aa98bd2cb1 · matheusbsbs/matheus-bagliano

@@ -1,9 +1,9 @@

1 To: git@vger.kernel.org

2 -Subject: What's cooking in git.git (Sep 2009, #05; Wed, 23)

3 -X-master-at: a6dbf8814f433a7fbfa9cde6333c98019f6db1e4

4 -X-next-at: a9dd6c4e8ab38f5c4af9c0710381d3844f1d904f

2 +Subject: What's cooking in git.git (Oct 2009, #01; Wed, 07)

3 +X-master-at: 8fd2cfa7accd6b8f877014bf3e4bf8547b8e0d2a

4 +X-next-at: df7ac2094d6cd7280e3254a772061a64edac2d37

5

6 -What's cooking in git.git (Sep 2009, #05; Wed, 23)

6 +What's cooking in git.git (Oct 2009, #01; Wed, 07)

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

8

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

@@ -54,50 +54,108 @@ backward compatible. Currently planeed changes are:

54 We are in pre-release feature freeze. 'next' will hold topics meant for

55 1.6.6 and 1.7.0.

56

57 -Tonight's tip of 'master' is 1.6.5-rc2, and I'll disappear for about a

58 -week. Hopefully when I come back in early October all the regressions are

59 -found and already squashed ;-)

57 +Tonight's tip of 'master' is 1.6.5-rc3.

58

59 --------------------------------------------------

60 [New Topics]

61

64 -* jc/fix-tree-walk (2009-09-14) 9 commits

65 - - read-tree --debug-unpack

66 - - unpack-trees.c: look ahead in the index

67 - - unpack-trees.c: prepare for looking ahead in the index

68 - - traverse_trees(): handle D/F conflict case sanely

69 - - more D/F conflict tests

70 - - tests: move convenience regexp to match object names to test-lib.sh

71 - - unpack_callback(): use unpack_failed() consistently

72 - - unpack-trees: typofix

73 - - diff-lib.c: fix misleading comments on oneway_diff()

62 +* ch/am-header (2009-09-25) 2 commits

63 + (merged to 'next' on 2009-09-25 at f86e197)

64 + + git-am: force egrep to use correct characters set

65 + + git-am: fixed patch_format detection according to RFC2822

66

75 -This is my replacement for Linus's lt/maint-traverse-trees-fix patch. It

76 -is not so much as a counter-proposal; I originally thought it might make

77 -sense to walk the index and drive the walker to return the entries from

78 -trees to match entries from the index, but I ended up doing pretty much

79 -what Linus outlined --- walk the trees, and have the index walker follow

80 -it. It turned out that the index side also needed some hairy look-ahead,

81 -and I am only half satisfied with the current status of the series.

67 +* dk/blame-el (2009-09-29) 1 commit

68 + - git-blame.el: Change how blame information is shown.

69

83 -To fix the resolve merge regression seen in t6035, git-merge-resolve needs

84 -to be rewritten not to use the one-path-at-a-time "git merge-index".

70 +* ef/msvc-noreturn (2009-09-30) 2 commits

71 + (merged to 'next' on 2009-10-07 at 66137a0)

72 + + add NORETURN_PTR for function pointers

73 + + increase portability of NORETURN declarations

74

86 -* jp/fetch-tag-match (2009-09-17) 1 commit

87 - - fetch: Speed up fetch by rewriting find_non_local_tags

75 +jk: This is the latest round and I think should be ready for at least

76 +'next' (maybe even 'master' as it is really about the build and not about

77 +functionality).

78

89 -I did not have much energy left while dealing with the "fix-tree-walk"

90 -series, so I just queued this without reading nor thinking about it very

91 -much. I personally liked my version that had far smaller number of lines

92 -changed (which means I can be fairly certain that it did not introduce any

93 -regression), but perhaps the majorly rewritten logic this patch gives us

94 -may be easier to follow and maintain. We'll see.

79 +* ef/msys-imap (2009-10-03) 7 commits

80 + - mingw: enable OpenSSL

81 + - mingw: wrap SSL_set_(w|r)fd to call _get_osfhandle

82 + - imap-send: provide fall-back random-source

83 + - imap-send: build imap-send on Windows

84 + - imap-send: fix compilation-error on Windows

85 + - imap-send: use run-command API for tunneling

86 + - imap-send: use separate read and write fds

87 +

88 +jk: This is from an RFC which has generated some comments. He should be

89 +posting another round soon. 'pu' at best.

90 +

91 +* fc/mutt-alias (2009-09-30) 1 commit

92 + (merged to 'next' on 2009-10-07 at df7ac20)

93 + + send-email: fix mutt regex for grouped aliases

94 +

95 +jk: Latest round that addressed comments. Ready for 'next' if not

96 +'master'.

97 +

98 +* jk/reflog-date (2009-09-24) 1 commit

99 + (merged to 'next' on 2009-09-29 at 43d444a)

100 + + improve reflog date/number heuristic

101 +

102 +* jn/gitweb-patch (2009-09-30) 1 commit

103 + - gitweb: Do not show 'patch' link in 'commit' view for merges

104 +

105 +jk: After some comments with Jakub, I think the code is right but he

106 +promised a re-roll with more in the commit message.

107 +

108 +* mr/gitweb-snapshot (2009-09-26) 2 commits

109 + - gitweb: append short hash ids to snapshot files

110 + - gitweb: check given hash before trying to create snapshot

111 +

112 +jk: He posted a v5 of his series. I didn't look at it closely, but Jakub

113 +ack'd it.

114 +

115 +* mr/instaweb-cgid (2009-09-26) 1 commit

116 + (merged to 'next' on 2009-09-29 at 3524604)

117 + + instaweb: support mod_cgid for apache2

118 +

119 +* tf/doc-pt-br (2009-09-23) 1 commit

120 + - Documentation: update pt-BR

121 +

122 +The current AsciiDoc may barf on NOME and SINOPSE, as pt_BR language

123 +definition is not widely distributed yet (it just hit the development

124 +tree). Need to revert these headings (or change the length of the section

125 +underlines to match the length of translated names).

126 +

127 +* jc/pretty-lf (2009-10-04) 1 commit

128 + - Pretty-format: %[+-]x to tweak inter-item newlines

129 +

130 +I am not happy with this one yet. I am contemplating to introduce a new

131 +syntax "%[magic(param)<anything>%]" to generalize expressions of this and

132 +line wrapping features in an extensible way.

133 +

134 +* js/log-rewrap (2008-11-10) 3 commits

135 + . Add "%w" to pretty formats, which rewraps the commit message

136 + - Add strbuf_add_wrapped_text() to utf8.[ch]

137 + - print_wrapped_text(): allow hard newlines

138 +

139 +... and the first two from this series will be useful to implement an

140 +example magic "wrap", e.g. "%{wrap(i,j,w)%s%+b%]".

141

142 * jg/log-format-body-indent (2009-09-19) 1 commit

97 - - [NEEDSWORK: tests] git-log --format: Add %B tag with %B(x) option

143 + . git-log --format: Add %B tag with %B(x) option

144

99 -* jl/submodule-add-noname (2009-09-22) 1 commit

100 - - git submodule add: make the <path> parameter optional

145 +I think we should redo this on top of the first two patches from

146 +js/log-rewrap series; %B(x) is just a special case %B(x,x,0), no? If a

147 +magic value 0 (or negative) given to wrap-width does not disable wrapping,

148 +we probably should make it so. I merged this to 'pu' but then ejected it

149 +because it seems to break at least t6006.

150 +

151 +* bg/rebase-reword (2009-10-07) 1 commit

152 + - Teach 'rebase -i' the command "reword"

153 +

154 +* js/diff-verbose-submodule (2009-10-04) 1 commit

155 + - Add the --submodule-summary option to the diff option family

156 +

157 +Dscho sounded like he has some corrections after list comments, but I did

158 +not pick up his interdiff in the middle.

159

160 --------------------------------------------------

161 [Stalled]

@@ -134,12 +192,53 @@ Builds on db/vcs-helper. There is a re-roll planned.

192 - basic revision cache system, no integration or features

193 - man page and technical discussion for rev-cache

194

137 -Tonight's 'pu' ships with this and this series seems to break a few

138 -tests. I didn't debug.

195 +I merged this to 'pu' but then ejected it because it seems to break at

196 +least t6001.

197

198 --------------------------------------------------

199 [Cooking]

200

201 +* jl/submodule-add-noname (2009-09-22) 1 commit

202 + - git submodule add: make the <path> parameter optional

203 +

204 +Dscho started an interesting discussion regarding the larger workflow in

205 +which the "submodule add" is used. I think the patch itself makes sense

206 +but at the same time it probably makes sense to also take the <path> and

207 +infer the <repository> as Dscho suggested, probably in "git submodule

208 +add", not in "git add" proper, at least initially.

209 +

210 +* jc/fix-tree-walk (2009-09-14) 9 commits

211 + - read-tree --debug-unpack

212 + - unpack-trees.c: look ahead in the index

213 + - unpack-trees.c: prepare for looking ahead in the index

214 + - traverse_trees(): handle D/F conflict case sanely

215 + - more D/F conflict tests

216 + - tests: move convenience regexp to match object names to test-lib.sh

217 + - unpack_callback(): use unpack_failed() consistently

218 + - unpack-trees: typofix

219 + - diff-lib.c: fix misleading comments on oneway_diff()

220 +

221 +This is my replacement for Linus's lt/maint-traverse-trees-fix patch. It

222 +is not so much as a counter-proposal; I originally thought it might make

223 +sense to walk the index and drive the walker to return the entries from

224 +trees to match entries from the index, but I ended up doing pretty much

225 +what Linus outlined --- walk the trees, and have the index walker follow

226 +it. It turned out that the index side also needed some hairy look-ahead,

227 +and I am only half satisfied with the current status of the series.

228 +

229 +To fix the resolve merge regression seen in t6035, git-merge-resolve needs

230 +to be rewritten not to use the one-path-at-a-time "git merge-index".

231 +

232 +* jp/fetch-tag-match (2009-09-17) 1 commit

233 + - fetch: Speed up fetch by rewriting find_non_local_tags

234 +

235 +I did not have much energy left while dealing with the "fix-tree-walk"

236 +series, so I just queued this without reading nor thinking about it very

237 +much. I personally liked my version that had far smaller number of lines

238 +changed (which means I can be fairly certain that it did not introduce any

239 +regression), but perhaps the majorly rewritten logic this patch gives us

240 +may be easier to follow and maintain. We'll see.

241 +

242 * jc/maint-blank-at-eof (2009-09-15) 0 commits.

243 (this branch uses jc/maint-1.6.0-blank-at-eof.)

244

What's cooking (2009/10 #01)