git (Jul 2016, #03; Fri, 8)

4 X-master-at: 5c589a73de4394ad125a4effac227b3aec856fa1

5 X-next-at: 281e4d64a61b891e32e0344654b580b4203effa6

6

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

7 +What's cooking in git.git (Jul 2016, #03; Fri, 8)

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

9

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

@@ -21,325 +21,100 @@ of the repositories listed at

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

22

23 --------------------------------------------------

24 -[Graduated to "master"]

25 -

26 -* ak/t7800-wo-readlink (2016-06-21) 1 commit

27 - (merged to 'next' on 2016-06-28 at 403bbbb)

28 - + t7800: readlink may not be available

29 -

30 - One among four invocations of readlink(1) in our test suite has

31 - been rewritten so that the test can run on systems without the

32 - command (others are in valgrind test framework and t9802).

33 -

34 -

35 -* ao/p4-has-branch-prefix-fix (2016-06-22) 1 commit

36 - (merged to 'next' on 2016-06-28 at 333f90d)

37 - + git-p4: correct hasBranchPrefix verbose output

38 -

39 - A bug, which caused "git p4" while running under verbose mode to

40 - report paths that are omitted due to branch prefix incorrectly, has

41 - been fixed; the command said "Ignoring file outside of prefix" for

42 - paths that are _inside_.

43 -

44 -

45 -* cb/t7810-test-label-fix (2016-06-21) 1 commit

46 - (merged to 'next' on 2016-06-28 at 4d85a00)

47 - + t7810: fix duplicated test title

48 -

49 - Test clean-up.

50 -

51 -

52 -* ep/http-curl-trace (2016-05-24) 2 commits

53 - (merged to 'next' on 2016-06-27 at c290515)

54 - + imap-send.c: introduce the GIT_TRACE_CURL enviroment variable

55 - + http.c: implement the GIT_TRACE_CURL environment variable

56 -

57 - HTTP transport gained an option to produce more detailed debugging

58 - trace.

59 -

60 -

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

62 - (merged to 'next' on 2016-06-28 at ee0a088)

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

64 - + mailsplit: support unescaping mboxrd messages

65 - + pretty: support "mboxrd" output format

66 -

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

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

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

70 -

71 -

72 -* jc/t2300-setup (2016-06-22) 1 commit

73 - (merged to 'next' on 2016-06-28 at 62b902a)

74 - + t2300: "git --exec-path" is not usable in $PATH on Windows as-is

75 -

76 - Portability fix for Windows.

77 -

78 -

79 -* jk/add-i-diff-compact-heuristics (2016-06-16) 1 commit

80 - (merged to 'next' on 2016-06-27 at 568f892)

81 - + add--interactive: respect diff.compactionHeuristic

82 -

83 - "git add -i/-p" learned to honor diff.compactionHeuristic

84 - experimental knob, so that the user can work on the same hunk split

85 - as "git diff" output.

86 -

87 -

88 -* jk/bisect-show-tree (2016-06-16) 1 commit

89 - (merged to 'next' on 2016-06-27 at 6970f87e)

90 - + bisect: always call setup_revisions after init_revisions

91 -

92 - "git bisect" makes an internal call to "git diff-tree" when

93 - bisection finds the culprit, but this call did not initialize the

94 - data structure to pass to the diff-tree API correctly.

95 -

96 -

97 -* jk/gpg-interface-cleanup (2016-06-17) 7 commits

98 - (merged to 'next' on 2016-06-28 at eed3498)

99 - + gpg-interface: check gpg signature creation status

100 - + sign_buffer: use pipe_command

101 - + verify_signed_buffer: use pipe_command

102 - + run-command: add pipe_command helper

103 - + verify_signed_buffer: use tempfile object

104 - + verify_signed_buffer: drop pbuf variable

105 - + gpg-interface: use child_process.args

106 -

107 - A new run-command API function pipe_command() is introduced to

108 - sanely feed data to the standard input while capturing data from

109 - the standard output and the standard error of an external process,

110 - which is cumbersome to hand-roll correctly without deadlocking.

111 -

112 - The codepath to sign data in a prepared buffer with GPG has been

113 - updated to use this API to read from the status-fd to check for

114 - errors (instead of relying on GPG's exit status).

115 -

116 -

117 -* jk/parseopt-string-list (2016-06-13) 3 commits

118 - (merged to 'next' on 2016-06-27 at 27462e6)

119 - + blame,shortlog: don't make local option variables static

120 - + interpret-trailers: don't duplicate option strings

121 - + parse_opt_string_list: stop allocating new strings

122 - (this branch is used by jk/string-list-static-init.)

123 -

124 - The command line argument parsing that uses OPT_STRING_LIST() often

125 - made a copy of the argv[] element, which was unnecessary.

126 -

127 -

128 -* jk/repack-keep-unreachable (2016-06-14) 3 commits

129 - (merged to 'next' on 2016-06-28 at 802b849)

130 - + repack: extend --keep-unreachable to loose objects

131 - + repack: add --keep-unreachable option

132 - + repack: document --unpack-unreachable option

133 -

134 - "git repack" learned the "--keep-unreachable" option, which sends

135 - loose unreachable objects to a pack instead of leaving them loose.

136 - This helps heuristics based on the number of loose objects

137 - (e.g. "gc --auto").

138 -

139 -

140 -* jk/send-pack-stdio (2016-06-10) 2 commits

141 - (merged to 'next' on 2016-06-27 at 2cfb0ff)

142 - + write_or_die: remove the unused write_or_whine() function

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

144 -

145 - Code clean-up.

146 -

147 -

148 -* jk/string-list-static-init (2016-06-13) 2 commits

149 - (merged to 'next' on 2016-06-27 at 3d4b2fa)

150 - + use string_list initializer consistently

151 - + Merge branch 'jk/parseopt-string-list' into jk/string-list-static-init

152 - (this branch uses jk/parseopt-string-list.)

153 -

154 - Instead of taking advantage of a struct string_list that is

155 - allocated with all NULs happens to be STRING_LIST_INIT_NODUP kind,

156 - initialize them explicitly as such, to document their behaviour

157 - better.

158 -

159 -

160 -* jk/tzoffset-fix (2016-06-20) 3 commits

161 - (merged to 'next' on 2016-06-28 at 08ec8c5)

162 - + local_tzoffset: detect errors from tm_to_time_t

163 - + t0006: test various date formats

164 - + t0006: rename test-date's "show" to "relative"

165 -

166 - The internal code used to show local timezone offset is not

167 - prepared to handle timestamps beyond year 2100, and gave a

168 - bogus offset value to the caller. Use a more benign looking

169 - +0000 instead and let "git log" going in such a case, instead

170 - of aborting.

171 -

172 -

173 -* jk/upload-pack-hook (2016-06-02) 7 commits

174 - (merged to 'next' on 2016-06-28 at e5e5bb6)

175 - + upload-pack: provide a hook for running pack-objects

176 - + t1308: do not get fooled by symbolic links to the source tree

177 - + config: add a notion of "scope"

178 - + config: return configset value for current_config_ functions

179 - + config: set up config_source for command-line config

180 - + git_config_parse_parameter: refactor cleanup code

181 - + git_config_with_options: drop "found" counting

182 -

183 - "upload-pack" allows a custom "git pack-objects" replacement when

184 - responding to "fetch/clone" via the uploadpack.packObjectsHook.

185 -

186 -

187 -* js/mingw-parameter-less-c-functions (2016-06-20) 1 commit

188 - (merged to 'next' on 2016-06-28 at e673c65)

189 - + mingw: let the build succeed with DEVELOPER=1

190 -

191 - Some platform-specific code had non-ANSI strict declarations of C

192 - functions that do not take any parameters, which has been

193 - corrected.

194 -

195 -

196 -* js/perf-on-apple (2016-06-21) 1 commit

197 - (merged to 'next' on 2016-06-28 at 83529b9)

198 - + perf: accommodate for MacOSX

199 -

200 - t/perf needs /usr/bin/time with GNU extension; the invocation of it

201 - is updated to "gtime" on Darwin.

202 -

203 -

204 -* km/fetch-do-not-free-remote-name (2016-06-14) 1 commit

205 - (merged to 'next' on 2016-06-27 at 4bc34c4)

206 - + builtin/fetch.c: don't free remote->name after fetch

207 -

208 - The ownership rule for the piece of memory that hold references to

209 - be fetched in "git fetch" was screwy, which has been cleaned up.

210 -

211 -

212 -* lc/shell-default-value-noexpand (2016-06-19) 1 commit

213 - (merged to 'next' on 2016-06-28 at 2dc7727)

214 - + sh-setup: enclose setting of ${VAR=default} in double-quotes

215 -

216 - Fix unnecessarily waste in the idiomatic use of ': ${VAR=default}'

217 - to set the default value, without enclosing it in double quotes.

218 -

219 -

220 -* lf/sideband-returns-void (2016-06-16) 2 commits

221 - (merged to 'next' on 2016-06-27 at 558c781)

222 - + upload-pack.c: make send_client_data() return void

223 - + sideband.c: make send_sideband() return void

224 -

225 - A small internal API cleanup.

226 -

227 -

228 -* mg/signature-doc (2016-06-17) 4 commits

229 - (merged to 'next' on 2016-06-28 at 2732a37)

230 - + Documentation/technical: signed merge tag format

231 - + Documentation/technical: signed commit format

232 - + Documentation/technical: signed tag format

233 - + Documentation/technical: describe signature formats

234 -

235 - Formats of the various data (and how to validate them) where we use

236 - GPG signature have been documented.

237 -

238 -

239 -* nd/graph-width-padded (2016-06-16) 2 commits

240 - (merged to 'next' on 2016-06-28 at 0f733ef)

241 - + pretty.c: support <direction>|(<negative number>) forms

242 - + pretty: pass graph width to pretty formatting for use in '%>|(N)'

243 -

244 - "log --graph --format=" learned that "%>|(N)" specifies the width

245 - relative to the terminal's left edge, not relative to the area to

246 - draw text that is to the right of the ancestry-graph section. It

247 - also now accepts negative N that means the column limit is relative

248 - to the right border.

249 -

24 +[New Topics]

25

251 -* nd/test-lib-httpd-show-error-log-in-verbose (2016-06-13) 1 commit

252 - (merged to 'next' on 2016-06-27 at 9793d81)

253 - + lib-httpd.sh: print error.log on error

26 +* rs/help-c-source-with-gitattributes (2016-07-07) 1 commit

27 + - .gitattributes: set file type for C files

28

255 - HTTPd tests learned to show the server error log to help diagnosing

256 - a failing tests.

29 + The .c/.h sources are marked as such in our .gitattributes file so

30 + that "git diff -W" and friends would work better.

31

32 + Will merge to 'next'.

33

259 -* nd/worktree-cleanup-post-head-protection (2016-05-24) 6 commits

260 - (merged to 'next' on 2016-06-28 at 55414fa)

261 - + worktree: simplify prefixing paths

262 - + worktree: avoid 0{40}, too many zeroes, hard to read

263 - + worktree.c: use is_dot_or_dotdot()

264 - + git-worktree.txt: keep subcommand listing in alphabetical order

265 - + worktree.c: rewrite mark_current_worktree() to avoid strbuf

266 - + completion: support git-worktree

267 - (this branch is used by nd/worktree-lock.)

34

269 - Further preparatory clean-up for "worktree" feature continues.

35 +* rs/notes-merge-no-toctou (2016-07-07) 1 commit

36 + - notes-merge: use O_EXCL to avoid overwriting existing files

37

38 + "git notes merge" had a code to see if a path exists (and fails if

39 + it does) and then open the path for writing (when it doesn't).

40 + Replace it with open with O_EXCL.

41

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

273 - (merged to 'next' on 2016-06-27 at 0f01ce1)

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

42

276 - Code clean-up.

43 +* rw/make-needs-librt (2016-07-07) 1 commit

44 + - Makefile: add NEEDS_LIBRT to optionally link with librt

45

46 + Makefile assumed that -lrt is always available on platforms that

47 + want to use clock_gettime() and CLOCK_MONOTONIC, which is not a

48 + case for recent Mac OS X. The necessary symbols are often found in

49 + libc on many modern systems and having -lrt on the command line, as

50 + long as the library exists, had no effect, but when the platform

51 + removes librt.a that is a different matter--having -lrt will break

52 + the linkage.

53

279 -* sb/clone-shallow-passthru (2016-06-20) 1 commit

280 - (merged to 'next' on 2016-06-28 at d515d7d)

281 - + clone: do not let --depth imply --shallow-submodules

282 - (this branch is used by sb/t5614-modernize.)

54 + This change could be seen as a regression for those who do need to

55 + specify -lrt, as they now specifically ask for NEEDS_LIBRT when

56 + building. Hopefully they are in the minority these days.

57

284 - Fix an unintended regression in v2.9 that breaks "clone --depth"

285 - that recurses down to submodules by forcing the submodules to also

286 - be cloned shallowly, which many server instances that host upstream

287 - of the submodules are not prepared for.

58 + Will merge to 'next'.

59

60

290 -* sb/t5614-modernize (2016-06-21) 1 commit

291 - (merged to 'next' on 2016-06-28 at ee05326)

292 - + t5614: don't use subshells

293 - (this branch uses sb/clone-shallow-passthru.)

61 +* jc/renormalize-merge-kill-safer-crlf (2016-07-08) 2 commits

62 + - merge: avoid "safer crlf" during recording of merge results

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

64 + (this branch is tangled with tb/convert-peek-in-index.)

65

295 - Test clean-up.

66 + "git merge" with renormalization did not work well with

67 + merge-recursive, due to "safer crlf" conversion kicking in when it

68 + shouldn't.

69

297 ---------------------------------------------------

298 -[New Topics]

70

300 -* ak/lazy-prereq-mktemp (2016-07-06) 1 commit

301 - - t7610: test for mktemp before test execution

71 +* jk/printf-format (2016-07-08) 3 commits

72 + - commit.c: remove print_commit_list()

73 + - avoid using sha1_to_hex output as printf format

74 + - walker: let walker_say take arbitrary formats

75

303 - A test that unconditionally used "mktemp" learned that the command

304 - is not necessarily available everywhere.

76 + Code clean-up to avoid using a variable string that compilers may

77 + feel untrustable as printf-style format given to write_file()

78 + helper function.

79

80 Will merge to 'next'.

81

82

309 -* ew/svn-bad-ref (2016-07-06) 1 commit

310 - - git-svn: warn instead of dying when commit data is missing

311 -

312 - An earlier update to "git svn" tightened the parsing of refs given

313 - as the command line arguments too much. This has been fixed by

314 - loosening the check a bit.

315 -

316 - I'm likely to discard this, favouring a direct pull request from

317 - the subsystem maintainer directly going to 'master'.

83 +* jk/write-file (2016-07-08) 9 commits

84 + - branch: use write_file_buf instead of write_file

85 + - use write_file_buf where applicable

86 + - write_file: add format attribute

87 + - write_file: add pointer+len variant

88 + - write_file: use xopen

89 + - write_file: drop "gently" form

90 + - branch: use non-gentle write_file for branch description

91 + - am: ignore return value of write_file()

92 + - config: fix bogus fd check when setting up default config

93

94 + General code clean-up around a helper function to write a

95 + single-liner to a file.

96

320 -* jk/difftool-in-subdir (2016-07-06) 1 commit

321 - - difftool: fix argument handling in subdirs

322 -

323 - "git difftool <paths>..." started in a subdirectory failed to

324 - interpret the paths relative to that directory, which has been

325 - fixed.

326 -

327 - Waiting for an ack.

97 + Will merge to 'next'.

98 + While it is textually independent, jk/printf-format needs to be

99 + merged before this one.

100

101

330 -* nd/test-helpers (2016-07-06) 2 commits

331 - - Makefile: use VCSSVN_LIB to refer to svn library

332 - - Makefile: drop extra dependencies for test helpers

102 +* js/am-call-theirs-theirs-in-fallback-3way (2016-07-08) 1 commit

103 + - am: counteract gender bias

104

334 - Build clean-up.

105 + One part of "git am" had an oddball helper function that called

106 + stuff from outside "his" as opposed to calling what we have "ours",

107 + which was not gender-neutral and also inconsistent with the rest of

108 + the system where outside stuff is usuall called "theirs" in

109 + contrast to "ours".

110

111 Will merge to 'next'.

112

113 --------------------------------------------------

114 [Stalled]

115

341 -* jh/clean-smudge-annex (2016-07-06) 10 commits

342 - - SQUASH???

116 +* jh/clean-smudge-annex (2016-07-08) 9 commits

117 + - SQUASH??? compilation fix

118 - use smudgeToFile filter in recursive merge

119 - use smudgeToFile filter in git am

120 - better recovery from failure of smudgeToFile filter

@@ -348,23 +123,21 @@ of the repositories listed at

123 - use cleanFromFile in git add

124 - add smudgeToFile and cleanFromFile filter configs

125 - clarify %f documentation

351 - - Merge branch 'tb/convert-peek-in-index' into jh/clean-smudge-annex

352 - (this branch uses tb/convert-peek-in-index.)

126

127 The interface to "clean/smudge" filters require Git to feed the

355 - whole contents via pipe, which is suboptimal for some

356 - applications. "cleanFromFile/smudgeToFile" commands are the moral

357 - equilvalents for these filters but they interact with the files on

358 - the filesystem directly.

128 + whole contents via pipe, which is suboptimal for some applications.

129 + "cleanFromFile/smudgeToFile" commands are the moral equilvalents

130 + for these filters but they interact with the files on the

131 + filesystem directly.

132

133

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

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

363 - - correct ce_compare_data() in a middle of a merge

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

135 + . correct ce_compare_data() in a middle of a merge

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

137 - convert: unify the "auto" handling of CRLF

365 - (this branch is used by jh/clean-smudge-annex.)

138 + (this branch is tangled with jc/renormalize-merge-kill-safer-crlf.)

139

367 - Needs review.

140 + Tentatively replaced with jc/renormalize-merge-kill-safer-crlf

141

142

143 * sb/bisect (2016-04-15) 22 commits

@@ -465,6 +238,45 @@ of the repositories listed at

238 --------------------------------------------------

239 [Cooking]

240

241 +* ak/lazy-prereq-mktemp (2016-07-06) 1 commit

242 + - t7610: test for mktemp before test execution

243 +

244 + A test that unconditionally used "mktemp" learned that the command

245 + is not necessarily available everywhere.

246 +

247 + Will merge to 'next'.

248 +

249 +

250 +* ew/svn-bad-ref (2016-07-06) 1 commit

251 + - git-svn: warn instead of dying when commit data is missing

252 +

253 + An earlier update to "git svn" tightened the parsing of refs given

254 + as the command line arguments too much. This has been fixed by

255 + loosening the check a bit.

256 +

257 + I'm likely to discard this, favouring a direct pull request from

258 + the subsystem maintainer directly going to 'master'.

259 +

260 +

261 +* jk/difftool-in-subdir (2016-07-06) 1 commit

262 + - difftool: fix argument handling in subdirs

263 +

264 + "git difftool <paths>..." started in a subdirectory failed to

265 + interpret the paths relative to that directory, which has been

266 + fixed.

267 +

268 + Waiting for an ack.

269 +

270 +

271 +* nd/test-helpers (2016-07-06) 2 commits

272 + - Makefile: use VCSSVN_LIB to refer to svn library

273 + - Makefile: drop extra dependencies for test helpers

274 +

275 + Build clean-up.

276 +

277 + Will merge to 'next'.

278 +

279 +

280 * dk/blame-move-no-reason-for-1-line-context (2016-05-29) 1 commit

281 (merged to 'next' on 2016-07-06 at 087490c)

282 + blame: require 0 context lines while finding moved lines with -M

@@ -480,7 +292,7 @@ of the repositories listed at

292

293 Prepare jk/commin-main-2.8 topic to be mergeable to 2.9 and later.

294

483 - Will merge to 'master'.

295 + Will merge to 'next'.

296

297

298 * jk/common-main-2.8 (2016-07-06) 6 commits

@@ -604,13 +416,13 @@ of the repositories listed at

416 Will merge to 'master'.

417

418

607 -* ls/p4-tmp-refs (2016-06-29) 1 commit

419 +* ls/p4-tmp-refs (2016-07-08) 1 commit

420 - git-p4: place temporary refs used for branch import under refs/git-p4-tmp

421

422 "git p4" used a location outside $GIT_DIR/refs/ to place its

423 temporary branches, which has been moved to refs/git-p4-tmp/.

424

613 - Needs an ack from "git p4" experts.

425 + Will merge to 'next'.

426

427

428 * jc/pull-rebase-ff (2016-06-29) 1 commit

@@ -652,7 +464,7 @@ of the repositories listed at

464 The test framework learned a new helper test_match_signal to check

465 an exit code from getting killed by an expected signal.

466

655 - Is this what everybody can agree on?

467 + Will merge to 'next'.

468

469

470 * ah/unpack-trees-advice-messages (2016-06-27) 1 commit

@@ -783,7 +595,7 @@ of the repositories listed at

595 eventually the warning can be turned into a hard error, upgrading

596 the deprecation into removal of this (mis)feature.

597

786 - Will wait for further comments for a bit before merging to 'next'.

598 + Will merge to 'next'.

599

600

601 * jk/ansi-color (2016-06-23) 7 commits

@@ -894,8 +706,7 @@ of the repositories listed at

706 Error handling in the codepaths that updates refs has been

707 improved.

708

897 - Is everybody happy with this version?

898 - If so, will merge to 'next'.

709 + Will merge to 'next'.

710

711

712 * js/find-commit-subject-ignore-leading-blanks (2016-06-29) 5 commits

@@ -1087,8 +898,7 @@ of the repositories listed at

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

899 has been revamped.

900

1090 - Is everybody happy with this version?

1091 - If so, will merge to 'next'.

901 + Will merge to 'next'.

902

903

904 * mh/connect (2016-06-06) 10 commits

@@ -1161,10 +971,9 @@ of the repositories listed at

971 Will merge to 'master'.

972

973

1164 -* nd/worktree-lock (2016-07-01) 7 commits

974 +* nd/worktree-lock (2016-07-08) 6 commits

975 - worktree.c: find_worktree() search by path suffix

976 - worktree: add "unlock" command

1167 - - fixup! worktree: add "lock" command

977 - worktree: add "lock" command

978 - worktree.c: add is_worktree_locked()

979 - worktree.c: add is_main_worktree()

@@ -1175,7 +984,7 @@ of the repositories listed at

984 command learned a dedicated command pair to create and remoev such

985 a file, so that the users do not have to do this with editor.

986

1178 - Need to squash one fixup before merging to 'next'.

987 + Will merge to 'next'.

988

989

990 * sb/submodule-default-paths (2016-06-20) 5 commits

@@ -1290,8 +1099,7 @@ of the repositories listed at

1099 Further preparatory work on the refs API before the pluggable

1100 backend series can land.

1101

1293 - Is everybody happy with this version?

1294 - If so, will merge to 'next'.

1102 + Will merge to 'next'.

1103

1104

1105 * jc/send-email-skip-backup (2016-04-12) 1 commit

What's cooking (2016/07 #03)