What's cooking (2015/11 #03)
Jeff King committed
Nov 20, 2015 at 08:55 UTC
aa8d3e403b9972afc9ac0322b4e6b04c97d707d6
1 file changed
+299
-127
whats-cooking.txt
+299
-127
@@ -1,94 +1,269 @@
1
To: git@vger.kernel.org
2
Bcc: lwn@lwn.net
3
-Subject: What's cooking in git.git (Nov 2015, #02; Fri, 6)
4
-X-master-at: f34be46e47773d03e9d09641209121591a6b37c8
5
-X-next-at: c10b11f912265621bc9c1988f048b8c065303d87
3
+Subject: What's cooking in git.git (Nov 2015, #03; Fri, 20)
4
+X-master-at: 0c83680e9c047170614fb08ef222ea4f460e514d
5
+X-next-at: dc6ae483b6062dec4979462c5b072d883a323fee
6
7
-What's cooking in git.git (Nov 2015, #02; Fri, 6)
7
+What's cooking in git.git (Nov 2015, #03; Fri, 20)
8
--------------------------------------------------
9
10
Here are the topics that have been cooking. Commits prefixed with
11
'-' are only in 'pu' (proposed updates) while commits prefixed with
12
'+' are in 'next'.
13
14
-I'll be offline on vacation for a few weeks, and Jeff King
15
-graciously agreed to help shepherd the project forward in the
16
-meantime as an interim maintainer. Please be gentle.
14
+Sorry for the extreme delay in pushing out my tree. I ended up busier
15
+than I expected during the first week of Junio's absence, and then it
16
+took me a while to dig myself out of the ever-filling hole of incoming
17
+topics. :)
18
18
-You can find the changes described here in the integration branches
19
-of the repositories listed at
19
+Expect quite a few new topics in this edition. A number of them have
20
+seen review already, and I'm planning to merge them to "next" fairly
21
+quickly (I may do another integration cycle late tonight or tomorrow
22
+with those changes).
23
+
24
+I still have a few threads to comb through on the list. If you don't see
25
+the latest version of your topic here, please feel free to ping me, or
26
+to repost.
27
+
28
+You can find the normal integration branches at:
29
+
30
+ https://github.com/git/git/
31
+
32
+and all topic branches at:
33
+
34
+ https://github.com/peff/git/
35
+
36
+But note that I will _not_ be pushing to kernel.org.
37
21
- http://git-blame.blogspot.com/p/git-public-repositories.html
38
39
--------------------------------------------------
40
[New Topics]
41
26
-* dt/refs-backend-pre-vtable (2015-11-03) 26 commits
27
- - refs: break out ref conflict checks
28
- - refs: create a shared version of verify_refname_available
29
- - files_log_ref_write: new function
30
- - initdb: make safe_create_dir public
31
- - refs: move should_autocreate_reflog to common code
32
- - refs: move peel_object to the common code
33
- - refs: move copy_msg to the common code
34
- - refs: move refname_is_safe to the common code
35
- - refs: move transaction functions to the common code
36
- - refs: move head_ref_namespaced to the common code
37
- - refs: move ref iterators to the common code
38
- - refs: move prettify_refname to the common code
39
- - refs: move is_branch to the common code
40
- - refs: move check_refname_format to the common code
41
- - refs: move resolve_refdup to common
42
- - refs: move read_ref, read_ref_full and ref_exists to the common code
43
- - refs: move warn_if_dangling_symref* to the common code
44
- - refs: move dwim and friend functions to the common refs code
45
- - refs: move the hidden refs functions to the common code
46
- - refs: move read_ref_at to the common refs file
47
- - refs: move delete_pseudoref and delete_ref to the common code
48
- - refs: move update_ref to refs/refs.c
49
- - refs: add a new file, refs/refs.c, to hold common refs code
50
- - refs/files-backend.c: new file, renamed from refs.c
51
- - refs: make is_branch public
52
- - Merge branch 'jk/war-on-sprintf' into HEAD
53
-
54
- This is a verbatim copy of refs-backend-pre-vtable branch taken
55
- from https://github.com/mhagger/git; not even signed-off by me, and
56
- "thanks for noticing; Junio can you amend?" obviously has not been
57
- done. We would want to rebase the series (after discarding the
58
- bottom "Merge into HEAD") on top of 228905f in order to do the
59
- amending and then re-merge, but I ran out of time (and patience).
42
+* ap/doc-env-variable-format (2015-11-11) 1 commit
43
+ - Documentation: make environment variable formatting more consistent
44
45
+ Minor documentation fixup.
46
62
-* rc/configure-use-libs-when-checking-a-lib (2015-11-06) 1 commit
63
- - configure.ac: use $LIBS not $CFLAGS when testing -lpthread
47
+ Will merge to 'next'.
48
65
- The "configure" script did not test for -lpthread correctly, which
66
- upset some linkers.
49
+
50
+* cb/hook-sigpipe (2015-11-16) 1 commit
51
+ - allow hooks to ignore their standard input stream
52
+
53
+ We now consistently allow all hooks to ignore their standard input,
54
+ rather than having git complain of SIGPIPE.
55
+
56
+ Will merge to 'next'.
57
+
58
+
59
+* dg/subtree-test-cleanup (2015-11-13) 7 commits
60
+ - contrib/subtree: Handle '--prefix' argument with a slash appended
61
+ - contrib/subtree: Make each test self-contained
62
+ - contrib/subtree: Add split tests
63
+ - contrib/subtree: Add merge tests
64
+ - contrib/subtree: Add tests for subtree add
65
+ - contrib/subtree: Add test for missing subtree
66
+ - contrib/subtree: Clean and refactor test code
67
+
68
+ Test cleanups for the subtree project.
69
+
70
+ Will merge to 'next'.
71
+
72
+
73
+* dk/check-ignore-docs (2015-11-20) 1 commit
74
+ - check-ignore: correct documentation about output
75
+
76
+ Documentation clarification for "check-ignore" without "--verbose".
77
+
78
+ This may be losing information, and we would want to enhance the
79
+ "--verbose" documentation.
80
+
81
+ Will hold for comments.
82
+
83
+
84
+* dk/rerere-train-quoting (2015-11-20) 1 commit
85
+ - Escape Git's exec path in contrib/rerere-train.sh script
86
+
87
+ Fix shell quoting in contrib script.
88
+
89
+ Will merge to 'next'.
90
+
91
+
92
+* dk/t5813-unc-paths (2015-11-20) 1 commit
93
+ - t5813: avoid creating urls that break on cygwin
94
+
95
+ Test portability fix for a topic in v2.6.1.
96
+
97
+ Will merge to 'next'.
98
+
99
+
100
+* ec/annotate-deleted (2015-11-20) 1 commit
101
+ - annotate: skip checking working tree if a revision is provided
102
+
103
+ Usability fix for annotate-specific "<file> <rev>" syntax with deleted
104
+ files.
105
+
106
+ Needs review.
107
+
108
+
109
+* fm/shell-path-whitespace (2015-11-13) 1 commit
110
+ - rebase-i-exec: Allow space in SHELL_PATH
111
+
112
+ Portability fix for Windows, which may rewrite $SHELL variable using
113
+ non-POSIX paths.
114
+
115
+ Will merge to 'next'.
116
+
117
+
118
+* jk/filter-branch-no-index (2015-11-06) 1 commit
119
+ - filter-branch: skip index read/write when possible
120
+
121
+ Speed up filter-branch for cases where we only care about rewriting
122
+ commits, not tree data.
123
+
124
+ Needs review.
125
+
126
+
127
+* jk/send-email-ca-path (2015-11-20) 1 commit
128
+ - send-email: die if CA path doesn't exist
129
+
130
+ Use a safer behavior when we hit errors verifying remote certificates.
131
+
132
+ We may want to update this to guide users toward workarounds.
133
+
134
+
135
+* jk/send-email-complete-aliases (2015-11-20) 2 commits
136
+ - completion: add support for completing email aliases
137
+ - sendemail: teach git-send-email to dump alias names
138
+
139
+ Teach send-email to dump mail aliases, so that we can do tab completion
140
+ on the command line.
141
+
142
+ This is v4. Is everybody happy with this one?
143
+
144
+
145
+* jk/send-email-expand-paths (2015-11-20) 2 commits
146
+ - send-email: expand paths in sendemail.{to,cc}cmd config
147
+ - send-email: expand path in sendemail.smtpsslcertpath config
148
+
149
+ Expand paths in some send-email config variables.
150
+
151
+ Needs review.
152
+
153
+
154
+* js/test-modernize-t9300 (2015-11-20) 7 commits
155
+ - modernize t9300: move test preparations into test_expect_success
156
+ - modernize t9300: mark here-doc words to ignore tab indentation
157
+ - modernize t9300: use test_when_finished for clean-up
158
+ - modernize t9300: wrap lines after &&
159
+ - modernize t9300: use test_must_be_empty
160
+ - modernize t9300: use test_must_fail
161
+ - modernize t9300: single-quote placement and indentation
162
+
163
+ Clean up style in an ancient test.
164
165
Will merge to 'next'.
166
167
+
168
+* ls/p4-test-timeouts (2015-11-20) 3 commits
169
+ - git-p4: add trap to kill p4d on test exit
170
+ - git-p4: add p4d timeout in tests
171
+ - git-p4: retry kill/cleanup operations in tests with timeout
172
+
173
+ Work around some test flakiness with p4d.
174
+
175
+ Will merge to 'next'.
176
+
177
+
178
+* ls/test-must-fail-sigpipe (2015-11-20) 2 commits
179
+ - add "ok=sigpipe" to test_must_fail and use it to fix flaky tests
180
+ - implement test_might_fail using a refactored test_must_fail
181
+
182
+ Fix some racy client/server tests by treating SIGPIPE the same as a
183
+ normal non-zero exit.
184
+
185
+ Will merge to 'next'.
186
+
187
+
188
+* mc/push-recurse-submodules-config (2015-11-20) 1 commit
189
+ - push: add recurseSubmodules config option
190
+
191
+ Add new config to avoid typing "--recurse-submodules" on each push.
192
+
193
+ People seemed to like v1, but v2 didn't see a lot of review. Is
194
+ everybody happy?
195
+
196
+
197
+* mk/blame-first-parent (2015-11-20) 1 commit
198
+ - blame: fix object casting regression
199
+
200
+ Regression fix for a topic already in master.
201
+
202
+ Will merge to 'next'.
203
+
204
+
205
+* np/credential-cache-sighup (2015-11-20) 1 commit
206
+ - credential-cache: new option to ignore sighup
207
+
208
+ Workaround for using credential-cache with emacs.
209
+
210
+ Will merge to 'next'.
211
+
212
+
213
+* rs/fsck-nul-header (2015-11-20) 2 commits
214
+ - fsck: treat a NUL in a tag header as an error
215
+ - t1450: add tests for NUL in headers of commits and tags
216
+
217
+ Fsck did not correctly detect a NUL-truncated header in a tag.
218
+
219
+ Will merge to 'next'.
220
+
221
+
222
+* rs/parseopt-short-help (2015-11-20) 5 commits
223
+ - show-ref: stop using PARSE_OPT_NO_INTERNAL_HELP
224
+ - grep: stop using PARSE_OPT_NO_INTERNAL_HELP
225
+ - parse-options: allow -h as a short option
226
+ - parse-options: inline parse_options_usage() at its only remaining caller
227
+ - parse-options: deduplicate parse_options_usage() calls
228
+
229
+ Make "-h" command line option work more consistently in all commands.
230
+
231
+ Needs review.
232
+
233
+
234
--------------------------------------------------
235
[Graduated to "master"]
236
73
-* dt/http-range (2015-11-02) 2 commits
74
- (merged to 'next' on 2015-11-03 at 7c3cc60)
75
- + http: use off_t to store partial file size
76
- + http.c: use CURLOPT_RANGE for range requests
237
+* dg/git-svn-empty-dirs (2015-11-05) 1 commit
238
+ - git-svn: improve rebase/mkdirs performance
239
78
- A Range: request can be responded with a full response and when
79
- asked properly libcurl knows how to strip the result down to the
80
- requested range. However, we were hand-crafting a range request
81
- and it did not kick in.
240
+ This came via a pull from Eric's git-svn tree, and I dropped the
241
+ topic branch that Junio had created.
242
243
84
-* ea/checkout-progress (2015-11-01) 1 commit
85
- (merged to 'next' on 2015-11-03 at 08a8d1f)
86
- + checkout: add --progress option
244
+* dk/gc-idx-wo-pack (2015-11-04) 3 commits
245
+ (merged to 'next' on 2015-11-05 at af058fa)
246
+ + gc: remove garbage .idx files from pack dir
247
+ + t5304: test cleaning pack garbage
248
+ + prepare_packed_git(): refactor garbage reporting in pack directory
249
+
250
+ Having a leftover .idx file without corresponding .pack file in
251
+ the repository hurts performance; "git gc" learned to prune them.
252
+
253
+ We may want to do the same for .bitmap (and notice but not prune
254
+ .keep) without corresponding .pack, but that can be a separate
255
+ topic.
256
+
257
+
258
+* lf/ref-is-hidden-namespace (2015-11-05) 4 commits
259
+ (merged to 'next' on 2015-11-05 at b245404)
260
+ + t5509: add basic tests for hideRefs
261
+ + hideRefs: add support for matching full refs
262
+ + upload-pack: strip refs before calling ref_is_hidden()
263
+ + config.txt: document the semantics of hideRefs with namespaces
264
+
265
+ Extend transfer.hideRefs to work better with use of namespaces.
266
88
- "git checkout" did not follow the usual "--[no-]progress"
89
- convention and implemented only "--quiet" that is essentially
90
- a superset of "--no-progress". Extend the command to support the
91
- usual "--[no-]progress".
267
268
--------------------------------------------------
269
[Stalled]
@@ -165,6 +340,36 @@ of the repositories listed at
340
--------------------------------------------------
341
[Cooking]
342
343
+* dt/refs-backend-pre-vtable (2015-11-20) 10 commits
344
+ - refs: break out ref conflict checks
345
+ - files_log_ref_write: new function
346
+ - initdb: make safe_create_dir public
347
+ - refs: split filesystem-based refs code into a new file
348
+ - refs/refs-internal.h: new header file
349
+ - refname_is_safe(): improve docstring
350
+ - pack_if_possible_fn(): use ref_type() instead of is_per_worktree_ref()
351
+ - copy_msg(): rename to copy_reflog_msg()
352
+ - verify_refname_available(): new function
353
+ - verify_refname_available(): rename function
354
+
355
+ Code preparation for pluggable ref backends.
356
+
357
+ This is the slimmed-down version from Michael, v8. It applies a lot
358
+ more pleasantly than the old one.
359
+
360
+ Will merge to 'next'.
361
+
362
+
363
+* rc/configure-use-libs-when-checking-a-lib (2015-11-06) 1 commit
364
+ (merged to 'next' on 2015-11-20 at 37c26aa)
365
+ + configure.ac: use $LIBS not $CFLAGS when testing -lpthread
366
+
367
+ The "configure" script did not test for -lpthread correctly, which
368
+ upset some linkers.
369
+
370
+ Will merge to 'master'.
371
+
372
+
373
* ad/sha1-update-chunked (2015-11-05) 2 commits
374
- sha1: allow limiting the size of the data passed to SHA1_Update()
375
- sha1: provide another level of indirection for the SHA-1 functions
@@ -173,8 +378,7 @@ of the repositories listed at
378
more than 4GB at a time, and we now have a compile-time workaround
379
for it.
380
176
- I think this is more or less ready. I am skeptical about the file
177
- location reorg ([PATCH 3/3] $gmane/280912) and did not queue it.
381
+ Will merge to 'next'.
382
383
384
* vl/grep-configurable-threads (2015-11-01) 1 commit
@@ -183,14 +387,13 @@ of the repositories listed at
387
"git grep" can now be configured (or told from the command line)
388
how many threads to use when searching in the working tree files.
389
186
- Expecting a reroll.
187
- ($gmane/280854)
390
+ I haven't reviewed v6 yet. More eyes are welcome.
391
392
393
* kf/http-proxy-auth-methods (2015-11-04) 3 commits
191
- - SQUASH???
192
- - http: use credential API to handle proxy authentication
193
- - http: allow selection of proxy authentication method
394
+ . SQUASH???
395
+ . http: use credential API to handle proxy authentication
396
+ . http: allow selection of proxy authentication method
397
398
New http.proxyAuthMethod configuration variable can be used to
399
specify what authentication method to use, as a way to work around
@@ -199,57 +402,30 @@ of the repositories listed at
402
has been taught to use credential API to store the authentication
403
material in user's keyrings.
404
405
+ I ejected this from pu for the moment, as it conflicts with the
406
+ pt/http-socks-proxy topic.
407
+
408
Still being worked on.
409
($gmane/280925).
410
411
206
-* lf/ref-is-hidden-namespace (2015-11-05) 4 commits
207
- (merged to 'next' on 2015-11-05 at b245404)
208
- + t5509: add basic tests for hideRefs
209
- + hideRefs: add support for matching full refs
210
- + upload-pack: strip refs before calling ref_is_hidden()
211
- + config.txt: document the semantics of hideRefs with namespaces
212
-
213
- Extend transfer.hideRefs to work better with use of namespaces.
214
-
215
- Will merge to 'master'.
216
-
217
-
218
-* dg/git-svn-empty-dirs (2015-11-05) 1 commit
219
- - git-svn: improve rebase/mkdirs performance
220
-
221
-
412
* fp/subtree-todo-update (2015-11-06) 1 commit
223
- - contrib/subtree: remove "push" command from the "todo" file
224
-
225
- Will merge to 'next' and then 'master'.
226
-
227
-
228
-* dk/gc-idx-wo-pack (2015-11-04) 3 commits
229
- (merged to 'next' on 2015-11-05 at af058fa)
230
- + gc: remove garbage .idx files from pack dir
231
- + t5304: test cleaning pack garbage
232
- + prepare_packed_git(): refactor garbage reporting in pack directory
233
-
234
- Having a leftover .idx file without corresponding .pack file in
235
- the repository hurts performance; "git gc" learned to prune them.
236
-
237
- We may want to do the same for .bitmap (and notice but not prune
238
- .keep) without corresponding .pack, but that can be a separate
239
- topic.
413
+ (merged to 'next' on 2015-11-20 at 3861698)
414
+ + contrib/subtree: remove "push" command from the "todo" file
415
416
Will merge to 'master'.
417
418
244
-* sb/submodule-parallel-update (2015-11-05) 10 commits
419
+* sb/submodule-parallel-update (2015-11-20) 11 commits
420
- clone: allow an explicit argument for parallel submodule clones
421
- submodule update: expose parallelism to the user
422
- git submodule update: have a dedicated helper for cloning
423
- fetching submodules: respect `submodule.jobs` config option
249
- - submodule config: update parse_config()
250
- - submodule config: remove name_and_item_from_var
251
- - submodule config: keep update strategy around
252
- - run_processes_parallel: add output to tracing messages
424
+ - submodule-config: introduce parse_generic_submodule_config
425
+ - submodule-config: remove name_and_item_from_var
426
+ - submodule-config: drop check against NULL
427
+ - submodule-config: keep update strategy around
428
+ - run_processes_parallel: delimit intermixed task output
429
- Merge branch 'rs/daemon-plug-child-leak' into sb/submodule-parallel-update
430
- Merge branch 'sb/submodule-parallel-fetch' into sb/submodule-parallel-update
431
(this branch uses sb/submodule-parallel-fetch.)
@@ -266,6 +442,8 @@ of the repositories listed at
442
in-place instead of this follow-up topic tweaking them as past
443
mistakes and inflexibilities).
444
445
+ I picked up v4 from the list, but it needs review.
446
+
447
448
* jc/strbuf-gets (2015-10-28) 17 commits
449
- test-sha1-array: read command stream with strbuf_gets()
@@ -318,24 +496,17 @@ of the repositories listed at
496
497
498
* jc/test-must-fail-with-sigpipe (2015-10-30) 1 commit
321
- - test: accept death by SIGPIPE as a valid failure mode
322
-
323
- In some tests we expect a git command that transfers objects to
324
- result in failure, but it is normal for the failure to come from
325
- SIGPIPE when the other end of the connection detects an (expected)
326
- error and hangs up. However, test_expect_failure helper function
327
- was not prepared to see a death by sigpipe as a normal failure
328
- exit.
329
-
330
- Replacement was floated as a part of Travis CI patch series, so it
331
- is likely that this will be replaced while I (Junio) am away.
499
+ . test: accept death by SIGPIPE as a valid failure mode
500
501
+ Discarded in favor of ls/test-must-fail-sigpipe.
502
334
-* pt/http-socks-proxy (2015-10-26) 1 commit
335
- - remote-http(s): support SOCKS proxies
503
+* pt/http-socks-proxy (2015-11-20) 1 commit
504
+ (merged to 'next' on 2015-11-20 at dc6ae48)
505
+ + remote-http(s): support SOCKS proxies
506
507
Add support for talking http/https over socks proxy.
338
- Needs a sign-off from the original author.
508
+
509
+ Will cook in 'next'.
510
511
512
* mh/notes-allow-reading-treeish (2015-10-08) 3 commits
@@ -446,8 +617,9 @@ of the repositories listed at
617
618
619
* sb/submodule-parallel-fetch (2015-11-05) 16 commits
449
- - strbuf: update documentation for strbuf_read_once()
450
- - run-command: remove set_nonblocking()
620
+ (merged to 'next' on 2015-11-20 at 89fc723)
621
+ + strbuf: update documentation for strbuf_read_once()
622
+ + run-command: remove set_nonblocking()
623
(merged to 'next' on 2015-10-23 at 8f04bbd)
624
+ run-command: fix missing output from late callbacks
625
+ test-run-command: increase test coverage
@@ -469,9 +641,9 @@ of the repositories listed at
641
Add a framework to spawn a group of processes in parallel, and use
642
it to run "git fetch --recurse-submodules" in parallel.
643
472
- Still being worked on, but it seems that we are seeing light at the
473
- end of the tunnel.
474
- ($gmane/280937)
644
+ I merged the removal of the non-blocking descriptors to 'next'.
645
+
646
+ Will cook in 'next'.
647
648
649
* ad/cygwin-wants-rename (2015-08-07) 1 commit