What's cooking (2009/09 #02)
Junio C Hamano committed
Sep 7, 2009 at 17:50 UTC
235f87a0a89892f35600957a96985630d5a18df3
1 file changed
+232
-161
whats-cooking.txt
+232
-161
@@ -1,9 +1,9 @@
1
To: git@vger.kernel.org
2
-Subject: What's cooking in git.git (Sep 2009, #01; Sat, 05)
3
-X-master-at: 79b4fde573eb5e86a8109d100ad182be1eb5fc54
4
-X-next-at: 85dc3ec32928a03c0f4996b00d7a74f5a8f9045e
2
+Subject: What's cooking in git.git (Sep 2009, #02; Mon, 07)
3
+X-master-at: 5ad9dce7e691106fecde413de8cc321b937367a6
4
+X-next-at: 5f6b0ffff13f5cd762d0a5a4e1c4dede58e8a537
5
6
-What's cooking in git.git (Sep 2009, #01; Sat, 05)
6
+What's cooking in git.git (Sep 2009, #02; Mon, 07)
7
--------------------------------------------------
8
9
Here are the topics that have been cooking. Commits prefixed with '-' are
@@ -11,55 +11,174 @@ only in 'pu' while commits prefixed with '+' are in 'next'. The ones
11
marked with '.' do not appear in any of the integration branches, but I am
12
still holding onto them.
13
14
-After the 1.6.5 cycle, the next release will be 1.7.0, and we will push
15
-out the planned "push safety" change. 1.7.0 would be a good time to
16
-introduce "justifiable" changes that are not strictly backward compatible.
14
+In 1.7.0, we plan to correct handful of warts in the interfaces everybody
15
+agrees that they were mistakes. The resulting system may not be strictly
16
+backward compatible. Currently planeed changes are:
17
+
18
+ * refuse push to update the checked out branch in a non-bare repo by
19
+ default
20
+
21
+ Make "git push" into a repository to update the branch that is checked
22
+ out fail by default. You can countermand this default by setting a
23
+ configuration variable in the receiving repository.
24
+
25
+ http://thread.gmane.org/gmane.comp.version-control.git/107758/focus=108007
26
+
27
+ * refuse push to delete the current branch by default
28
+
29
+ Make "git push $there :$killed" to delete the branch that is pointed at
30
+ by its HEAD fail by default. You can countermand this default by
31
+ setting a configuration variable in the receiving repository.
32
+
33
+ http://thread.gmane.org/gmane.comp.version-control.git/108862/focus=108936
34
+
35
+ * git-send-email won't make deep threads by default
36
+
37
+ Many people said that by default when sending more than 2 patches the
38
+ threading git-send-email makes by default is hard to read, and they
39
+ prefer the default be one cover letter and each patch as a direct
40
+ follow-up to the cover letter. You can countermand this by setting a
41
+ configuration variable.
42
+
43
+ http://article.gmane.org/gmane.comp.version-control.git/109790
44
+
45
+ * git-status won't be "git-commit --dry-run" anymore
46
+
47
+ http://thread.gmane.org/gmane.comp.version-control.git/125989/focus=125993
48
+
49
+ * "git-diff -w --exit-code" will exit success if only differences it
50
+ found are whitespace changes that are stripped away from the output.
51
+
52
+ http://thread.gmane.org/gmane.comp.version-control.git/119731/focus=119751
53
54
During 1.6.5 cycle, 'next' will hold topics meant for 1.6.5 and 1.7.0.
55
20
-I will probably do 1.6.5-rc0 this weekend, leaving some topics still
21
-cooking in 'next'.
56
+I tagged and pushed out 1.6.5-rc0. There are still some topics cooking in
57
+'next' I would rather not leave out, but I've been hoping to keep this
58
+cycle short, and I'm also hoping to tag the real 1.6.5 before I vanish for
59
+about a week and half later in the month. so in the worst case there may
60
+have to be a 1.6.6 to include the leftover topics.
61
62
--------------------------------------------------
63
[Graduated to "master"]
64
26
-* lt/approxidate (2009-08-30) 6 commits
27
- (merged to 'next' on 2009-08-30 at e016e3d)
28
- + fix approxidate parsing of relative months and years
29
- + tests: add date printing and parsing tests
30
- + refactor test-date interface
31
- + Add date formatting and parsing functions relative to a given time
32
- (merged to 'next' on 2009-08-26 at 62853f9)
33
- + Further 'approxidate' improvements
34
- + Improve on 'approxidate'
35
-
36
-* mr/gitweb-snapshot (2009-08-25) 3 commits
37
- (merged to 'next' on 2009-08-30 at e4edd0b)
38
- + gitweb: add t9501 tests for checking HTTP status codes
39
- + gitweb: split test suite into library and tests
40
- + gitweb: improve snapshot error handling
41
-
42
-* tf/diff-whitespace-incomplete-line (2009-08-23) 2 commits.
43
- (merged to 'next' on 2009-08-26 at 4fc7784)
44
- + xutils: Fix xdl_recmatch() on incomplete lines
45
- + xutils: Fix hashing an incomplete line with whitespaces at the end
65
+* np/maint-1.6.3-deepen (2009-09-03) 2 commits
66
+ (merged to 'next' on 2009-09-06 at d597fba)
67
+ + pack-objects: free preferred base memory after usage
68
+ + make shallow repository deepening more network efficient
69
+
70
+Further optimization and fixing of the object transfer around the "shallow
71
+repository" codepaths.
72
+
73
+* js/stash-dwim (2009-07-27) 1 commit.
74
+ (merged to 'next' on 2009-08-16 at 67896c4)
75
+ + Make 'git stash -k' a short form for 'git stash save --keep-index'
76
+ (this branch is used by tr/reset-checkout-patch.)
77
+
78
+* tr/reset-checkout-patch (2009-08-18) 10 commits.
79
+ (merged to 'next' on 2009-09-03 at d4f2ed7)
80
+ + stash: simplify defaulting to "save" and reject unknown options
81
+ (merged to 'next' on 2009-08-27 at d314281)
82
+ + Make test case number unique
83
+ (merged to 'next' on 2009-08-18 at e465bb3)
84
+ + tests: disable interactive hunk selection tests if perl is not available
85
+ (merged to 'next' on 2009-08-16 at 67896c4)
86
+ + DWIM 'git stash save -p' for 'git stash -p'
87
+ + Implement 'git stash save --patch'
88
+ + Implement 'git checkout --patch'
89
+ + Implement 'git reset --patch'
90
+ + builtin-add: refactor the meat of interactive_add()
91
+ + Add a small patch-mode testing library
92
+ + git-apply--interactive: Refactor patch mode code
93
+ (this branch uses js/stash-dwim.)
94
+
95
+DWIMmery of the two series tightened for safety a few days ago. Otherwise
96
+the same as what have been cooking in 'next' for a few weeks.
97
+
98
+* jc/upload-pack-hook (2009-08-28) 2 commits
99
+ (merged to 'next' on 2009-08-31 at f9933a5)
100
+ + upload-pack: feed "kind [clone|fetch]" to post-upload-pack hook
101
+ + upload-pack: add a trigger for post-upload-pack hook
102
+
103
+* jk/clone-b (2009-08-26) 1 commit
104
+ (merged to 'next' on 2009-08-30 at 10a68d1)
105
+ + clone: add --branch option to select a different HEAD
106
+
107
+* jc/mailinfo-scissors (2009-08-26) 5 commits
108
+ (merged to 'next' on 2009-08-30 at 5fc6248)
109
+ + mailinfo.scissors: new configuration
110
+ + am/mailinfo: Disable scissors processing by default
111
+ + Documentation: describe the scissors mark support of "git am"
112
+ + Teach mailinfo to ignore everything before -- >8 -- mark
113
+ + builtin-mailinfo.c: fix confusing internal API to mailinfo()
114
+
115
+I didn't pick up the patch to simplify the definition of scissors. I do
116
+not have strong opinion on it either way, and nobody on the list seemed to
117
+care too much either.
118
119
--------------------------------------------------
120
[New Topics]
121
50
-* pk/fast-import-tars (2009-09-03) 1 commit
51
- - import-tars: Allow per-tar author and commit message.
52
-
53
-* jc/maint-1.6.0-blank-at-eof (2009-09-04) 9 commits.
54
- - diff --color: color blank-at-eof
55
- - diff --whitespace=warn/error: fix blank-at-eof check
56
- - diff --whitespace=warn/error: obey blank-at-eof
57
- - diff.c: the builtin_diff() deals with only two-file comparison
58
- - apply --whitespace: warn blank but not necessarily empty lines at EOF
59
- - apply --whitespace=warn/error: diagnose blank at EOF
60
- - apply.c: split check_whitespace() into two
61
- - apply --whitespace=fix: detect new blank lines at eof correctly
62
- - apply --whitespace=fix: fix handling of blank lines at the eof
122
+* cb/maint-1.6.3-grep-relative-up (2009-09-05) 2 commits.
123
+ (merged to 'next' on 2009-09-07 at f9b5b48)
124
+ + grep: accept relative paths outside current working directory
125
+ + grep: fix exit status if external_grep() punts
126
+
127
+1.6.5 material.
128
+
129
+* jk/1.7.0-status (2009-09-05) 5 commits
130
+ - docs: note that status configuration affects only long format
131
+ (merged to 'next' on 2009-09-07 at 8a7c563)
132
+ + commit: support alternate status formats
133
+ + status: add --porcelain output format
134
+ + status: refactor format option parsing
135
+ + status: refactor short-mode printing to its own function
136
+ (this branch uses jc/1.7.0-status.)
137
+
138
+Gives the --short output format to post 1.7.0 "git commit --dry-run" that
139
+is similar to that of post 1.7.0 "git status".
140
+
141
+It might be a good idea to make the --short format part of 1.6.6 without
142
+waiting for 1.7.0; it would require some branch shuffling to bring the
143
+short-status patch earlier than the one that makes "status" different from
144
+"commit --dry-run", though.
145
+
146
+* jk/unwanted-advices (2009-09-06) 2 commits
147
+ - status: make "how to stage" messages optional
148
+ - push: make non-fast-forward help message configurable
149
+
150
+Post 1.3.0 "newbie friendliness drive" introduced a lot of "helpful"
151
+advice messages that are sometimes found to be irritating once the user
152
+got used to working with git. This series introduces a general mechanism
153
+to allow us to make them less loud using the config mechanism. The config
154
+option hierarchy in these two are "message.*" but it is planned to be
155
+renamed to "advice.*".
156
+
157
+* jn/gitweb-show-size (2009-09-07) 1 commit
158
+ - gitweb: Add 'show-sizes' feature to show blob sizes in tree view
159
+
160
+* jt/pushinsteadof (2009-09-07) 2 commits
161
+ - Add url.<base>.pushInsteadOf: URL rewriting for push only
162
+ - Wrap rewrite globals in a struct in preparation for adding another set
163
+
164
+Probably 1.6.5 material.
165
+
166
+* lt/maint-traverse-trees-fix (2009-09-06) 1 commit.
167
+ - Prepare 'traverse_trees()' for D/F conflict lookahead
168
+
169
+Beginning of the fix to a rather nasty longstanding issue of merging trees
170
+with ("a" "a-b"), ("a/b" "a-b") and just ("a-b").
171
+
172
+* ne/rev-cache (2009-09-07) 7 commits
173
+ . support for commit grafts, slight change to general mechanism
174
+ . support for path name caching in rev-cache
175
+ . full integration of rev-cache into git, completed test suite
176
+ . administrative functions for rev-cache, start of integration into git
177
+ . support for non-commit object caching in rev-cache
178
+ . basic revision cache system, no integration or features
179
+ . man page and technical discussion for rev-cache
180
+
181
+Replaced but I do not think this is ready for 'pu' yet.
182
183
--------------------------------------------------
184
[Stalled]
@@ -81,8 +200,15 @@ cooking in 'next'.
200
I heard the cvs-helper series depends on this one. It seems that the
201
fan-out strategy is being rethought?
202
84
---------------------------------------------------
85
-[Cooking]
203
+* je/send-email-no-subject (2009-08-05) 1 commit
204
+ (merged to 'next' on 2009-08-30 at b6455c2)
205
+ + send-email: confirm on empty mail subjects
206
+
207
+The existing tests to covers the positive case (i.e. as long as the user
208
+says "yes" to the "do you really want to send this message that lacks
209
+subject", the message is sent) of this feature, but the feature itself
210
+needs its own test to verify the negative case (i.e. does it correctly
211
+stop if the user says "no"?)
212
213
* db/vcs-helper (2009-09-03) 16 commits
214
- Allow helpers to report in "list" command that the ref is unchanged
@@ -104,7 +230,11 @@ fan-out strategy is being rethought?
230
+ git-http-fetch: not a builtin
231
+ Use an external program to implement fetching with curl
232
+ Add support for external programs for handling native fetches
107
- (this branch is used by jh/cvs-helper)
233
+ (this branch is used by jh/cvs-helper.)
234
+
235
+I'd really want to have this in 1.6.5 so that we can eject -lcurl from the
236
+main "git" binary. The patches in 'pu' got some review comments, and I
237
+thought Daniel's responses were sensible. Comments?
238
239
* jh/cvs-helper (2009-08-18) 8 commits
240
- More fixes to the git-remote-cvs installation procedure
@@ -115,82 +245,10 @@ fan-out strategy is being rethought?
245
- 1/2: Add Python support library for CVS remote helper
246
- Basic build infrastructure for Python scripts
247
- Allow helpers to request marks for fast-import
118
- (this branch uses db/vcs-helper)
248
+ (this branch uses db/vcs-helper.)
249
250
Builds on db/vcs-helper. There is a re-roll planned.
251
122
-* pk/fast-import-dirs (2009-09-03) 1 commit
123
- - Add script for importing bits-and-pieces to Git.
124
-
125
-With an updated key-value quoting rules, which I haven't carefully looked
126
-at. Aren't there standard libraries to do this sort of thing without us
127
-having to hand-roll these logic every time, I have to wonder...
128
-
129
-* jn/gitweb-blame (2009-09-01) 5 commits
130
- - gitweb: Minify gitweb.js if JSMIN is defined
131
- - gitweb: Create links leading to 'blame_incremental' using JavaScript
132
- - gitweb: Colorize 'blame_incremental' view during processing
133
- - gitweb: Incremental blame (using JavaScript)
134
- - gitweb: Add optional "time to generate page" info in footer
135
-
136
-Ajax-y blame. The first part should be in 'next' shortly.
137
-
138
-* js/stash-dwim (2009-07-27) 1 commit.
139
- (merged to 'next' on 2009-08-16 at 67896c4)
140
- + Make 'git stash -k' a short form for 'git stash save --keep-index'
141
- (this branch is used by tr/reset-checkout-patch.)
142
-
143
-* tr/reset-checkout-patch (2009-08-18) 10 commits.
144
- (merged to 'next' on 2009-09-03 at d4f2ed7)
145
- + stash: simplify defaulting to "save" and reject unknown options
146
- (merged to 'next' on 2009-08-27 at d314281)
147
- + Make test case number unique
148
- (merged to 'next' on 2009-08-18 at e465bb3)
149
- + tests: disable interactive hunk selection tests if perl is not available
150
- (merged to 'next' on 2009-08-16 at 67896c4)
151
- + DWIM 'git stash save -p' for 'git stash -p'
152
- + Implement 'git stash save --patch'
153
- + Implement 'git checkout --patch'
154
- + Implement 'git reset --patch'
155
- + builtin-add: refactor the meat of interactive_add()
156
- + Add a small patch-mode testing library
157
- + git-apply--interactive: Refactor patch mode code
158
- (this branch uses js/stash-dwim.)
159
-
160
-DWIMmery of the two series tightened for safety. This should be ready for
161
-1.6.5.
162
-
163
-* jc/upload-pack-hook (2009-08-28) 2 commits
164
- (merged to 'next' on 2009-08-31 at f9933a5)
165
- + upload-pack: feed "kind [clone|fetch]" to post-upload-pack hook
166
- + upload-pack: add a trigger for post-upload-pack hook
167
-
168
-* jk/clone-b (2009-08-26) 1 commit
169
- (merged to 'next' on 2009-08-30 at 10a68d1)
170
- + clone: add --branch option to select a different HEAD
171
-
172
-* je/send-email-no-subject (2009-08-05) 1 commit
173
- (merged to 'next' on 2009-08-30 at b6455c2)
174
- + send-email: confirm on empty mail subjects
175
-
176
-The existing tests to covers the positive case (i.e. as long as the user
177
-says "yes" to the "do you really want to send this message that lacks
178
-subject", the message is sent) of this feature, but the feature itself
179
-needs its own test to verify the negative case (i.e. does it correctly
180
-stop if the user says "no"?)
181
-
182
-* jc/mailinfo-scissors (2009-08-26) 5 commits
183
- (merged to 'next' on 2009-08-30 at 5fc6248)
184
- + mailinfo.scissors: new configuration
185
- + am/mailinfo: Disable scissors processing by default
186
- + Documentation: describe the scissors mark support of "git am"
187
- + Teach mailinfo to ignore everything before -- >8 -- mark
188
- + builtin-mailinfo.c: fix confusing internal API to mailinfo()
189
-
190
-I didn't pick up the patch to simplify the definition of scissors. I do
191
-not have strong opinion on it either way, but the list would hopefully
192
-decide it before too long.
193
-
252
* cc/sequencer-rebase-i (2009-08-28) 15 commits
253
- rebase -i: use "git sequencer--helper --cherry-pick"
254
- sequencer: add "--cherry-pick" option to "git sequencer--helper"
@@ -209,16 +267,59 @@ decide it before too long.
267
- sequencer: add "builtin-sequencer--helper.c"
268
269
Migrating "rebase -i" bit by bit to C.
270
+Not a 1.6.5 material yet.
271
+
272
+--------------------------------------------------
273
+[Cooking]
274
+
275
+* pk/fast-import-tars (2009-09-03) 1 commit
276
+ (merged to 'next' on 2009-09-07 at 8fbf027)
277
+ + import-tars: Allow per-tar author and commit message.
278
+
279
+* jc/maint-1.6.0-blank-at-eof (2009-09-05) 10 commits.
280
+ (merged to 'next' on 2009-09-07 at 165dc3c)
281
+ + core.whitespace: split trailing-space into blank-at-{eol,eof}
282
+ + diff --color: color blank-at-eof
283
+ + diff --whitespace=warn/error: fix blank-at-eof check
284
+ + diff --whitespace=warn/error: obey blank-at-eof
285
+ + diff.c: the builtin_diff() deals with only two-file comparison
286
+ + apply --whitespace: warn blank but not necessarily empty lines at EOF
287
+ + apply --whitespace=warn/error: diagnose blank at EOF
288
+ + apply.c: split check_whitespace() into two
289
+ + apply --whitespace=fix: detect new blank lines at eof correctly
290
+ + apply --whitespace=fix: fix handling of blank lines at the eof
291
+
292
+Hopefully in 1.6.5
293
213
-* sr/gfi-options (2009-09-02) 6 commits
214
- - fast-import: test the new option command
215
- - fast-import: add option command
216
- - fast-import: test the new feature command
217
- - fast-import: add feature command
218
- - fast-import: put marks reading in it's own function
219
- - fast-import: put option parsing code in separate functions
294
+* pk/fast-import-dirs (2009-09-03) 1 commit
295
+ (merged to 'next' on 2009-09-07 at 836cba2)
296
+ + Add script for importing bits-and-pieces to Git.
297
+
298
+With an updated key-value quoting rules, which I haven't carefully looked
299
+at. I have to wonder if there are standard libraries to do this sort of
300
+thing without having to hand-roll these logic every time. But perhaps I
301
+shouldn't be too picky on contrib/ material.
302
221
-Re-rolled again.
303
+* jn/gitweb-blame (2009-09-01) 5 commits
304
+ - gitweb: Minify gitweb.js if JSMIN is defined
305
+ - gitweb: Create links leading to 'blame_incremental' using JavaScript
306
+ (merged to 'next' on 2009-09-07 at 3622199)
307
+ + gitweb: Colorize 'blame_incremental' view during processing
308
+ + gitweb: Incremental blame (using JavaScript)
309
+ + gitweb: Add optional "time to generate page" info in footer
310
+
311
+Ajax-y blame.
312
+
313
+* sr/gfi-options (2009-09-06) 6 commits
314
+ (merged to 'next' on 2009-09-07 at 5f6b0ff)
315
+ + fast-import: test the new option command
316
+ + fast-import: add option command
317
+ + fast-import: test the new feature command
318
+ + fast-import: add feature command
319
+ + fast-import: put marks reading in it's own function
320
+ + fast-import: put option parsing code in separate functions
321
+
322
+Perhaps 1.6.5 material but I wasn't sure.
323
324
* nd/sparse (2009-08-20) 19 commits
325
- sparse checkout: inhibit empty worktree
@@ -244,13 +345,16 @@ Re-rolled again.
345
--------------------------------------------------
346
[For 1.7.0]
347
247
-* jc/1.7.0-status (2009-08-15) 3 commits
348
+* jc/1.7.0-status (2009-09-05) 4 commits
349
+ (merged to 'next' on 2009-09-06 at 19d4beb)
350
+ + status: typo fix in usage
351
(merged to 'next' on 2009-08-22 at b3507bb)
352
+ git status: not "commit --dry-run" anymore
353
+ git stat -s: short status output
354
+ git stat: the beginning of "status that is not a dry-run of commit"
355
+ (this branch is used by jk/1.7.0-status.)
356
253
-With this, "git status" is no longer "git commit --preview".
357
+With this, "git status" is no longer "git commit --dry-run".
358
359
* jc/1.7.0-send-email-no-thread-default (2009-08-22) 1 commit
360
(merged to 'next' on 2009-08-22 at 5106de8)
@@ -287,40 +391,7 @@ Maybe some people care about this. I dunno.
391
392
Maybe some people care about this. I dunno.
393
290
-* ar/maint-1.6.2-merge-recursive-d-f (2009-05-11) 2 commits.
291
- . Fix for a merge where a branch has an F->D transition
292
- . Add a reminder test case for a merge with F/D transition
293
-
294
-* jc/merge-convert (2009-01-26) 1 commit.
295
- . git-merge-file: allow converting the results for the work tree
296
-
394
* lt/read-directory (2009-05-15) 3 commits.
395
. Add initial support for pathname conversion to UTF-8
396
. read_directory(): infrastructure for pathname character set conversion
397
. Add 'fill_directory()' helper function for directory traversal
301
-
302
-* ps/blame (2009-03-12) 1 commit.
303
- . blame.c: start libifying the blame infrastructure
304
-
305
-* pb/tracking (2009-07-16) 7 commits.
306
- . branch.c: if remote is not config'd for branch, don't try delete push config
307
- . branch, checkout: introduce autosetuppush
308
- . move deletion of merge configuration to branch.c
309
- . remote: add per-remote autosetupmerge and autosetuprebase configuration
310
- . introduce a struct tracking_config
311
- . branch: install_branch_config and struct tracking refactoring
312
- . config: allow false and true values for branch.autosetuprebase
313
-
314
-Has been ejected from 'pu' for some time, expecting a reroll.
315
-
316
-* ne/rev-cache (2009-08-21) 6 commits
317
- . support for path name caching in rev-cache
318
- . full integration of rev-cache into git, completed test suite
319
- . administrative functions for rev-cache, start of integration into git
320
- . support for non-commit object caching in rev-cache
321
- . basic revision cache system, no integration or features
322
- . man page and technical discussion for rev-cache
323
-
324
-Updated but seems to break upload-pack tests when merged to 'pu'; given
325
-what this series touches, breakages in that area are expected.
326
-May discard if a working reroll comes, to give it a fresh start.