What's cooking (2016/10 #07)
Junio C Hamano committed
Oct 26, 2016 at 15:27 UTC
1ba9a6af1862ac04e0cbe006b2dbcaa6cd1b1d5b
1 file changed
+427
-498
whats-cooking.txt
+427
-498
@@ -1,10 +1,10 @@
1
To: git@vger.kernel.org
2
Bcc: lwn@lwn.net
3
-Subject: What's cooking in git.git (Oct 2016, #06; Mon, 24)
4
-X-master-at: 659889482ac63411daea38b2c3d127842ea04e4d
5
-X-next-at: 5073a4de2dac30ca7cfaaa2d4bdd8eead108e4ca
3
+Subject: What's cooking in git.git (Oct 2016, #07; Wed, 26)
4
+X-master-at: 2cc2e70264e0fcba04f9ef791d144bbc8b501206
5
+X-next-at: d2da68a14ada34b7842a51c88751ffe4aaf8a1af
6
7
-What's cooking in git.git (Oct 2016, #06; Mon, 24)
7
+What's cooking in git.git (Oct 2016, #07; Wed, 26)
8
--------------------------------------------------
9
10
Here are the topics that have been cooking. Commits prefixed with
@@ -12,64 +12,138 @@ 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
-Originally I planed to start concluding this cycle today, but
16
-waiting for the conclusion of a few test breakages on Windows, I
17
-didn't tag -rc0 today.
18
-
19
-Here are my current thinking on the notable topics.
20
-
21
- - the "off-by-one fix" part of sb/submodule-ignore-trailing-slash
22
- needs to be in the upcoming release but the "trailing /. in base
23
- should not affect the resolution of ../relative/path" part that
24
- is still under discussion can wait. Which means we'd need a few
25
- more !MINGW prerequisites in the tests by -rc0.
26
-
27
- - js/prepare-sequencer topic is not yet in 'next' but it would be a
28
- nice-to-have in the upcoming release if we can. It does not yet
29
- touch "rebase -i", but does touch the sequencer code that is used
30
- in cherry-pick and revert, so I'd prefer to cook it for at least
31
- a week and half in 'next' before merging.
32
-
33
- - ls/filter-process topic has been in 'next' with one known test
34
- breakage on Windows, whose resolution ls/git-open-cloexec is
35
- close to its final shape. Perhaps we can cook them for at least
36
- a week and half in 'next' and have it in the upcoming release.
37
-
38
- - ex/deprecate-empty-pathspec-as-match-all topic that makes it
39
- illegal to say 'git add ""' when you mean 'git add .' has been in
40
- 'next' for more than a cycle. I am inclined to merge it in the
41
- upcoming release.
42
-
43
- - jc/merge-drop-old-syntax is relatively new in 'next' after all
44
- known in-tree dependents have been updated. I am planning to
45
- keep it cooking in 'next' but add a backward incompatibility
46
- notice to the release notes to the upcoming release.
47
-
48
- - lt/abbrev-auto and its follow-up jk/abbrev-auto are about auto
49
- scaling the default abbreviation length when Git produces a short
50
- object name to adjust to the modern times. Peff noticed one
51
- fallout from it recently and its fix jc/abbrev-auto is not yet in
52
- 'next'. I would not be surprised if there are other uncovered
53
- fallouts remaining in the code, but at the same time, I expect
54
- they are all cosmetic kind that do not affect correctness, so I
55
- am inclined to include all of them in the upcoming release.
56
-
57
-I plan to merge other smallish topics that have been in 'next' to
58
-'master' soonish, and relabel the remainder that have been labeled
59
-as "Will merge to 'master'" to "Will hold" and keep cooking them in
60
-'next'. For this reason, please do not take the "Will merge to
61
-'master'" label too literally in this issue of "What's cooking"
62
-report. It is always true that the label only means "the topic will
63
-be in 'master' eventually", not "the topic will be in the upcoming
64
-release", but in this issue that is even more true than usual.
65
-
15
You can find the changes described here in the integration branches
16
of the repositories listed at
17
18
http://git-blame.blogspot.com/p/git-public-repositories.html
19
20
--------------------------------------------------
72
-[New Topics]
21
+[Graduated to "master"]
22
+
23
+* ab/gitweb-abbrev-links (2016-10-14) 3 commits
24
+ (merged to 'next' on 2016-10-17 at 4868def05e)
25
+ + gitweb: link to "git describe"'d commits in log messages
26
+ + gitweb: link to 7-char+ SHA-1s, not only 8-char+
27
+ + gitweb: fix a typo in a comment
28
+
29
+ In addition to purely abbreviated commit object names, "gitweb"
30
+ learned to turn "git describe" output (e.g. v2.9.3-599-g2376d31787)
31
+ into clickable links in its output.
32
+
33
+
34
+* bw/ls-files-recurse-submodules (2016-10-10) 4 commits
35
+ (merged to 'next' on 2016-10-17 at f0e398946a)
36
+ + ls-files: add pathspec matching for submodules
37
+ + ls-files: pass through safe options for --recurse-submodules
38
+ + ls-files: optionally recurse into submodules
39
+ + git: make super-prefix option
40
+
41
+ "git ls-files" learned "--recurse-submodules" option that can be
42
+ used to get a listing of tracked files across submodules (i.e. this
43
+ only works with "--cached" option, not for listing untracked or
44
+ ignored files). This would be a useful tool to sit on the upstream
45
+ side of a pipe that is read with xargs to work on all working tree
46
+ files from the top-level superproject.
47
+
48
+
49
+* bw/submodule-branch-dot-doc (2016-10-19) 1 commit
50
+ (merged to 'next' on 2016-10-21 at 18aad25ba8)
51
+ + submodules doc: update documentation for "." used for submodule branches
52
+
53
+ Recent git allows submodule.<name>.branch to use a special token
54
+ "." instead of the branch name; the documentation has been updated
55
+ to describe it.
56
+
57
+
58
+* dk/worktree-dup-checkout-with-bare-is-ok (2016-10-14) 1 commit
59
+ (merged to 'next' on 2016-10-17 at 24594d3e56)
60
+ + worktree: allow the main brach of a bare repository to be checked out
61
+
62
+ In a worktree connected to a repository elsewhere, created via "git
63
+ worktree", "git checkout" attempts to protect users from confusion
64
+ by refusing to check out a branch that is already checked out in
65
+ another worktree. However, this also prevented checking out a
66
+ branch, which is designated as the primary branch of a bare
67
+ reopsitory, in a worktree that is connected to the bare
68
+ repository. The check has been corrected to allow it.
69
+
70
+
71
+* ex/deprecate-empty-pathspec-as-match-all (2016-06-22) 1 commit
72
+ (merged to 'next' on 2016-09-21 at e19148ea63)
73
+ + pathspec: warn on empty strings as pathspec
74
+
75
+ Originally merged to 'next' on 2016-07-13
76
+
77
+ An empty string used as a pathspec element has always meant
78
+ 'everything matches', but it is too easy to write a script that
79
+ finds a path to remove in $path and run 'git rm "$paht"', which
80
+ ends up removing everything. Start warning about this use of an
81
+ empty string used for 'everything matches' and ask users to use a
82
+ more explicit '.' for that instead.
83
+
84
+ The hope is that existing users will not mind this change, and
85
+ eventually the warning can be turned into a hard error, upgrading
86
+ the deprecation into removal of this (mis)feature.
87
+
88
+
89
+* jc/cocci-xstrdup-or-null (2016-10-12) 1 commit
90
+ (merged to 'next' on 2016-10-17 at 55ceaa465a)
91
+ + cocci: refactor common patterns to use xstrdup_or_null()
92
+
93
+ Code cleanup.
94
+
95
+
96
+* jc/diff-unique-abbrev-comments (2016-09-30) 1 commit
97
+ (merged to 'next' on 2016-10-17 at c7fb286102)
98
+ + diff_unique_abbrev(): document its assumption and limitation
99
+ (this branch is used by jk/no-looking-at-dotgit-outside-repo.)
100
+
101
+ A bit more comments in a tricky code.
102
+
103
+
104
+* jc/ws-error-highlight (2016-10-04) 4 commits
105
+ (merged to 'next' on 2016-10-17 at ecbdc57d77)
106
+ + diff: introduce diff.wsErrorHighlight option
107
+ + diff.c: move ws-error-highlight parsing helpers up
108
+ + diff.c: refactor parse_ws_error_highlight()
109
+ + t4015: split out the "setup" part of ws-error-highlight test
110
+
111
+ "git diff/log --ws-error-highlight=<kind>" lacked the corresponding
112
+ configuration variable to set it by default.
113
+
114
+
115
+* jk/ambiguous-short-object-names (2016-10-12) 1 commit
116
+ (merged to 'next' on 2016-10-19 at e7c55a9da5)
117
+ + t1512: become resilient to GETTEXT_POISON build
118
+
119
+ A test fixup to recently graduated topic.
120
+
121
+
122
+* jk/diff-submodule-diff-inline (2016-10-20) 1 commit
123
+ (merged to 'next' on 2016-10-21 at 13f300805e)
124
+ + rev-list: use hdr_termination instead of a always using a newline
125
+
126
+ A recently graduated topic regressed "git rev-list --header"
127
+ output, breaking "gitweb". This has been fixed.
128
+
129
+
130
+* jk/fetch-quick-tag-following (2016-10-14) 1 commit
131
+ (merged to 'next' on 2016-10-19 at d7718dcdf5)
132
+ + fetch: use "quick" has_sha1_file for tag following
133
+
134
+ When fetching from a remote that has many tags that are irrelevant
135
+ to branches we are following, we used to waste way too many cycles
136
+ when checking if the object pointed at by a tag (that we are not
137
+ going to fetch!) exists in our repository too carefully.
138
+
139
+
140
+* jk/merge-base-fork-point-without-reflog (2016-10-12) 1 commit
141
+ (merged to 'next' on 2016-10-19 at 00a6797f62)
142
+ + merge-base: handle --fork-point without reflog
143
+
144
+ "git rebase" immediately after "git clone" failed to find the fork
145
+ point from the upstream.
146
+
147
148
* jk/tap-verbose-fix (2016-10-24) 4 commits
149
(merged to 'next' on 2016-10-24 at 5073a4de2d)
@@ -86,10 +160,80 @@ of the repositories listed at
160
new mode to run our tests in the test harness to send the verbose
161
output separately to the log file.
162
89
- Will merge to 'master'.
163
164
+* jk/tighten-alloc (2016-10-17) 2 commits
165
+ (merged to 'next' on 2016-10-19 at 548522a520)
166
+ + inline xalloc_flex() into FLEXPTR_ALLOC_MEM
167
+ + avoid pointer arithmetic involving NULL in FLEX_ALLOC_MEM
168
+
169
+ Protect our code from over-eager compilers.
170
+
171
+
172
+* jk/upload-pack-use-prio-queue (2016-10-11) 1 commit
173
+ (merged to 'next' on 2016-10-19 at 1d6efb07ac)
174
+ + upload-pack: use priority queue in reachable() check
175
+
176
+ "git upload-pack" had its code cleaned-up and performance improved
177
+ by reducing use of timestamp-ordered commit-list, which was
178
+ replaced with a priority queue.
179
+
180
+
181
+* js/libify-require-clean-work-tree (2016-10-07) 6 commits
182
+ (merged to 'next' on 2016-10-17 at f5c20df38b)
183
+ + wt-status: begin error messages with lower-case
184
+ + wt-status: teach has_{unstaged,uncommitted}_changes() about submodules
185
+ + wt-status: export also the has_un{staged,committed}_changes() functions
186
+ + wt-status: make the require_clean_work_tree() function reusable
187
+ + pull: make code more similar to the shell script again
188
+ + pull: drop confusing prefix parameter of die_on_unclean_work_tree()
189
+
190
+ The require_clean_work_tree() helper was recreated in C when "git
191
+ pull" was rewritten from shell; the helper is now made available to
192
+ other callers in preparation for upcoming "rebase -i" work.
193
+
194
+
195
+* mg/gpg-richer-status (2016-10-12) 1 commit
196
+ (merged to 'next' on 2016-10-17 at 8843a6a8be)
197
+ + gpg-interface: use more status letters
198
+
199
+ The GPG verification status shown in "%G?" pretty format specifier
200
+ was not rich enough to differentiate a signature made by an expired
201
+ key, a signature made by a revoked key, etc. New output letters
202
+ have been assigned to express them.
203
92
-* po/fix-doc-merge-base-illustration (2016-10-21) 1 commit
204
+
205
+* mm/credential-libsecret (2016-10-11) 1 commit
206
+ (merged to 'next' on 2016-10-17 at 1b4af03ba4)
207
+ + contrib: add credential helper for libsecret
208
+
209
+ A new credential helper that talks via "libsecret" with
210
+ implementations of XDG Secret Service API has been added to
211
+ contrib/credential/.
212
+
213
+
214
+* mm/send-email-cc-cruft-after-address (2016-10-21) 3 commits
215
+ (merged to 'next' on 2016-10-21 at c7ec2b5025)
216
+ + Git.pm: add comment pointing to t9000
217
+ + t9000-addresses: update expected results after fix
218
+ (merged to 'next' on 2016-10-19 at 41e3f876cd)
219
+ + parse_mailboxes: accept extra text after <...> address
220
+
221
+ "git send-email" attempts to pick up valid e-mails from the
222
+ trailers, but people in real world write non-addresses there, like
223
+ "Cc: Stable <add@re.ss> # 4.8+", which broke the output depending
224
+ on the availability and vintage of Mail::Address perl module.
225
+
226
+
227
+* pb/test-parse-options-expect (2016-10-17) 1 commit
228
+ (merged to 'next' on 2016-10-19 at d3517d592f)
229
+ + t0040: convert all possible tests to use `test-parse-options --expect`
230
+
231
+ Test clean-up.
232
+
233
+
234
+* po/fix-doc-merge-base-illustration (2016-10-24) 2 commits
235
+ (merged to 'next' on 2016-10-24 at 6539e97fcf)
236
+ + doc: fix the 'revert a faulty merge' ASCII art tab spacing
237
(merged to 'next' on 2016-10-21 at ac6f04a6c5)
238
+ doc: fix merge-base ASCII art tab spacing
239
@@ -97,84 +241,139 @@ of the repositories listed at
241
tabs in it. Adjust a few of them in merge-base documentation to
242
work around them.
243
100
- Will merge to 'master'.
244
245
+* pt/gitgui-updates (2016-10-20) 35 commits
246
+ (merged to 'next' on 2016-10-21 at 4c8214095a)
247
+ + Merge tag 'gitgui-0.21.0' of git://repo.or.cz/git-gui
248
+ + git-gui: set version 0.21
249
+ + Merge branch 'as/bulgarian' into pu
250
+ + git-gui: Mark 'All' in remote.tcl for translation
251
+ + git-gui i18n: Updated Bulgarian translation (565,0f,0u)
252
+ + Merge branch 'os/preserve-author' into pu
253
+ + git-gui: avoid persisting modified author identity
254
+ + git-gui: Do not reset author details on amend
255
+ + Merge branch 'kb/unicode' into pu
256
+ + git-gui: handle the encoding of Git's output correctly
257
+ + git-gui: unicode file name support on windows
258
+ + Merge branch 'dr/ru' into pu
259
+ + git-gui: Update Russian translation
260
+ + git-gui: maintain backwards compatibility for merge syntax
261
+ + Merge branch 'va/i18n_2' into pu
262
+ + git-gui i18n: mark string in lib/error.tcl for translation
263
+ + git-gui: fix incorrect use of Tcl append command
264
+ + git-gui i18n: mark "usage:" strings for translation
265
+ + git-gui i18n: internationalize use of colon punctuation
266
+ + Merge branch 'pt/non-mouse-usage' into pu
267
+ + Amend tab ordering and text widget border and highlighting.
268
+ + Allow keyboard control to work in the staging widgets.
269
+ + Merge branch 'pt/git4win-mods' into pu
270
+ + git-gui (Windows): use git-gui.exe in `Create Desktop Shortcut`
271
+ + git-gui: fix detection of Cygwin
272
+ + Merge branch 'patches' into pu
273
+ + git-gui: ensure the file in the diff pane is in the list of selected files
274
+ + git-gui: support for $FILENAMES in tool definitions
275
+ + git-gui: fix initial git gui message encoding
276
+ + git-gui/po/glossary/txt-to-pot.sh: use the $( ... ) construct for command substitution
277
+ + Merge branch 'va/i18n' into pu
278
+ + Merge branch 'rs/use-modern-git-merge-syntax' into pu
279
+ + Merge branch 'js/commit-gpgsign' into pu
280
+ + Merge branch 'sy/i18n' into pu
281
+ + git-gui: sort entries in tclIndex
282
103
-* jc/abbrev-auto (2016-10-22) 4 commits
104
- - transport: compute summary-width dynamically
105
- - transport: allow summary-width to be computed dynamically
106
- - fetch: pass summary_width down the callchain
107
- - transport: pass summary_width down the callchain
108
- (this branch uses jk/abbrev-auto and lt/abbrev-auto.)
283
+ A new version of git-gui, now at its 0.21.0 tag.
284
110
- "git push" and "git fetch" reports from what old object to what new
111
- object each ref was updated, using abbreviated refnames, and they
112
- attempt to align the columns for this and other pieces of
113
- information. The way these codepaths compute how many display
114
- columns to allocate for the object names portion of this output has
115
- been updated to match the recent "auto scale the default
116
- abbreviation length" change.
285
118
- Will merge to 'next'.
286
+* tg/add-chmod+x-fix (2016-10-20) 1 commit
287
+ (merged to 'next' on 2016-10-21 at 1585ac7139)
288
+ + t3700: fix broken test under !SANITY
289
290
+ A hot-fix for a test added by a recent topic that went to both
291
+ 'master' and 'maint' already.
292
121
-* jc/reset-unmerge (2016-10-24) 1 commit
122
- - reset: --unmerge
293
124
- After "git add" is run prematurely during a conflict resolution,
125
- "git diff" can no longer be used as a way to sanity check by
126
- looking at the combined diff. "git reset" learned a new
127
- "--unmerge" option to recover from this situation.
294
+* va/i18n (2016-10-17) 7 commits
295
+ (merged to 'next' on 2016-10-19 at b7d733698b)
296
+ + i18n: diff: mark warnings for translation
297
+ + i18n: credential-cache--daemon: mark advice for translation
298
+ + i18n: convert mark error messages for translation
299
+ + i18n: apply: mark error message for translation
300
+ + i18n: apply: mark error messages for translation
301
+ + i18n: apply: mark info messages for translation
302
+ + i18n: apply: mark plural string for translation
303
304
+ More i18n.
305
306
131
-* jk/daemon-path-ok-check-truncation (2016-10-24) 1 commit
132
- - daemon: detect and reject too-long paths
307
+* yk/git-tag-remove-mention-of-old-layout-in-doc (2016-10-20) 1 commit
308
+ (merged to 'next' on 2016-10-21 at 8d9e23b023)
309
+ + doc: remove reference to the traditional layout in git-tag.txt
310
134
- "git daemon" used fixed-length buffers to turn URL to the
135
- repository the client asked for into the server side directory
136
- path, using snprintf() to avoid overflowing these buffers, but
137
- allowed possibly truncated paths to the directory. This has been
138
- tightened to reject such a request that causes overlong path to be
139
- required to serve.
311
+ Shorten description of auto-following in "git tag" by removing a
312
+ mention of historical remotes layout which is not relevant to the
313
+ main topic.
314
141
- Will merge to 'next'.
315
+--------------------------------------------------
316
+[New Topics]
317
318
+* aw/numbered-stash (2016-10-26) 1 commit
319
+ (merged to 'next' on 2016-10-26 at 8d9325fa3a)
320
+ + stash: allow stashes to be referenced by index only
321
144
-* ls/git-open-cloexec (2016-10-24) 3 commits
145
- - SQUASH???
146
- - read-cache: make sure file handles are not inherited by child processes
147
- - sha1_file: open window into packfiles with CLOEXEC
322
+ The user always has to say "stash@{$N}" when naming a single
323
+ element in the default location of the stash, i.e. reflogs in
324
+ refs/stash. The "git stash" command learned to accept "git stash
325
+ apply 4" as a short-hand for "git stash apply stash@{4}".
326
149
- Git generally does not explicitly close file descriptors that were
150
- open in the parent process when spawning a child process, but most
151
- of the time the child does not want to access them. As Windows does
152
- not allow removing or renaming a file that has a file descriptor
153
- open, a slow-to-exit child can even break the parent process by
154
- holding onto them. Use O_CLOEXEC flag to open files in various
155
- codepaths.
327
+ Will merge to 'master'.
328
157
- Under discussion.
158
- cf. <20161024183900.GA12769@starla>
329
160
- This needs to be merged before ls/filter-process so that it won't
161
- break Windows.
330
+* jk/no-looking-at-dotgit-outside-repo-final (2016-10-26) 1 commit
331
+ (merged to 'next' on 2016-10-26 at 220e160451)
332
+ + setup_git_env: avoid blind fall-back to ".git"
333
+ (this branch uses jk/no-looking-at-dotgit-outside-repo.)
334
335
+ This is the endgame of the topic to avoid blindly falling back to
336
+ ".git" when the setup sequence said we are _not_ in Git repository.
337
+ A corner case that happens to work right now may be broken by a
338
+ call to die("BUG").
339
164
-* rs/ring-buffer-wraparound (2016-10-24) 1 commit
165
- - hex: make wraparound of the index into ring-buffer explicit
340
+ Will cook in 'next'.
341
167
- The code that we have used for the past 10+ years to cycle
168
- 4-element ring buffers turns out to be not quite portable in
169
- theoretical world.
342
171
- Under discussion.
172
- cf. <b1f9054e-fadb-c2d3-bf95-00e88e1fb85b@web.de>
343
+* ew/svn-wt (2016-10-14) 2 commits
344
+ - git-svn: "git worktree" awareness
345
+ - git-svn: reduce scope of input record separator change
346
+
347
+ Will replace with a direct pulling from Eric to 'master'.
348
174
-
349
--------------------------------------------------
350
[Stalled]
351
352
+* hv/submodule-not-yet-pushed-fix (2016-10-10) 3 commits
353
+ - batch check whether submodule needs pushing into one call
354
+ - serialize collection of refs that contain submodule changes
355
+ - serialize collection of changed submodules
356
+
357
+ The code in "git push" to compute if any commit being pushed in the
358
+ superproject binds a commit in a submodule that hasn't been pushed
359
+ out was overly inefficient, making it unusable even for a small
360
+ project that does not have any submodule but have a reasonable
361
+ number of refs.
362
+
363
+ Waiting for review.
364
+ cf. <cover.1475851621.git.hvoigt@hvoigt.net>
365
+
366
+
367
+* sb/push-make-submodule-check-the-default (2016-10-10) 2 commits
368
+ - push: change submodule default to check when submodules exist
369
+ - submodule add: extend force flag to add existing repos
370
+
371
+ Turn the default of "push.recurseSubmodules" to "check" when
372
+ submodules seem to be in use.
373
+
374
+ Will hold to wait for hv/submodule-not-yet-pushed-fix
375
+
376
+
377
* jc/bundle (2016-03-03) 6 commits
378
- index-pack: --clone-bundle option
379
- Merge branch 'jc/index-pack' into jc/bundle
@@ -288,144 +487,103 @@ of the repositories listed at
487
--------------------------------------------------
488
[Cooking]
489
291
-* jc/merge-base-fp-only (2016-10-19) 8 commits
292
- . merge-base: fp experiment
293
- - merge: allow to use only the fp-only merge bases
294
- - merge-base: limit the output to bases that are on first-parent chain
295
- - merge-base: mark bases that are on first-parent chain
296
- - merge-base: expose get_merge_bases_many_0() a bit more
297
- - merge-base: stop moving commits around in remove_redundant()
298
- - sha1_name: remove ONELINE_SEEN bit
299
- - commit: simplify fastpath of merge-base
300
-
301
- An experiment of merge-base that ignores common ancestors that are
302
- not on the first parent chain.
303
-
304
-
305
-* bw/submodule-branch-dot-doc (2016-10-19) 1 commit
306
- (merged to 'next' on 2016-10-21 at 18aad25ba8)
307
- + submodules doc: update documentation for "." used for submodule branches
308
-
309
- Recent git allows submodule.<name>.branch to use a special token
310
- "." instead of the branch name; the documentation has been updated
311
- to describe it.
312
-
313
- Will merge to 'master'.
314
-
315
-
316
-* tg/add-chmod+x-fix (2016-10-20) 1 commit
317
- (merged to 'next' on 2016-10-21 at 1585ac7139)
318
- + t3700: fix broken test under !SANITY
319
-
320
- A hot-fix for a test added by a recent topic that went to both
321
- 'master' and 'maint' already.
322
-
323
- Will merge to 'master'.
324
-
325
-
326
-* jk/diff-submodule-diff-inline (2016-10-20) 1 commit
327
- (merged to 'next' on 2016-10-21 at 13f300805e)
328
- + rev-list: use hdr_termination instead of a always using a newline
490
+* jc/abbrev-auto (2016-10-22) 4 commits
491
+ (merged to 'next' on 2016-10-26 at 92fdb66807)
492
+ + transport: compute summary-width dynamically
493
+ + transport: allow summary-width to be computed dynamically
494
+ + fetch: pass summary_width down the callchain
495
+ + transport: pass summary_width down the callchain
496
+ (this branch uses jk/abbrev-auto and lt/abbrev-auto.)
497
330
- A recently graduated topic regressed "git rev-list --header"
331
- output, breaking "gitweb". This has been fixed.
498
+ "git push" and "git fetch" reports from what old object to what new
499
+ object each ref was updated, using abbreviated refnames, and they
500
+ attempt to align the columns for this and other pieces of
501
+ information. The way these codepaths compute how many display
502
+ columns to allocate for the object names portion of this output has
503
+ been updated to match the recent "auto scale the default
504
+ abbreviation length" change.
505
506
Will merge to 'master'.
507
508
336
-* jk/no-looking-at-dotgit-outside-repo (2016-10-20) 8 commits
337
- - setup_git_env: avoid blind fall-back to ".git"
338
- - diff: handle sha1 abbreviations outside of repository
339
- - diff_aligned_abbrev: use "struct oid"
340
- - diff_unique_abbrev: rename to diff_aligned_abbrev
341
- - find_unique_abbrev: use 4-buffer ring
342
- - test-*-cache-tree: setup git dir
343
- - read info/{attributes,exclude} only when in repository
344
- - Merge branch 'jc/diff-unique-abbrev-comments' into jk/no-looking-at-dotgit-outside-repo
345
- (this branch uses jc/diff-unique-abbrev-comments.)
346
-
347
- Update "git diff --no-index" codepath not to try to peek into .git/
348
- directory that happens to be under the current directory, when we
349
- know we are operating outside any repository.
509
+* jc/reset-unmerge (2016-10-24) 1 commit
510
+ - reset: --unmerge
511
351
- Will wait until 'jc/diff-unique-abbrev-comments' graduates, rebase
352
- onto 'master' and then cook in 'next'.
512
+ After "git add" is run prematurely during a conflict resolution,
513
+ "git diff" can no longer be used as a way to sanity check by
514
+ looking at the combined diff. "git reset" learned a new
515
+ "--unmerge" option to recover from this situation.
516
517
355
-* pt/gitgui-updates (2016-10-20) 35 commits
356
- (merged to 'next' on 2016-10-21 at 4c8214095a)
357
- + Merge tag 'gitgui-0.21.0' of git://repo.or.cz/git-gui
358
- + git-gui: set version 0.21
359
- + Merge branch 'as/bulgarian' into pu
360
- + git-gui: Mark 'All' in remote.tcl for translation
361
- + git-gui i18n: Updated Bulgarian translation (565,0f,0u)
362
- + Merge branch 'os/preserve-author' into pu
363
- + git-gui: avoid persisting modified author identity
364
- + git-gui: Do not reset author details on amend
365
- + Merge branch 'kb/unicode' into pu
366
- + git-gui: handle the encoding of Git's output correctly
367
- + git-gui: unicode file name support on windows
368
- + Merge branch 'dr/ru' into pu
369
- + git-gui: Update Russian translation
370
- + git-gui: maintain backwards compatibility for merge syntax
371
- + Merge branch 'va/i18n_2' into pu
372
- + git-gui i18n: mark string in lib/error.tcl for translation
373
- + git-gui: fix incorrect use of Tcl append command
374
- + git-gui i18n: mark "usage:" strings for translation
375
- + git-gui i18n: internationalize use of colon punctuation
376
- + Merge branch 'pt/non-mouse-usage' into pu
377
- + Amend tab ordering and text widget border and highlighting.
378
- + Allow keyboard control to work in the staging widgets.
379
- + Merge branch 'pt/git4win-mods' into pu
380
- + git-gui (Windows): use git-gui.exe in `Create Desktop Shortcut`
381
- + git-gui: fix detection of Cygwin
382
- + Merge branch 'patches' into pu
383
- + git-gui: ensure the file in the diff pane is in the list of selected files
384
- + git-gui: support for $FILENAMES in tool definitions
385
- + git-gui: fix initial git gui message encoding
386
- + git-gui/po/glossary/txt-to-pot.sh: use the $( ... ) construct for command substitution
387
- + Merge branch 'va/i18n' into pu
388
- + Merge branch 'rs/use-modern-git-merge-syntax' into pu
389
- + Merge branch 'js/commit-gpgsign' into pu
390
- + Merge branch 'sy/i18n' into pu
391
- + git-gui: sort entries in tclIndex
518
+* jk/daemon-path-ok-check-truncation (2016-10-24) 1 commit
519
+ (merged to 'next' on 2016-10-26 at 70c08241f6)
520
+ + daemon: detect and reject too-long paths
521
393
- A new version of git-gui, now at its 0.21.0 tag.
522
+ "git daemon" used fixed-length buffers to turn URL to the
523
+ repository the client asked for into the server side directory
524
+ path, using snprintf() to avoid overflowing these buffers, but
525
+ allowed possibly truncated paths to the directory. This has been
526
+ tightened to reject such a request that causes overlong path to be
527
+ required to serve.
528
529
Will merge to 'master'.
530
531
398
-* yk/git-tag-remove-mention-of-old-layout-in-doc (2016-10-20) 1 commit
399
- (merged to 'next' on 2016-10-21 at 8d9e23b023)
400
- + doc: remove reference to the traditional layout in git-tag.txt
532
+* ls/git-open-cloexec (2016-10-25) 3 commits
533
+ (merged to 'next' on 2016-10-26 at f7259cbddb)
534
+ + read-cache: make sure file handles are not inherited by child processes
535
+ + sha1_file: open window into packfiles with O_CLOEXEC
536
+ + sha1_file: rename git_open_noatime() to git_open()
537
402
- Shorten description of auto-following in "git tag" by removing a
403
- mention of historical remotes layout which is not relevant to the
404
- main topic.
538
+ Git generally does not explicitly close file descriptors that were
539
+ open in the parent process when spawning a child process, but most
540
+ of the time the child does not want to access them. As Windows does
541
+ not allow removing or renaming a file that has a file descriptor
542
+ open, a slow-to-exit child can even break the parent process by
543
+ holding onto them. Use O_CLOEXEC flag to open files in various
544
+ codepaths.
545
546
Will merge to 'master'.
547
548
409
-* dk/worktree-dup-checkout-with-bare-is-ok (2016-10-14) 1 commit
410
- (merged to 'next' on 2016-10-17 at 24594d3e56)
411
- + worktree: allow the main brach of a bare repository to be checked out
549
+* rs/ring-buffer-wraparound (2016-10-26) 1 commit
550
+ (merged to 'next' on 2016-10-26 at d2da68a14a)
551
+ + hex: make wraparound of the index into ring-buffer explicit
552
413
- In a worktree connected to a repository elsewhere, created via "git
414
- worktree", "git checkout" attempts to protect users from confusion
415
- by refusing to check out a branch that is already checked out in
416
- another worktree. However, this also prevented checking out a
417
- branch, which is designated as the primary branch of a bare
418
- reopsitory, in a worktree that is connected to the bare
419
- repository. The check has been corrected to allow it.
553
+ The code that we have used for the past 10+ years to cycle
554
+ 4-element ring buffers turns out to be not quite portable in
555
+ theoretical world.
556
557
Will merge to 'master'.
558
559
424
-* jc/cocci-xstrdup-or-null (2016-10-12) 1 commit
425
- (merged to 'next' on 2016-10-17 at 55ceaa465a)
426
- + cocci: refactor common patterns to use xstrdup_or_null()
560
+* jc/merge-base-fp-only (2016-10-19) 8 commits
561
+ . merge-base: fp experiment
562
+ - merge: allow to use only the fp-only merge bases
563
+ - merge-base: limit the output to bases that are on first-parent chain
564
+ - merge-base: mark bases that are on first-parent chain
565
+ - merge-base: expose get_merge_bases_many_0() a bit more
566
+ - merge-base: stop moving commits around in remove_redundant()
567
+ - sha1_name: remove ONELINE_SEEN bit
568
+ - commit: simplify fastpath of merge-base
569
428
- Code cleanup.
570
+ An experiment of merge-base that ignores common ancestors that are
571
+ not on the first parent chain.
572
+
573
+
574
+* jk/no-looking-at-dotgit-outside-repo (2016-10-26) 6 commits
575
+ (merged to 'next' on 2016-10-26 at 4aa877b578)
576
+ + diff: handle sha1 abbreviations outside of repository
577
+ + diff_aligned_abbrev: use "struct oid"
578
+ + diff_unique_abbrev: rename to diff_aligned_abbrev
579
+ + find_unique_abbrev: use 4-buffer ring
580
+ + test-*-cache-tree: setup git dir
581
+ + read info/{attributes,exclude} only when in repository
582
+ (this branch is used by jk/no-looking-at-dotgit-outside-repo-final.)
583
+
584
+ Update "git diff --no-index" codepath not to try to peek into .git/
585
+ directory that happens to be under the current directory, when we
586
+ know we are operating outside any repository.
587
588
Will merge to 'master'.
589
@@ -444,48 +602,6 @@ of the repositories listed at
602
Waiting for review.
603
604
447
-* jk/ambiguous-short-object-names (2016-10-12) 1 commit
448
- (merged to 'next' on 2016-10-19 at e7c55a9da5)
449
- + t1512: become resilient to GETTEXT_POISON build
450
-
451
- A test fixup to recently graduated topic.
452
-
453
- Will merge to 'master'.
454
-
455
-
456
-* jk/merge-base-fork-point-without-reflog (2016-10-12) 1 commit
457
- (merged to 'next' on 2016-10-19 at 00a6797f62)
458
- + merge-base: handle --fork-point without reflog
459
-
460
- "git rebase" immediately after "git clone" failed to find the fork
461
- point from the upstream.
462
-
463
- Will merge to 'master'.
464
-
465
-
466
-* jk/upload-pack-use-prio-queue (2016-10-11) 1 commit
467
- (merged to 'next' on 2016-10-19 at 1d6efb07ac)
468
- + upload-pack: use priority queue in reachable() check
469
-
470
- Code clean-up and performance improvement to reduce use of
471
- timestamp-ordered commit-list by replacing it with a priority
472
- queue.
473
-
474
- Will merge to 'master'.
475
-
476
-
477
-* jk/fetch-quick-tag-following (2016-10-14) 1 commit
478
- (merged to 'next' on 2016-10-19 at d7718dcdf5)
479
- + fetch: use "quick" has_sha1_file for tag following
480
-
481
- When fetching from a remote that has many tags that are irrelevant
482
- to branches we are following, we used to waste way too many cycles
483
- when checking if the object pointed at by a tag (that we are not
484
- going to fetch!) exists in our repository too carefully.
485
-
486
- Will merge to 'master'.
487
-
488
-
605
* jt/trailer-with-cruft (2016-10-21) 8 commits
606
- trailer: support values folded to multiple lines
607
- trailer: forbid leading whitespace in trailers
@@ -504,55 +620,6 @@ of the repositories listed at
620
Waiting for review.
621
622
507
-* mm/send-email-cc-cruft-after-address (2016-10-21) 3 commits
508
- (merged to 'next' on 2016-10-21 at c7ec2b5025)
509
- + Git.pm: add comment pointing to t9000
510
- + t9000-addresses: update expected results after fix
511
- (merged to 'next' on 2016-10-19 at 41e3f876cd)
512
- + parse_mailboxes: accept extra text after <...> address
513
-
514
- "git send-email" attempts to pick up valid e-mails from the
515
- trailers, but people in real world write non-addresses there, like
516
- "Cc: Stable <add@re.ss> # 4.8+", which broke the output depending
517
- on the availability and vintage of Mail::Address perl module.
518
-
519
- Will merge to 'master'.
520
-
521
-
522
-* va/i18n (2016-10-17) 7 commits
523
- (merged to 'next' on 2016-10-19 at b7d733698b)
524
- + i18n: diff: mark warnings for translation
525
- + i18n: credential-cache--daemon: mark advice for translation
526
- + i18n: convert mark error messages for translation
527
- + i18n: apply: mark error message for translation
528
- + i18n: apply: mark error messages for translation
529
- + i18n: apply: mark info messages for translation
530
- + i18n: apply: mark plural string for translation
531
-
532
- More i18n.
533
-
534
- Will merge to 'master'.
535
-
536
-
537
-* jk/tighten-alloc (2016-10-17) 2 commits
538
- (merged to 'next' on 2016-10-19 at 548522a520)
539
- + inline xalloc_flex() into FLEXPTR_ALLOC_MEM
540
- + avoid pointer arithmetic involving NULL in FLEX_ALLOC_MEM
541
-
542
- Protect our code from over-eager compilers.
543
-
544
- Will merge to 'master'.
545
-
546
-
547
-* pb/test-parse-options-expect (2016-10-17) 1 commit
548
- (merged to 'next' on 2016-10-19 at d3517d592f)
549
- + t0040: convert all possible tests to use `test-parse-options --expect`
550
-
551
- Test clean-up.
552
-
553
- Will merge to 'master'.
554
-
555
-
623
* pb/bisect (2016-10-18) 27 commits
624
- bisect--helper: remove the dequote in bisect_start()
625
- bisect--helper: retire `--bisect-auto-next` subcommand
@@ -587,70 +654,53 @@ of the repositories listed at
654
Waiting for review.
655
656
590
-* ab/gitweb-abbrev-links (2016-10-14) 3 commits
591
- (merged to 'next' on 2016-10-17 at 4868def05e)
592
- + gitweb: link to "git describe"'d commits in log messages
593
- + gitweb: link to 7-char+ SHA-1s, not only 8-char+
594
- + gitweb: fix a typo in a comment
595
-
596
- In addition to purely abbreviated commit object names, "gitweb"
597
- learned to turn "git describe" output (e.g. v2.9.3-599-g2376d31787)
598
- into clickable links in its output.
599
-
600
- Will merge to 'master'.
601
-
602
-
657
* js/prepare-sequencer (2016-10-21) 27 commits
604
- - sequencer: mark all error messages for translation
605
- - sequencer: start error messages consistently with lower case
606
- - sequencer: quote filenames in error messages
607
- - sequencer: mark action_name() for translation
608
- - sequencer: remove overzealous assumption in rebase -i mode
609
- - sequencer: teach write_message() to append an optional LF
610
- - sequencer: refactor write_message() to take a pointer/length
611
- - sequencer: roll back lock file if write_message() failed
612
- - sequencer: stop releasing the strbuf in write_message()
613
- - sequencer: left-trim lines read from the script
614
- - sequencer: support cleaning up commit messages
615
- - sequencer: support amending commits
616
- - sequencer: allow editing the commit message on a case-by-case basis
617
- - sequencer: introduce a helper to read files written by scripts
618
- - sequencer: prepare for rebase -i's commit functionality
619
- - sequencer: remember the onelines when parsing the todo file
620
- - sequencer: get rid of the subcommand field
621
- - sequencer: avoid completely different messages for different actions
622
- - sequencer: strip CR from the todo script
623
- - sequencer: completely revamp the "todo" script parsing
624
- - sequencer: refactor the code to obtain a short commit name
625
- - sequencer: future-proof read_populate_todo()
626
- - sequencer: plug memory leaks for the option values
627
- - sequencer: future-proof remove_sequencer_state()
628
- - sequencer: avoid unnecessary indirection
629
- - sequencer: use memoized sequencer directory path
630
- - sequencer: use static initializers for replay_opts
658
+ (merged to 'next' on 2016-10-26 at 12be8ebe90)
659
+ + sequencer: mark all error messages for translation
660
+ + sequencer: start error messages consistently with lower case
661
+ + sequencer: quote filenames in error messages
662
+ + sequencer: mark action_name() for translation
663
+ + sequencer: remove overzealous assumption in rebase -i mode
664
+ + sequencer: teach write_message() to append an optional LF
665
+ + sequencer: refactor write_message() to take a pointer/length
666
+ + sequencer: roll back lock file if write_message() failed
667
+ + sequencer: stop releasing the strbuf in write_message()
668
+ + sequencer: left-trim lines read from the script
669
+ + sequencer: support cleaning up commit messages
670
+ + sequencer: support amending commits
671
+ + sequencer: allow editing the commit message on a case-by-case basis
672
+ + sequencer: introduce a helper to read files written by scripts
673
+ + sequencer: prepare for rebase -i's commit functionality
674
+ + sequencer: remember the onelines when parsing the todo file
675
+ + sequencer: get rid of the subcommand field
676
+ + sequencer: avoid completely different messages for different actions
677
+ + sequencer: strip CR from the todo script
678
+ + sequencer: completely revamp the "todo" script parsing
679
+ + sequencer: refactor the code to obtain a short commit name
680
+ + sequencer: future-proof read_populate_todo()
681
+ + sequencer: plug memory leaks for the option values
682
+ + sequencer: future-proof remove_sequencer_state()
683
+ + sequencer: avoid unnecessary indirection
684
+ + sequencer: use memoized sequencer directory path
685
+ + sequencer: use static initializers for replay_opts
686
687
Update of the sequencer codebase to make it reusable to reimplement
688
"rebase -i" continues.
689
635
- Will merge to 'next'.
690
+ Will merge to 'master'.
691
692
638
-* sb/submodule-ignore-trailing-slash (2016-10-18) 3 commits
639
- . submodule--helper: normalize funny urls
693
+* sb/submodule-ignore-trailing-slash (2016-10-25) 3 commits
694
+ (merged to 'next' on 2016-10-26 at e56a8ebb38)
695
+ + t0060: sidestep surprising path mangling results on Windows
696
(merged to 'next' on 2016-10-11 at e37425ed17)
697
+ submodule: ignore trailing slash in relative url
698
+ submodule: ignore trailing slash on superproject URL
699
700
A minor regression fix for "git submodule".
701
646
- It seems that POSIX emulation layer of Windows is not cooperating;
647
- this may have to wait (or tentatively reverted in Windows port) for
648
- the resolution of the issue.
649
-
650
- cf. <alpine.DEB.2.20.1610131255001.197091@virtualbox>
651
- cf. <CAGZ79kYrKGLEOO72aWuX5OOM-AecdFZFXRqBkRzhdAM-VbPFxA@mail.gmail.com>
652
-
653
- What's the current state of this topic?
702
+ Will merge to 'master'.
703
+ Queued with a test breakage workaround on Windows from j6t.
704
705
706
* st/verify-tag (2016-10-10) 7 commits
@@ -669,17 +719,6 @@ of the repositories listed at
719
cf. <20161007210721.20437-1-santiago@nyu.edu>
720
721
672
-* mm/credential-libsecret (2016-10-11) 1 commit
673
- (merged to 'next' on 2016-10-17 at 1b4af03ba4)
674
- + contrib: add credential helper for libsecret
675
-
676
- A new credential helper that talks via "libsecret" with
677
- implementations of XDG Secret Service API has been added to
678
- contrib/credential/.
679
-
680
- Will merge to 'master'.
681
-
682
-
722
* sb/attr (2016-10-24) 36 commits
723
- completion: clone can initialize specific submodules
724
- clone: add --init-submodule=<pathspec> switch
@@ -723,18 +762,7 @@ of the repositories listed at
762
Building on top of the updated API, the pathspec machinery learned
763
to select only paths with given attributes set.
764
726
-
727
-* jc/ws-error-highlight (2016-10-04) 4 commits
728
- (merged to 'next' on 2016-10-17 at ecbdc57d77)
729
- + diff: introduce diff.wsErrorHighlight option
730
- + diff.c: move ws-error-highlight parsing helpers up
731
- + diff.c: refactor parse_ws_error_highlight()
732
- + t4015: split out the "setup" part of ws-error-highlight test
733
-
734
- "git diff/log --ws-error-highlight=<kind>" lacked the corresponding
735
- configuration variable to set it by default.
736
-
737
- Will merge to 'master'.
765
+ Waiting for review.
766
767
768
* jk/abbrev-auto (2016-10-03) 1 commit
@@ -747,14 +775,15 @@ of the repositories listed at
775
which in turn needs to estimate how many hexdigits are necessary to
776
ensure uniqueness.
777
750
- Undecided.
778
+ Will merge to 'master'.
779
780
781
* nd/ita-empty-commit (2016-10-24) 4 commits
754
- - commit: don't be fooled by ita entries when creating initial commit
755
- - commit: fix empty commit creation when there's no changes but ita entries
756
- - diff: add --ita-[in]visible-in-index
757
- - diff-lib: allow ita entries treated as "not yet exist in index"
782
+ (merged to 'next' on 2016-10-26 at fb007cdae1)
783
+ + commit: don't be fooled by ita entries when creating initial commit
784
+ + commit: fix empty commit creation when there's no changes but ita entries
785
+ + diff: add --ita-[in]visible-in-index
786
+ + diff-lib: allow ita entries treated as "not yet exist in index"
787
788
When new paths were added by "git add -N" to the index, it was
789
enough to circumvent the check by "git commit" to refrain from
@@ -762,7 +791,7 @@ of the repositories listed at
791
prevented "git status" to show such a path as "new file" in the
792
"Changes not staged for commit" section.
793
765
- Will merge to 'next'.
794
+ Will merge to 'master'.
795
796
797
* lt/abbrev-auto (2016-10-03) 3 commits
@@ -779,16 +808,6 @@ of the repositories listed at
808
Will hold to see if people scream.
809
810
782
-* jc/diff-unique-abbrev-comments (2016-09-30) 1 commit
783
- (merged to 'next' on 2016-10-17 at c7fb286102)
784
- + diff_unique_abbrev(): document its assumption and limitation
785
- (this branch is used by jk/no-looking-at-dotgit-outside-repo.)
786
-
787
- A bit more comments in a tricky code.
788
-
789
- Will merge to 'master'.
790
-
791
-
811
* va/i18n-perl-scripts (2016-10-20) 14 commits
812
- i18n: difftool: mark warnings for translation
813
- i18n: send-email: mark string with interpolation for translation
@@ -823,51 +842,6 @@ of the repositories listed at
842
Will hold to see if people scream.
843
844
826
-* mg/gpg-richer-status (2016-10-12) 1 commit
827
- (merged to 'next' on 2016-10-17 at 8843a6a8be)
828
- + gpg-interface: use more status letters
829
-
830
- The GPG verification status shown in "%G?" pretty format specifier
831
- was not rich enough to differentiate a signature made by an expired
832
- key, a signature made by a revoked key, etc. New output letters
833
- have been assigned to express them.
834
-
835
- Will merge to 'master'.
836
-
837
-
838
-* js/libify-require-clean-work-tree (2016-10-07) 6 commits
839
- (merged to 'next' on 2016-10-17 at f5c20df38b)
840
- + wt-status: begin error messages with lower-case
841
- + wt-status: teach has_{unstaged,uncommitted}_changes() about submodules
842
- + wt-status: export also the has_un{staged,committed}_changes() functions
843
- + wt-status: make the require_clean_work_tree() function reusable
844
- + pull: make code more similar to the shell script again
845
- + pull: drop confusing prefix parameter of die_on_unclean_work_tree()
846
-
847
- The require_clean_work_tree() helper was recreated in C when "git
848
- pull" was rewritten from shell; the helper is now made available to
849
- other callers in preparation for upcoming "rebase -i" work.
850
-
851
- Will merge to 'master'.
852
-
853
-
854
-* bw/ls-files-recurse-submodules (2016-10-10) 4 commits
855
- (merged to 'next' on 2016-10-17 at f0e398946a)
856
- + ls-files: add pathspec matching for submodules
857
- + ls-files: pass through safe options for --recurse-submodules
858
- + ls-files: optionally recurse into submodules
859
- + git: make super-prefix option
860
-
861
- "git ls-files" learned "--recurse-submodules" option that can be
862
- used to get a listing of tracked files across submodules (i.e. this
863
- only works with "--cached" option, not for listing untracked or
864
- ignored files). This would be a useful tool to sit on the upstream
865
- side of a pipe that is read with xargs to work on all working tree
866
- files from the top-level superproject.
867
-
868
- Will merge to 'master'.
869
-
870
-
845
* ls/filter-process (2016-10-17) 14 commits
846
(merged to 'next' on 2016-10-19 at ffd0de042c)
847
+ contrib/long-running-filter: add long running filter example
@@ -892,22 +866,7 @@ of the repositories listed at
866
all filtering need is served by this single process for multiple
867
paths, reducing the process creation overhead.
868
895
- Will wait for ls/git-open-cloexec.
896
-
897
-
898
-* hv/submodule-not-yet-pushed-fix (2016-10-10) 3 commits
899
- - batch check whether submodule needs pushing into one call
900
- - serialize collection of refs that contain submodule changes
901
- - serialize collection of changed submodules
902
-
903
- The code in "git push" to compute if any commit being pushed in the
904
- superproject binds a commit in a submodule that hasn't been pushed
905
- out was overly inefficient, making it unusable even for a small
906
- project that does not have any submodule but have a reasonable
907
- number of refs.
908
-
909
- Waiting for review.
910
- cf. <cover.1475851621.git.hvoigt@hvoigt.net>
869
+ Will merge to 'master'.
870
871
872
* sg/fix-versioncmp-with-common-suffix (2016-09-08) 5 commits
@@ -926,16 +885,6 @@ of the repositories listed at
885
cf. <20160908223727.Horde.jVOOJ278ssZ3qkyjkmyqZD-@webmail.informatik.kit.edu>
886
887
929
-* sb/push-make-submodule-check-the-default (2016-10-10) 2 commits
930
- - push: change submodule default to check when submodules exist
931
- - submodule add: extend force flag to add existing repos
932
-
933
- Turn the default of "push.recurseSubmodules" to "check" when
934
- submodules seem to be in use.
935
-
936
- Will hold to wait for hv/submodule-not-yet-pushed-fix
937
-
938
-
888
* jc/pull-rebase-ff (2016-07-28) 1 commit
889
- pull: fast-forward "pull --rebase=true"
890
@@ -946,26 +895,6 @@ of the repositories listed at
895
Needs a real log message and a few tests.
896
897
949
-* ex/deprecate-empty-pathspec-as-match-all (2016-06-22) 1 commit
950
- (merged to 'next' on 2016-09-21 at e19148ea63)
951
- + pathspec: warn on empty strings as pathspec
952
-
953
- Originally merged to 'next' on 2016-07-13
954
-
955
- An empty string used as a pathspec element has always meant
956
- 'everything matches', but it is too easy to write a script that
957
- finds a path to remove in $path and run 'git rm "$paht"', which
958
- ends up removing everything. Start warning about this use of an
959
- empty string used for 'everything matches' and ask users to use a
960
- more explicit '.' for that instead.
961
-
962
- The hope is that existing users will not mind this change, and
963
- eventually the warning can be turned into a hard error, upgrading
964
- the deprecation into removal of this (mis)feature.
965
-
966
- Will hold to see if people scream.
967
-
968
-
898
* jc/merge-drop-old-syntax (2015-04-29) 1 commit
899
(merged to 'next' on 2016-10-11 at 8928c8b9b3)
900
+ merge: drop 'git merge <message> HEAD <commit>' syntax
@@ -978,4 +907,4 @@ of the repositories listed at
907
which needs to be fixed before this final step can proceed.
908
cf. <5671DB28.8020901@kdbg.org>
909
981
- Will merge to 'master'.
910
+ Will cook in 'next'.