1901a58e9f · matheusbsbs/matheus-bagliano

@@ -1,275 +1,215 @@

1 To: git@vger.kernel.org

2 -Subject: What's cooking in git.git (Jul 2012, #05; Sun, 15)

3 -X-master-at: 31c79549b85c6393be4f40432f5b86ebc097fc7e

4 -X-next-at: fdb3850c586fd1eced5e0456ef599f9c0c9191d9

2 +Subject: What's cooking in git.git (Jul 2012, #06; Sun, 22)

3 +X-master-at: e7719469154fc9bbf910df89b80a5145c7d93959

4 +X-next-at: 70d4bc158cb1d6d848984267f267177390e71989

5

6 -What's cooking in git.git (Jul 2012, #05; Sun, 15)

6 +What's cooking in git.git (Jul 2012, #06; Sun, 22)

7 --------------------------------------------------

8

9 Here are the topics that have been cooking. Commits prefixed with '-' are

10 only in 'pu' (proposed updates) while commits prefixed with '+' are in 'next'.

11

12 -The sixth batch of topics are now in 'master'.

12 +I think we are more or less done for this cycle except for a handful

13 +of topics that are to be rerolled, even though I haven't tagged the

14 +tip of the 'master' with 1.7.12-rc0 yet.

15

16 You can find the changes described here in the integration branches of the

17 repositories listed at

18

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

20

19 ---------------------------------------------------

20 -[New Topics]

21 -

22 -* sn/doc-typofix (2012-07-14) 1 commit

23 - - doc: A few minor copy edits.

24 -

21 --------------------------------------------------

22 [Graduated to "master"]

23

28 -* ar/clone-honor-umask-at-top (2012-07-09) 3 commits

29 - (merged to 'next' on 2012-07-10 at 680bc22)

30 - + add: create ADD_EDIT.patch with mode 0666

31 - + rerere: make rr-cache fanout directory honor umask

32 - + Restore umasks influence on the permissions of work tree created by clone

33 -

34 -A handful of files and directories we create had tighter than

35 -necessary permission bits when the user wanted to have group

36 -writability (e.g. by setting "umask 002").

37 -

38 -* cw/amend-commit-without-message (2012-07-09) 1 commit

39 - (merged to 'next' on 2012-07-10 at e7c75d5)

40 - + Allow edit of empty message with commit --amend

41 -

42 -"commit --amend" used to refuse amending a commit with an empty log

43 -message, with or without "--allow-empty-message".

44 -

45 -* cw/rebase-i-root (2012-07-05) 3 commits

46 - (merged to 'next' on 2012-07-05 at aeb4c56)

47 - + t3404: make test 57 work with dash and others

48 - (merged to 'next' on 2012-07-03 at 231c0a4)

49 - + Add tests for rebase -i --root without --onto

50 - + rebase -i: support --root without --onto

51 -

52 -"git rebase [-i] --root $tip" can now be used to rewrite all the

53 -history down to the root.

54 -

55 -* jc/apply-3way (2012-07-09) 19 commits

56 - (merged to 'next' on 2012-07-10 at fb5c872)

57 - + apply: tests for the --3way option

58 - + apply: document --3way option

59 - + apply: allow rerere() to work on --3way results

60 - + apply: register conflicted stages to the index

61 - + apply: --3way with add/add conflict

62 - + apply: move verify_index_match() higher

63 - + apply: plug the three-way merge logic in

64 - + apply: fall back on three-way merge

65 - + apply: accept -3/--3way command line option

66 - + apply: move "already exists" logic to check_to_create()

67 - + apply: move check_to_create_blob() closer to its sole caller

68 - + apply: further split load_preimage()

69 - + apply: refactor "previous patch" logic

70 - + apply: split load_preimage() helper function out

71 - + apply: factor out checkout_target() helper function

72 - + apply: refactor read_file_or_gitlink()

73 - + apply: clear_image() clears things a bit more

74 - + apply: a bit more comments on PATH_TO_BE_DELETED

75 - + apply: fix an incomplete comment in check_patch()

76 -

77 -"git apply" learns to wiggle the base version and perform three-way merge

78 -when a patch does not exactly apply to the version you have.

79 -

80 -* jk/index-pack-streaming-fix (2012-07-10) 1 commit

81 - (merged to 'next' on 2012-07-13 at c575338)

82 - + index-pack: loop while inflating objects in unpack_data

83 -

84 -The streaming index-pack introduced in 1.7.11 had a data corruption

85 -bug, and this should fix it.

86 -

87 -* jk/maint-commit-amend-only-no-paths (2012-07-10) 1 commit

88 - (merged to 'next' on 2012-07-10 at 1c650c2)

89 - + commit: fix "--amend --only" with no pathspec

90 -

91 -"git commit --amend --only --" was meant to allow "Clever" people to

92 -rewrite the commit message without making any change even when they

93 -have already changes for the next commit added to their index, but

94 -it never worked as advertised since it was introduced in 1.3.0 era.

95 -

96 -* jk/push-delete-ref-error-message (2012-07-03) 1 commit

97 - (merged to 'next' on 2012-07-09 at 82c6bd2)

98 - + push: don't guess at qualifying remote refs on deletion

99 -

100 -The error message from "git push $there :bogo" mentioned we tried

101 -and failed to guess what ref is being deleted based on the LHS of

102 -the refspec, which we don't.

103 -

104 -* jn/makefile-cleanup (2012-07-09) 14 commits

105 - (merged to 'next' on 2012-07-10 at f55e6a1)

106 - + Makefile: document ground rules for target-specific dependencies

107 - + Makefile: move GIT-VERSION-FILE dependencies closer to use

108 - + Makefile: build instaweb similar to other scripts

109 - + Makefile: update scripts when build-time parameters change

110 - + Makefile: do not replace @@GIT_VERSION@@ in shell scripts

111 - + Makefile: split prefix flags from GIT-CFLAGS

112 - + Makefile: be silent when only GIT_USER_AGENT changes

113 - + Makefile: split GIT_USER_AGENT from GIT-CFLAGS

114 - + Makefile: do not replace @@GIT_USER_AGENT@@ in scripts

115 - + Makefile: apply dependencies consistently to sparse/asm targets

116 - + Makefile: do not have git.o depend on common-cmds.h

117 - + Makefile: fold XDIFF_H and VCSSVN_H into LIB_H

118 - + Makefile: fold MISC_H into LIB_H

119 - + Makefile: sort LIB_H list

120 -

121 -Tightens dependency rules to avoid unnecessary recompilation.

122 -

123 -* pw/git-p4-jobs (2012-07-05) 3 commits

124 - (merged to 'next' on 2012-07-10 at c6ef8f9)

125 - + git p4: notice Jobs lines in git commit messages

126 - + git p4 test: refactor marshal_dump

127 - + git p4: remove unused P4Submit interactive setting

128 -

129 -Teach "git p4" to notice "Jobs:" in the log message and relay it to

130 -Perforce to trigger its "jobs" support.

131 -

132 -* pw/git-p4-move (2012-07-12) 2 commits

133 - (merged to 'next' on 2012-07-13 at 648a8ad)

134 - + git p4: add support for 'p4 move' in P4Submit

135 - + git p4: refactor diffOpts calculation

136 -

137 -* tg/maint-cache-name-compare (2012-07-11) 1 commit

138 - (merged to 'next' on 2012-07-13 at 0d07ddc)

139 - + cache_name_compare(): do not truncate while comparing paths

140 - (this branch is used by tg/ce-namelen-field.)

141 -

142 -Even though the index can record pathnames longer than 1<<12 bytes,

143 -in some places we were not comparing them in full, potentially

144 -replacing index entries instead of adding.

145 -

146 -Will merge down to older maintenance releases.

147 -

148 ---------------------------------------------------

149 -[Stalled]

150 -

151 -* jc/maint-push-refs-all (2012-05-04) 2 commits

152 - - get_fetch_map(): tighten checks on dest refs

153 - - fetch/push: allow refs/*:refs/*

154 -

155 -Allows pushing and fetching refs/stash.

156 -There still seem to be other bugs hiding (e.g. try pushing twice).

157 -

158 -Not ready.

24 +* as/t4012-style-updates (2012-07-16) 8 commits

25 + (merged to 'next' on 2012-07-16 at 86c14f6)

26 + + t4012: Use test_must_fail instead of if-else

27 + + t4012: use 'printf' instead of 'dd' to generate a binary file

28 + + t4012: Re-indent test snippets

29 + + t4012: Make --shortstat test more robust

30 + + t4012: Break up pipe into serial redirections

31 + + t4012: Actually quote the sed script

32 + + t4012: Unquote git command fragment in test title

33 + + t4012: modernize style for quoting

34

160 -* jc/split-blob (2012-04-03) 6 commits

161 - - chunked-object: streaming checkout

162 - - chunked-object: fallback checkout codepaths

163 - - bulk-checkin: support chunked-object encoding

164 - - bulk-checkin: allow the same data to be multiply hashed

165 - - new representation types in the packstream

166 - - packfile: use varint functions

167 -

168 -Not ready.

35 +* jc/maint-filter-branch-epoch-date (2012-07-12) 3 commits

36 + (merged to 'next' on 2012-07-13 at d144342)

37 + + t7003: add test to filter a branch with a commit at epoch

38 + + date.c: Fix off by one error in object-header date parsing

39 + (merged to 'next' on 2012-07-10 at f9774ca)

40 + + filter-branch: do not forget the '@' prefix to force git-timestamp

41

170 -I finished the streaming checkout codepath, but as explained in

171 -127b177 (bulk-checkin: support chunked-object encoding, 2011-11-30),

172 -these are still early steps of a long and painful journey. At least

173 -pack-objects and fsck need to learn the new encoding for the series

174 -to be usable locally, and then index-pack/unpack-objects needs to

175 -learn it to be used remotely.

42 +In 1.7.9 era, we taught "git rebase" about the raw timestamp format

43 +but we did not teach the same trick to "filter-branch", which rolled

44 +a similar logic on its own.

45

177 -Given that I heard a lot of noise that people want large files, and

178 -that I was asked by somebody at GitTogether'11 privately for an

179 -advice on how to pay developers (not me) to help adding necessary

180 -support, I am somewhat dissapointed that the original patch series

181 -that was sent almost two months ago still remains here without much

182 -comments and updates from the developer community. I even made the

183 -interface to the logic that decides where to split chunks easily

184 -replaceable, and I deliberately made the logic in the original patch

185 -extremely stupid to entice others, especially the "bup" fanboys, to

186 -come up with a better logic, thinking that giving people an easy

187 -target to shoot for, they may be encouraged to help out. The plan is

188 -not working :-(.

46 +* jc/sha1-name-more (2012-07-13) 27 commits

47 + (merged to 'next' on 2012-07-13 at 534e6fe)

48 + + t1512: match the "other" object names

49 + (merged to 'next' on 2012-07-11 at d30ef51)

50 + + t1512: ignore whitespaces in wc -l output

51 + (merged to 'next' on 2012-07-10 at 669dd92)

52 + + rev-parse --disambiguate=<prefix>

53 + + rev-parse: A and B in "rev-parse A..B" refer to committish

54 + + reset: the command takes committish

55 + + commit-tree: the command wants a tree and commits

56 + + apply: --build-fake-ancestor expects blobs

57 + + sha1_name.c: add support for disambiguating other types

58 + + revision.c: the "log" family, except for "show", takes committish

59 + + revision.c: allow handle_revision_arg() to take other flags

60 + + sha1_name.c: introduce get_sha1_committish()

61 + + sha1_name.c: teach lookup context to get_sha1_with_context()

62 + + sha1_name.c: many short names can only be committish

63 + + sha1_name.c: get_sha1_1() takes lookup flags

64 + + sha1_name.c: get_describe_name() by definition groks only commits

65 + + sha1_name.c: teach get_short_sha1() a commit-only option

66 + + sha1_name.c: allow get_short_sha1() to take other flags

67 + + get_sha1(): fix error status regression

68 + + sha1_name.c: restructure disambiguation of short names

69 + + sha1_name.c: correct misnamed "canonical" and "res"

70 + + sha1_name.c: refactor find_short_packed_object()

71 + + sha1_name.c: rename "now" to "current"

72 + + sha1_name.c: clarify what "fake" is for in find_short_object_filename()

73 + + sha1_name.c: get rid of get_sha1_with_mode()

74 + + sha1_name.c: get rid of get_sha1_with_mode_1()

75 + + sha1_name.c: hide get_sha1_with_context_1() ugliness

76 + + sha1_name.c: indentation fix

77

190 ---------------------------------------------------

191 -[Cooking]

78 +Logic to disambiguate abbreviated object names have been taught to

79 +take advantage of object types that are expected in the context,

80 +e.g. XXXXXX in the "git describe" output v1.2.3-gXXXXXX must be a

81 +commit object, not a blob nor a tree. This will help us prolong the

82 +lifetime of abbreviated object names.

83

193 -* as/t4012-style-updates (2012-07-12) 7 commits

194 - - t4012: use 'printf' instead of 'dd' to generate a binary file

195 - - t4012: Re-indent test snippets

196 - - t4012: Make --shortstat test more robust

197 - - t4012: Break up pipe into serial redirections

198 - - t4012: Actually quote the sed script

199 - - t4012: Unquote git command fragment in test title

200 - - t4012: modernize style for quoting

84 +* jk/revision-walk-stop-at-max-count (2012-07-13) 1 commit

85 + (merged to 'next' on 2012-07-16 at 603975d)

86 + + revision: avoid work after --max-count is reached

87

202 -Expecting a reroll.

88 +"git log -n 1 -- rarely-touched-path" was spending unnecessary

89 +cycles after showing the first change to find the next one, only to

90 +discard it.

91

92 * jl/maint-1.7.10-recurse-submodules-with-symlink (2012-07-12) 1 commit

205 - - submodules: don't stumble over symbolic links when cloning recursively

93 + (merged to 'next' on 2012-07-16 at 2e10caf)

94 + + submodules: don't stumble over symbolic links when cloning recursively

95

96 When "git submodule add" clones a submodule repository, it can get

97 confused where to store the resulting submodule repository in the

98 superproject's .git/ directory when there is a symbolic link in the

99 path to the current directory.

100

212 -Will merge to 'next'.

213 -

214 -* mm/config-xdg (2012-07-12) 2 commits

215 - - fixup! config: fix several access(NULL) calls

216 - - config: fix several access(NULL) calls

101 +* lk/more-helpful-status-hints (2012-07-16) 1 commit

102 + (merged to 'next' on 2012-07-16 at 8f9b3cc)

103 + + status: color in-progress message like other header messages

104

218 -Either we do the stupid literal conversion (shown in fixup!), or the

219 -reason why we special case missing HOME needs to be better explained.

105 +Finishing touches.

106

221 -* jk/revision-walk-stop-at-max-count (2012-07-13) 1 commit

222 - - revision: avoid work after --max-count is reached

107 +* mm/config-xdg (2012-07-16) 1 commit

108 + (merged to 'next' on 2012-07-16 at 1a75d75)

109 + + config: fix several access(NULL) calls

110

224 -"git log -n 1 -- rarely-touched-path" was spending unnecessary

225 -cycles after showing the first change to find the next one, only to

226 -discard it.

111 +Finishing touches.

112

228 -Will merge to 'next'.

113 +* ms/daemon-doc-typo (2012-07-16) 1 commit

114 + (merged to 'next' on 2012-07-16 at 3976434)

115 + + Documentation/git-daemon: add missing word

116

117 * pg/maint-1.7.9-am-where-is-patch (2012-07-13) 1 commit

231 - - am: indicate where a failed patch is to be found

118 + (merged to 'next' on 2012-07-16 at 32f1e30)

119 + + am: indicate where a failed patch is to be found

120

121 When "git am" failed, old timers knew to check .git/rebase-apply/patch

122 to see what went wrong, but we never told the users about it.

123

236 -Will merge to 'next'.

124 +--------------------------------------------------

125 +[New Topics]

126

238 -* jc/maint-filter-branch-epoch-date (2012-07-12) 3 commits

239 - (merged to 'next' on 2012-07-13 at d144342)

240 - + t7003: add test to filter a branch with a commit at epoch

241 - + date.c: Fix off by one error in object-header date parsing

242 - (merged to 'next' on 2012-07-10 at f9774ca)

243 - + filter-branch: do not forget the '@' prefix to force git-timestamp

127 +* mm/mediawiki-usability (2012-07-17) 10 commits

128 + (merged to 'next' on 2012-07-22 at fe66a95)

129 + + git-remote-mediawiki: allow page names with a ':'

130 + + git-remote-mediawiki: fix incorrect test usage in test

131 + + git-remote-mediawiki: properly deal with invalid remote revisions

132 + + git-remote-mediawiki: show progress information when getting last remote revision

133 + + git-remote-mediawiki: show progress information when listing pages

134 + + git-remote-mediawiki: use --force when adding notes

135 + + git-remote-mediawiki: get rid of O(N^2) loop

136 + + git-remote-mediawiki: make mediafiles export optional

137 + + git-remote-mediawiki: actually send empty comment when they're empty

138 + + git-remote-mediawiki: don't split namespaces with spaces

139

245 -In 1.7.9 era, we taught "git rebase" about the raw timestamp format

246 -but we did not teach the same trick to "filter-branch", which rolled

247 -a similar logic on its own.

140 +Will merge to 'master'.

141 +

142 +* jk/mediawiki-credential (2012-07-18) 4 commits

143 + (merged to 'next' on 2012-07-22 at 2cb99b2)

144 + + mw-to-git: use git-credential's URL parser

145 + + credential: convert "url" attribute into its parsed subparts

146 + + mw-to-git: check blank credential attributes via length

147 + + docs/credential: minor clarity fixups

148

149 Will merge to 'master'.

150

251 -* mb/remote-default-nn-origin (2012-07-11) 6 commits

252 - - Teach get_default_remote to respect remote.default.

253 - - Test that plain "git fetch" uses remote.default when on a detached HEAD.

254 - - Teach clone to set remote.default.

255 - - Teach "git remote" about remote.default.

256 - - Teach remote.c about the remote.default configuration setting.

257 - - Rename remote.c's default_remote_name static variables.

151 +* jv/maint-no-ext-diff (2012-07-19) 2 commits

152 + (merged to 'next' on 2012-07-22 at eefcf45)

153 + + diff: test precedence of external diff drivers

154 + + diff: correctly disable external_diff with --no-ext-diff

155

259 -When the user does not specify what remote to interact with, we

260 -often attempt to use 'origin'. This can now be customized via a

261 -configuration variable.

156 +"git diff --no-ext-diff" did not output anything for a typechange

157 +filepair when GIT_EXTERNAL_DIFF is in effect.

158

263 -Expecting a reroll.

264 -"The first remote becomes the default" bit is better done as a

265 -separate step.

159 +Will merge to 'master'.

160

267 -* tg/ce-namelen-field (2012-07-11) 2 commits

268 - - Strip namelen out of ce_flags into a ce_namelen field

269 - - Merge branch 'tg/maint-cache-name-compare' into tg/ce-namelen-field

161 +* kk/maint-1.7.9-commit-tree (2012-07-17) 1 commit

162 + + commit-tree: resurrect command line parsing updates

163 + (this branch is used by kk/maint-commit-tree.)

164

271 -Split lower bits of ce_flags field and creates a new ce_namelen

272 -field in the in-core index structure.

165 +A more natural-looking "git commit-tree -p <parent> <tree>" syntax

166 +was introduced long time ago, but we accidentally broke it in the

167 +1.7.9 era.

168 +

169 +* kk/maint-commit-tree (2012-07-17) 2 commits

170 + (merged to 'next' on 2012-07-22 at ab15d56)

171 + + Revert "git-commit-tree(1): update synopsis"

172 + + Merge branch 'kk/maint-1.7.9-commit-tree' into kk/maint-commit-tree

173 + (this branch uses kk/maint-1.7.9-commit-tree.)

174 +

175 +Will merge to 'master'.

176 +

177 +* mz/rebase-range (2012-07-18) 7 commits

178 + - rebase (without -p): correctly calculate patches to rebase

179 + - rebase -p: don't request --left-right only to ignore left side

180 + - rebase -p: use --cherry-mark for todo file

181 + - git-rebase--interactive.sh: look up subject in add_pick_line

182 + - git-rebase--interactive: group all $preserve_merges code

183 + - git-rebase--interactive.sh: extract function for adding "pick" line

184 + - git-rebase--am.sh: avoid special-casing --keep-empty

185 +

186 +* sl/autoconf (2012-07-19) 7 commits

187 + - build: reconfigure automatically if configure.ac changes

188 + - build: "make clean" should not remove configure-generated files

189 + - autoconf: use AC_CONFIG_COMMANDS instead of ad-hoc 'config.mak.append'

190 + - autoconf: remove few redundant semicolons

191 + - autoconf: remove some redundant shell indirections

192 + - autoconf: GIT_CONF_APPEND_LINE -> GIT_CONF_SUBST

193 + - autoconf: GIT_CONF_APPEND_LINE: change signature

194 +

195 +Will merge to 'next'.

196 +

197 +* jx/i18n-1.7.11 (2012-07-22) 7 commits

198 + - i18n: merge-recursive: mark strings for translation

199 + - Remove unused and bad gettext block from git-am

200 + - i18n: am: mark more strings for translation

201 + - Remove obsolete LONG_USAGE which breaks xgettext

202 + - i18n: Rewrite gettext messages start with dash

203 + - i18n: rebase: mark strings for translation

204 + - i18n: New keywords for xgettext extraction from sh

205 +

206 +* th/difftool-diffall (2012-07-19) 1 commit

207 + - difftool: only copy back files modified during directory diff

208 +

209 +Will merge to 'next'.

210 +

211 +--------------------------------------------------

212 +[Stalled]

213

214 * jl/submodule-rm (2012-07-05) 2 commits

215 - rm: remove submodules from the index and the .gitmodules file

@@ -283,17 +223,6 @@ Expecting a reroll.

223

224 Expecting a reroll of the tip one.

225

286 -* nk/maint-gitweb-log-by-lines (2012-07-05) 3 commits

287 - (merged to 'next' on 2012-07-13 at 780e16a)

288 - + gitweb: Add support to Link: tag

289 - + gitweb: Handle other types of tag in git_print_log

290 - + gitweb: Cleanup git_print_log()

291 -

292 -Teach gitweb to pay attention to various forms of credits that are

293 -similar to "Signed-off-by:" lines.

294 -

295 -Needs to be eyeballed for the correctness of the esc_html() in the tip one.

296 -

226 * ph/stash-rerere (2012-07-08) 2 commits

227 - stash: invoke rerere in case of conflict

228 - test: git-stash conflict sets up rerere

@@ -311,6 +240,17 @@ sense" heuristics "git mailinfo" already uses.

240 A draft from Linus received privately without a log message.

241 Expecting a reroll.

242

243 +* nk/maint-gitweb-log-by-lines (2012-07-05) 3 commits

244 + (merged to 'next' on 2012-07-13 at 780e16a)

245 + + gitweb: Add support to Link: tag

246 + + gitweb: Handle other types of tag in git_print_log

247 + + gitweb: Cleanup git_print_log()

248 +

249 +Teach gitweb to pay attention to various forms of credits that are

250 +similar to "Signed-off-by:" lines.

251 +

252 +Needs to be eyeballed for the correctness of the esc_html() in the tip one.

253 +

254 * jc/test-lib-source-build-options-early (2012-06-24) 1 commit

255 - test-lib: reorder and include GIT-BUILD-OPTIONS a lot earlier

256

@@ -322,44 +262,77 @@ one of its early operations.

262 Needs to be eyeballed by people who run tests with exotic options

263 like valgrind, --root=/dev/shm/somewhere, etc.

264

325 -Will merge to 'next'.

265 +* mb/remote-default-nn-origin (2012-07-11) 6 commits

266 + - Teach get_default_remote to respect remote.default.

267 + - Test that plain "git fetch" uses remote.default when on a detached HEAD.

268 + - Teach clone to set remote.default.

269 + - Teach "git remote" about remote.default.

270 + - Teach remote.c about the remote.default configuration setting.

271 + - Rename remote.c's default_remote_name static variables.

272

327 -* jc/sha1-name-more (2012-07-13) 27 commits

328 - (merged to 'next' on 2012-07-13 at 534e6fe)

329 - + t1512: match the "other" object names

330 - (merged to 'next' on 2012-07-11 at d30ef51)

331 - + t1512: ignore whitespaces in wc -l output

332 - (merged to 'next' on 2012-07-10 at 669dd92)

333 - + rev-parse --disambiguate=<prefix>

334 - + rev-parse: A and B in "rev-parse A..B" refer to committish

335 - + reset: the command takes committish

336 - + commit-tree: the command wants a tree and commits

337 - + apply: --build-fake-ancestor expects blobs

338 - + sha1_name.c: add support for disambiguating other types

339 - + revision.c: the "log" family, except for "show", takes committish

340 - + revision.c: allow handle_revision_arg() to take other flags

341 - + sha1_name.c: introduce get_sha1_committish()

342 - + sha1_name.c: teach lookup context to get_sha1_with_context()

343 - + sha1_name.c: many short names can only be committish

344 - + sha1_name.c: get_sha1_1() takes lookup flags

345 - + sha1_name.c: get_describe_name() by definition groks only commits

346 - + sha1_name.c: teach get_short_sha1() a commit-only option

347 - + sha1_name.c: allow get_short_sha1() to take other flags

348 - + get_sha1(): fix error status regression

349 - + sha1_name.c: restructure disambiguation of short names

350 - + sha1_name.c: correct misnamed "canonical" and "res"

351 - + sha1_name.c: refactor find_short_packed_object()

352 - + sha1_name.c: rename "now" to "current"

353 - + sha1_name.c: clarify what "fake" is for in find_short_object_filename()

354 - + sha1_name.c: get rid of get_sha1_with_mode()

355 - + sha1_name.c: get rid of get_sha1_with_mode_1()

356 - + sha1_name.c: hide get_sha1_with_context_1() ugliness

357 - + sha1_name.c: indentation fix

273 +When the user does not specify what remote to interact with, we

274 +often attempt to use 'origin'. This can now be customized via a

275 +configuration variable.

276 +

277 +Expecting a reroll.

278 +"The first remote becomes the default" bit is better done as a

279 +separate step.

280

359 -Teaches the object name parser things like a "git describe" output

360 -is always a commit object, "A" in "git log A" must be a committish,

361 -and "A" and "B" in "git log A...B" both must be committish, etc., to

362 -prolong the lifetime of abbreviated object names.

281 +* jc/maint-push-refs-all (2012-05-04) 2 commits

282 + - get_fetch_map(): tighten checks on dest refs

283 + - fetch/push: allow refs/*:refs/*

284 +

285 +Allows pushing and fetching refs/stash.

286 +There still seem to be other bugs hiding (e.g. try pushing twice).

287 +

288 +Not ready.

289 +

290 +* jc/split-blob (2012-04-03) 6 commits

291 + - chunked-object: streaming checkout

292 + - chunked-object: fallback checkout codepaths

293 + - bulk-checkin: support chunked-object encoding

294 + - bulk-checkin: allow the same data to be multiply hashed

295 + - new representation types in the packstream

296 + - packfile: use varint functions

297 +

298 +Not ready.

299 +

300 +I finished the streaming checkout codepath, but as explained in

301 +127b177 (bulk-checkin: support chunked-object encoding, 2011-11-30),

302 +these are still early steps of a long and painful journey. At least

303 +pack-objects and fsck need to learn the new encoding for the series

304 +to be usable locally, and then index-pack/unpack-objects needs to

305 +learn it to be used remotely.

306 +

307 +Given that I heard a lot of noise that people want large files, and

308 +that I was asked by somebody at GitTogether'11 privately for an

309 +advice on how to pay developers (not me) to help adding necessary

310 +support, I am somewhat dissapointed that the original patch series

311 +that was sent almost two months ago still remains here without much

312 +comments and updates from the developer community. I even made the

313 +interface to the logic that decides where to split chunks easily

314 +replaceable, and I deliberately made the logic in the original patch

315 +extremely stupid to entice others, especially the "bup" fanboys, to

316 +come up with a better logic, thinking that giving people an easy

317 +target to shoot for, they may be encouraged to help out. The plan is

318 +not working :-(.

319 +

320 +--------------------------------------------------

321 +[Cooking]

322 +

323 +* sn/doc-typofix (2012-07-14) 1 commit

324 + (merged to 'next' on 2012-07-22 at 168bba9)

325 + + doc: A few minor copy edits.

326 +

327 +Will merge to 'master'.

328 +

329 +* tg/ce-namelen-field (2012-07-11) 2 commits

330 + (merged to 'next' on 2012-07-22 at 2d85b05)

331 + + Strip namelen out of ce_flags into a ce_namelen field

332 + + Merge branch 'tg/maint-cache-name-compare' into tg/ce-namelen-field

333 +

334 +Split lower bits of ce_flags field and creates a new ce_namelen

335 +field in the in-core index structure.

336

337 Will merge to 'master'.

338

What's cooking (2012/07 #06)