What's cooking (2017/11 #03)
Junio C Hamano committed
Nov 8, 2017 at 14:44 UTC
d9d8f5591f5f7f5fd3a4ebaab31001c3cf5c993e
1 file changed
+242
-494
whats-cooking.txt
+242
-494
@@ -1,10 +1,10 @@
1
To: git@vger.kernel.org
2
Bcc: lwn@lwn.net
3
-Subject: What's cooking in git.git (Nov 2017, #02; Mon, 6)
3
+Subject: What's cooking in git.git (Nov 2017, #03; Wed, 8)
4
X-master-at: 7668cbc60578f99a4c048f8f8f38787930b8147b
5
-X-next-at: f294e3d99a357e12ccf7ade78a150fe6de4f7826
5
+X-next-at: 9af6a3dea062db2d4b3807c453d2f678c7f12e9c
6
7
-What's cooking in git.git (Nov 2017, #02; Mon, 6)
7
+What's cooking in git.git (Nov 2017, #03; Wed, 8)
8
--------------------------------------------------
9
10
Here are the topics that have been cooking. Commits prefixed with
@@ -12,10 +12,8 @@ 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
-Early batches for the next cycle has been merged to 'master' and
16
-'next' has become thin enough so that we can rewind it and rebuild
17
-it as usual, optionally with a handful of topics kicked out of it.
18
-That hasn't happened yet, but it will shortly.
15
+The tip of 'next' has been rebuilt on top of v2.15, while kicking a
16
+few topics back to 'pu'.
17
18
You can find the changes described here in the integration branches
19
of the repositories listed at
@@ -23,309 +21,183 @@ of the repositories listed at
21
http://git-blame.blogspot.com/p/git-public-repositories.html
22
23
--------------------------------------------------
26
-[Graduated to "master"]
27
-
28
-* ao/check-resolve-ref-unsafe-result (2017-10-21) 1 commit
29
- (merged to 'next' on 2017-10-26 at 84302ab59a)
30
- + commit: check result of resolve_ref_unsafe
31
-
32
- "git commit", after making a commit, did not check for errors when
33
- asking on what branch it made the commit, which has been correted.
34
-
35
-
36
-* bc/object-id (2017-10-16) 25 commits
37
- (merged to 'next' on 2017-10-18 at ef8aa2698d)
38
- + refs/files-backend: convert static functions to object_id
39
- + refs: convert read_raw_ref backends to struct object_id
40
- + refs: convert peel_object to struct object_id
41
- + refs: convert resolve_ref_unsafe to struct object_id
42
- + worktree: convert struct worktree to object_id
43
- + refs: convert resolve_gitlink_ref to struct object_id
44
- + Convert remaining callers of resolve_gitlink_ref to object_id
45
- + sha1_file: convert index_path and index_fd to struct object_id
46
- + refs: convert reflog_expire parameter to struct object_id
47
- + refs: convert read_ref_at to struct object_id
48
- + refs: convert peel_ref to struct object_id
49
- + builtin/pack-objects: convert to struct object_id
50
- + pack-bitmap: convert traverse_bitmap_commit_list to object_id
51
- + refs: convert dwim_log to struct object_id
52
- + builtin/reflog: convert remaining unsigned char uses to object_id
53
- + refs: convert dwim_ref and expand_ref to struct object_id
54
- + refs: convert read_ref and read_ref_full to object_id
55
- + refs: convert resolve_refdup and refs_resolve_refdup to struct object_id
56
- + Convert check_connected to use struct object_id
57
- + refs: update ref transactions to use struct object_id
58
- + refs: prevent accidental NULL dereference in write_pseudoref
59
- + refs: convert update_ref and refs_update_ref to use struct object_id
60
- + refs: convert delete_ref and refs_delete_ref to struct object_id
61
- + refs/files-backend: convert struct ref_to_prune to object_id
62
- + walker: convert to struct object_id
63
- (this branch is used by mh/tidy-ref-update-flags.)
64
-
65
- Conversion from uchar[20] to struct object_id continues.
66
-
67
-
68
-* bw/grep-recurse-submodules (2017-11-02) 1 commit
69
- (merged to 'next' on 2017-11-02 at 50bbebe9e3)
70
- + grep: take the read-lock when adding a submodule
71
-
72
- A broken access to object databases in recent update to "git grep
73
- --recurse-submodules" has been fixed.
74
-
75
-
76
-* cn/diff-indent-no-longer-is-experimental (2017-11-02) 1 commit
77
- (merged to 'next' on 2017-11-02 at d3fe3952a7)
78
- + diff: --indent-heuristic is no longer experimental
79
-
80
- Doc update.
81
-
82
-
83
-* dm/run-command-ignored-hook-advise (2017-10-10) 1 commit
84
- (merged to 'next' on 2017-10-10 at 0827814922)
85
- + run-command: add hint when a hook is ignored
86
-
87
- A hook script that is set unexecutable is simply ignored. Git
88
- notifies when such a file is ignored, unless the message is
89
- squelched via advice.ignoredHook configuration.
90
-
91
-
92
-* ds/find-unique-abbrev-optim (2017-10-13) 4 commits
93
- (merged to 'next' on 2017-10-16 at 7f4479ac52)
94
- + sha1_name: minimize OID comparisons during disambiguation
95
- + sha1_name: parse less while finding common prefix
96
- + sha1_name: unroll len loop in find_unique_abbrev_r()
97
- + p4211-line-log.sh: add log --online --raw --parents perf test
98
-
99
- Optimize the code to find shortest unique prefix of object names.
100
-
101
-
102
-* ex/deprecate-empty-pathspec-as-match-all (2017-06-23) 2 commits
103
- (merged to 'next' on 2017-08-14 at 24db08a6e8)
104
- + pathspec: die on empty strings as pathspec
105
- + t0027: do not use an empty string as a pathspec element
106
-
107
- The final step to make an empty string as a pathspec element
108
- illegal. We started this by first deprecating and warning a
109
- pathspec that has such an element in 2.11 (Nov 2016).
110
-
111
-
112
-* gc/gitweb-filetest-acl (2017-10-24) 1 commit
113
- (merged to 'next' on 2017-10-26 at 5b490f53bf)
114
- + gitweb: use filetest to allow ACLs
115
-
116
- "gitweb" checks if a directory is searchable with Perl's "-x"
117
- operator, which can be enhanced by using "filetest 'access'"
118
- pragma, which now we do.
119
-
120
-
121
-* hv/fetch-moved-submodules-on-demand (2017-10-17) 3 commits
122
- (merged to 'next' on 2017-10-26 at c446ea3e9a)
123
- + submodule: simplify decision tree whether to or not to fetch
124
- + implement fetching of moved submodules
125
- + fetch: add test to make sure we stay backwards compatible
126
-
127
- "git fetch --recurse-submodules" now knows that submodules can be
128
- moved around in the superproject in addition to getting updated,
129
- and finds the ones that need to be fetched accordingly.
130
-
131
-
132
-* jc/check-ref-format-oor (2017-10-18) 3 commits
133
- (merged to 'next' on 2017-10-18 at 8b17566c80)
134
- + check-ref-format doc: --branch validates and expands <branch>
135
- + check-ref-format --branch: strip refs/heads/ using skip_prefix
136
- + check-ref-format --branch: do not expand @{...} outside repository
137
-
138
- "git check-ref-format --branch @{-1}" bit a "BUG()" when run
139
- outside a repository for obvious reasons; clarify the documentation
140
- and make sure we do not even try to expand the at-mark magic in
141
- such a case, but still call the validation logic for branch names.
142
-
143
-
144
-* jc/no-cmd-as-subroutine (2017-10-11) 2 commits
145
- (merged to 'next' on 2017-10-16 at 5f4ee919ac)
146
- + merge-ours: do not use cmd_*() as a subroutine
147
- + describe: do not use cmd_*() as a subroutine
148
-
149
- Calling cmd_foo() as if it is a general purpose helper function is
150
- a no-no. Correct two instances of such to set an example.
151
-
152
-
153
-* jc/t5601-copy-workaround (2017-10-17) 1 commit
154
- (merged to 'next' on 2017-10-18 at ea1e133968)
155
- + t5601: rm the target file of cp that could still be executing
24
+[New Topics]
25
157
- A (possibly flakey) test fix.
26
+* bw/rebase-i-ignored-submodule-fix (2017-11-07) 1 commit
27
+ - wt-status: actually ignore submodules when requested
28
29
+ "git rebase -i" recently started misbehaving when a submodule that
30
+ is configured with 'submodule.<name>.ignore' is dirty; this has
31
+ been corrected.
32
160
-* jk/diff-color-moved-fix (2017-10-21) 5 commits
161
- (merged to 'next' on 2017-10-26 at eb580e4c8c)
162
- + diff: handle NULs in get_string_hash()
163
- + diff: fix whitespace-skipping with --color-moved
164
- + t4015: test the output of "diff --color-moved -b"
165
- + t4015: check "negative" case for "-w --color-moved"
166
- + t4015: refactor --color-moved whitespace test
167
- (this branch is used by sb/diff-color-moved-use-xdl-recmatch.)
33
+ Will merge to 'next'.
34
+ I've edited in the tweak brought up in the discussion. Please
35
+ eyeball to sanity check.
36
169
- The experimental "color moved lines differently in diff output"
170
- feature was buggy around "ignore whitespace changes" edges, whihch
171
- has been corrected.
37
38
+* ar/unconfuse-three-dots (2017-11-07) 3 commits
39
+ - Documentation: revisions: add note about 3dots usages as continuation indications
40
+ - Documentation: user-manual: limit potentially confusing usage of 3dots (and 2dots)
41
+ - checkout: describe_detached_head: remove 3dots after committish
42
174
-* jk/misc-resolve-ref-unsafe-fixes (2017-10-21) 4 commits
175
- (merged to 'next' on 2017-10-26 at c327b2ec2f)
176
- + worktree: handle broken symrefs in find_shared_symref()
177
- + log: handle broken HEAD in decoration check
178
- + remote: handle broken symrefs
179
- + test-ref-store: avoid passing NULL to printf
43
+ Ancient part of codebase still shows dots after an abbreviated
44
+ object name just to show that it is not a full object name, but
45
+ these ellipses are confusing to people who newly discovered Git
46
+ who are used to seeing abbreviated object names and find them
47
+ confusing with the range syntax.
48
181
- Some codepaths did not check for errors when asking what branch the
182
- HEAD points at, which have been fixed.
49
50
+* jt/submodule-tests-cleanup (2017-11-08) 1 commit
51
+ - Tests: clean up and document submodule helpers
52
185
-* jk/rebase-i-exec-gitdir-fix (2017-11-02) 1 commit
186
- (merged to 'next' on 2017-11-02 at 9a519c7150)
187
- + sequencer: pass absolute GIT_DIR to exec commands
53
+ Will merge to 'next'.
54
189
- A recent regression in "git rebase -i" that broke execution of git
190
- commands from subdirectories via "exec" insn has been fixed.
55
+--------------------------------------------------
56
+[Stalled]
57
58
+* jn/ssh-wrappers (2017-10-30) 6 commits
59
+ - fixup! ssh: 'auto' variant to select between 'ssh' and 'simple'
60
+ - ssh: 'simple' variant does not support --port
61
+ - ssh: 'simple' variant does not support -4/-6
62
+ - ssh: 'auto' variant to select between 'ssh' and 'simple'
63
+ - connect: split ssh command line options into separate function
64
+ - connect: split git:// setup into a separate function
65
+ (this branch uses bw/protocol-v1.)
66
193
-* jk/revision-pruning-optim (2017-10-14) 1 commit
194
- (merged to 'next' on 2017-10-16 at 2662baa21d)
195
- + revision: quit pruning diff more quickly when possible
67
+ The ssh-variant 'simple' introduced earlier broke existing
68
+ installations by not passing --port/-4/-6 and not diagnosing an
69
+ attempt to pass these as an error. Instead, default to
70
+ automatically detect how compatible the GIT_SSH/GIT_SSH_COMMAND is
71
+ to OpenSSH convention and then error out an invocation to make it
72
+ easier to diagnose connection errors.
73
197
- Pathspec-limited revision traversal was taught not to keep finding
198
- unneeded differences once it knows two trees are different inside
199
- given pathspec.
74
+ Expecting a reroll.
75
+ cf. <20171025163243.xmm7szrkwgblpgcc@aiede.mtv.corp.google.com>
76
77
202
-* js/submodule-in-excluded (2017-10-26) 1 commit
203
- (merged to 'next' on 2017-10-26 at 2a262e6a0b)
204
- + status: do not get confused by submodules in excluded directories
78
+* jc/merge-symlink-ours-theirs (2017-09-26) 1 commit
79
+ - merge: teach -Xours/-Xtheirs to symbolic link merge
80
206
- "git status --ignored -u" did not stop at a working tree of a
207
- separate project that is embedded in an ignored directory and
208
- listed files in that other project, instead of just showing the
209
- directory itself as ignored.
81
+ "git merge -Xours/-Xtheirs" learned to use our/their version when
82
+ resolving a conflicting updates to a symbolic link.
83
84
+ Needs review.
85
212
-* kd/auto-col-with-pager-fix (2017-10-24) 2 commits
213
- (merged to 'next' on 2017-10-24 at ef3ff0f14b)
214
- + column: do not include pager.c
215
- (merged to 'next' on 2017-10-18 at fe89ff7e83)
216
- + column: show auto columns when pager is active
86
218
- "auto" as a value for the columnar output configuration ought to
219
- judge "is the output consumed by humans?" with the same criteria as
220
- "auto" for coloured output configuration, i.e. either the standard
221
- output stream is going to tty, or a pager is in use. We forgot the
222
- latter, which has been fixed.
87
+* mg/merge-base-fork-point (2017-09-17) 3 commits
88
+ - merge-base: find fork-point outside partial reflog
89
+ - merge-base: return fork-point outside reflog
90
+ - t6010: test actual test output
91
92
+ "merge-base --fork-point $branch $commit" is used to guess on which
93
+ commit among the commits that were once at the tip of the $branch the
94
+ $commit was built on top of, and it learns these historical tips from
95
+ the reflog of the $branch. When the true fork-point is lost due to
96
+ pruning of old reflog entries, the command does not give any output,
97
+ because it has no way to guess correctly and does not want to mislead
98
+ the user with a wrong guess.
99
225
-* ma/lockfile-fixes (2017-10-07) 12 commits
226
- (merged to 'next' on 2017-10-16 at 19fae5c138)
227
- + read_cache: roll back lock in `update_index_if_able()`
228
- + read-cache: leave lock in right state in `write_locked_index()`
229
- + read-cache: drop explicit `CLOSE_LOCK`-flag
230
- + cache.h: document `write_locked_index()`
231
- + apply: remove `newfd` from `struct apply_state`
232
- + apply: move lockfile into `apply_state`
233
- + cache-tree: simplify locking logic
234
- + checkout-index: simplify locking logic
235
- + tempfile: fix documentation on `delete_tempfile()`
236
- + lockfile: fix documentation on `close_lock_file_gently()`
237
- + treewide: prefer lockfiles on the stack
238
- + sha1_file: do not leak `lock_file`
100
+ The command has been updated to give the best but not known to be
101
+ correct guess, based on a hope that a merge-base between $commit and a
102
+ virtual merge across all the reflog entries that still are available
103
+ for $branch may still be a closer to the true fork-point than the
104
+ merge-base between $commit and the current tip of the $branch.
105
240
- An earlier update made it possible to use an on-stack in-core
241
- lockfile structure (as opposed to having to deliberately leak an
242
- on-heap one). Many codepaths have been updated to take advantage
243
- of this new facility.
106
+ This may have to be offered by an additional option, to allow the
107
+ users that are prepared to see a potentially incorrect guess to opt
108
+ into the feature, without affecting the current callers that may not
109
+ be prepared to accept a guess that is not known to be correct.
110
111
+ What's the doneness of this one?
112
246
-* mh/test-local-canary (2017-10-31) 1 commit
247
- (merged to 'next' on 2017-11-02 at f34927f687)
248
- + t0000: check whether the shell supports the "local" keyword
113
250
- We try to see if somebody runs our test suite with a shell that
251
- does not support "local" like bash/dash does.
114
+* cc/perf-run-config (2017-09-24) 9 commits
115
+ - perf: store subsection results in "test-results/$GIT_PERF_SUBSECTION/"
116
+ - perf/run: show name of rev being built
117
+ - perf/run: add run_subsection()
118
+ - perf/run: update get_var_from_env_or_config() for subsections
119
+ - perf/run: add get_subsections()
120
+ - perf/run: add calls to get_var_from_env_or_config()
121
+ - perf/run: add GIT_PERF_DIRS_OR_REVS
122
+ - perf/run: add get_var_from_env_or_config()
123
+ - perf/run: add '--config' option to the 'run' script
124
125
+ Needs review.
126
254
-* mp/push-pushoption-config (2017-10-24) 1 commit
255
- (merged to 'next' on 2017-10-26 at 2c72164d12)
256
- + builtin/push.c: add push.pushOption config
127
258
- The "--push-option=<string>" option to "git push" now defaults to a
259
- list of strings configured via push.pushOption variable.
128
+* jk/drop-ancient-curl (2017-08-09) 5 commits
129
+ - http: #error on too-old curl
130
+ - curl: remove ifdef'd code never used with curl >=7.19.4
131
+ - http: drop support for curl < 7.19.4
132
+ - http: drop support for curl < 7.16.0
133
+ - http: drop support for curl < 7.11.1
134
135
+ Some code in http.c that has bitrot is being removed.
136
262
-* pb/bisect-helper (2017-10-06) 6 commits
263
- (merged to 'next' on 2017-10-13 at ffc9aefe63)
264
- + bisect--helper: `is_expected_rev` & `check_expected_revs` shell function in C
265
- + t6030: explicitly test for bisection cleanup
266
- + bisect--helper: `bisect_clean_state` shell function in C
267
- + bisect--helper: `write_terms` shell function in C
268
- + bisect--helper: rewrite `check_term_format` shell function in C
269
- + bisect--helper: use OPT_CMDMODE instead of OPT_BOOL
270
- (this branch is used by pb/bisect-helper-2.)
137
+ Expecting a reroll.
138
272
- An early part of piece-by-piece rewrite of "git bisect".
139
140
+* mk/use-size-t-in-zlib (2017-08-10) 1 commit
141
+ . zlib.c: use size_t for size
142
275
-* pc/submodule-helper (2017-10-07) 3 commits
276
- (merged to 'next' on 2017-10-16 at 2b38de12cc)
277
- + submodule: port submodule subcommand 'status' from shell to C
278
- + submodule--helper: introduce for_each_listed_submodule()
279
- + submodule--helper: introduce get_submodule_displaypath()
143
+ The wrapper to call into zlib followed our long tradition to use
144
+ "unsigned long" for sizes of regions in memory, which have been
145
+ updated to use "size_t".
146
281
- GSoC.
147
+ Needs resurrecting by making sure the fix is good and still applies
148
+ (or adjusted to today's codebase).
149
150
284
-* sb/diff-color-moved-use-xdl-recmatch (2017-10-26) 2 commits
285
- (merged to 'next' on 2017-10-26 at 6711f24498)
286
- + diff.c: get rid of duplicate implementation
287
- + xdiff-interface: export comparing and hashing strings
288
- (this branch uses jk/diff-color-moved-fix.)
151
+* mg/status-in-progress-info (2017-05-10) 2 commits
152
+ - status --short --inprogress: spell it as --in-progress
153
+ - status: show in-progress info for short status
154
290
- Instead of using custom line comparison and hashing functions to
291
- implement "moved lines" coloring in the diff output, use the pair
292
- of these functions from lower-layer xdiff/ code.
155
+ "git status" learns an option to report various operations
156
+ (e.g. "merging") that the user is in the middle of.
157
158
+ cf. <xmqqmvakcdqw.fsf@gitster.mtv.corp.google.com>
159
295
-* tb/complete-checkout (2017-10-25) 1 commit
296
- (merged to 'next' on 2017-10-26 at beeaf5b00c)
297
- + completion: add remaining flags to checkout
160
299
- Command line completion (in contrib/) update.
161
+* nd/worktree-move (2017-04-20) 6 commits
162
+ - worktree remove: new command
163
+ - worktree move: refuse to move worktrees with submodules
164
+ - worktree move: accept destination as directory
165
+ - worktree move: new command
166
+ - worktree.c: add update_worktree_location()
167
+ - worktree.c: add validate_worktree()
168
169
+ "git worktree" learned move and remove subcommands.
170
302
-* tg/deprecate-stash-save (2017-10-27) 3 commits
303
- (merged to 'next' on 2017-10-27 at 73b1edb73b)
304
- + stash: remove now superfluos help for "stash push"
305
- + stash: mark "git stash save" deprecated in the man page
306
- + stash: replace "git stash save" with "git stash push" in the documentation
171
+ Expecting a reroll.
172
+ cf. <20170420101024.7593-1-pclouds@gmail.com>
173
+ cf. <20170421145916.mknekgqzhxffu7di@sigill.intra.peff.net>
174
+ cf. <d0e81b1e-5869-299e-f462-4d43dc997bd1@ramsayjones.plus.com>
175
308
- "git stash save" has been deprecated in favour of "git stash push".
176
+--------------------------------------------------
177
+[Cooking]
178
179
+* js/for-each-ref-remote-name-and-ref (2017-11-08) 3 commits
180
+ - for-each-ref: test :remotename and :remoteref
181
+ - for-each-ref: let upstream/push report the remote ref name
182
+ - for-each-ref: let upstream/push optionally report the remote name
183
311
-* wk/pull-signoff (2017-10-13) 1 commit
312
- (merged to 'next' on 2017-10-16 at 5e48f349d9)
313
- + pull: pass --signoff/--no-signoff to "git merge"
184
+ The "--format=..." option "git for-each-ref" takes learned to show
185
+ the name of the 'remote' repository and the ref at the remote side
186
+ that is affected for 'upstream' and 'push' via "%(push:remotename)"
187
+ and friends.
188
315
- "git pull" has been taught to accept "--[no-]signoff" option and
316
- pass it down to "git merge".
189
+ Will merge to 'next'.
190
318
---------------------------------------------------
319
-[New Topics]
191
192
* js/mingw-redirect-std-handles (2017-11-02) 3 commits
322
- - mingw: document the standard handle redirection
323
- - mingw: optionally redirect stderr/stdout via the same handle
324
- - mingw: add experimental feature to redirect standard handles
193
+ (merged to 'next' on 2017-11-07 at 9af6a3dea0)
194
+ + mingw: document the standard handle redirection
195
+ + mingw: optionally redirect stderr/stdout via the same handle
196
+ + mingw: add experimental feature to redirect standard handles
197
198
MinGW updates.
199
328
- Will merge to 'next'.
200
+ Will merge to 'master'.
201
202
203
* ma/bisect-leakfix (2017-11-06) 4 commits
@@ -339,9 +211,9 @@ of the repositories listed at
211
Will merge to 'next'.
212
213
342
-* ma/reduce-heads-leakfix (2017-11-06) 2 commits
214
+* ma/reduce-heads-leakfix (2017-11-08) 2 commits
215
- reduce_heads: fix memory leaks
344
- - builtin/merge-base: UNLEAK commit lists
216
+ - builtin/merge-base: free commit lists
217
218
Leak fixes.
219
@@ -349,12 +221,13 @@ of the repositories listed at
221
222
223
* sg/travis-fixes (2017-11-02) 2 commits
352
- - travis-ci: don't build Git for the static analysis job
353
- - travis-ci: fix running P4 and Git LFS tests in Linux build jobs
224
+ (merged to 'next' on 2017-11-07 at bbf39361b6)
225
+ + travis-ci: don't build Git for the static analysis job
226
+ + travis-ci: fix running P4 and Git LFS tests in Linux build jobs
227
228
TravisCI build updates.
229
357
- Will merge to 'next'.
230
+ Will merge to 'master'.
231
232
233
* sr/wrapper-quote-filenames (2017-11-06) 1 commit
@@ -366,7 +239,7 @@ of the repositories listed at
239
Will merge to 'next'.
240
241
369
-* jh/fsck-promisors (2017-11-03) 9 commits
242
+* jh/fsck-promisors (2017-11-08) 9 commits
243
- gc: do not repack promisor packfiles
244
- sha1_file: support lazily fetching missing objects
245
- introduce fetch-object: fetch one promisor object
@@ -379,7 +252,7 @@ of the repositories listed at
252
(this branch is used by jh/partial-clone; uses jh/object-filtering.)
253
254
382
-* jh/object-filtering (2017-11-03) 6 commits
255
+* jh/object-filtering (2017-11-08) 6 commits
256
- pack-objects: add list-objects filtering
257
- rev-list: add list-objects filtering support
258
- list-objects: filter objects in traverse_commit_list
@@ -389,7 +262,8 @@ of the repositories listed at
262
(this branch is used by jh/fsck-promisors and jh/partial-clone.)
263
264
392
-* jh/partial-clone (2017-11-03) 14 commits
265
+* jh/partial-clone (2017-11-08) 15 commits
266
+ - fixup! upload-pack: add object filtering for partial clone
267
- index-pack: silently assume missing objects are promisor
268
- fetch-pack: restore save_commit_buffer after use
269
- unpack-trees: batch fetching of missing blobs
@@ -408,196 +282,72 @@ of the repositories listed at
282
283
284
* js/early-config (2017-11-03) 1 commit
411
- - setup: avoid double slashes when looking for HEAD
285
+ (merged to 'next' on 2017-11-07 at 9477c7c8ea)
286
+ + setup: avoid double slashes when looking for HEAD
287
413
- Correct start-up sequence so that a repository could hang
288
+ Correct start-up sequence so that a repository could be placed
289
immediately under the root directory again (which was broken at
290
around Git 2.13).
291
417
- Will merge to 'next'.
292
+ Will merge to 'master'.
293
294
295
* ks/mailmap (2017-11-03) 1 commit
421
- - mailmap: use Kaartic Sivaraam's new address
296
+ (merged to 'next' on 2017-11-07 at 46975637c7)
297
+ + mailmap: use Kaartic Sivaraam's new address
298
423
- Will merge to 'next'.
299
+ Will merge to 'master'.
300
301
302
* rs/imap-send-next-arg-fix (2017-11-03) 2 commits
427
- - imap-send: handle missing response codes gracefully
428
- - imap-send: handle NULL return of next_arg()
303
+ (merged to 'next' on 2017-11-07 at f4df55fe5b)
304
+ + imap-send: handle missing response codes gracefully
305
+ + imap-send: handle NULL return of next_arg()
306
307
Error checking in "git imap-send" for empty response has been
308
improved.
309
433
- Will merge to 'next'.
310
+ Will merge to 'master'.
311
312
313
* dk/libsecret-unlock-to-load-fix (2017-11-04) 1 commit
437
- - credential-libsecret: unlock locked secrets
314
+ (merged to 'next' on 2017-11-07 at 57d1d76c8c)
315
+ + credential-libsecret: unlock locked secrets
316
317
The credential helper for libsecret (in contrib/) has been improved
318
to allow possibly prompting the end user to unlock secrets that are
319
currently locked (otherwise the secrets may not be loaded).
320
443
- Will merge to 'next'.
321
+ Will merge to 'master'.
322
323
324
* jm/relnotes-2.15-typofix (2017-11-06) 1 commit
447
- - fix typos in 2.15.0 release notes
448
-
449
- Will merge to 'next'.
325
+ (merged to 'next' on 2017-11-07 at 60fc937b62)
326
+ + fix typos in 2.15.0 release notes
327
328
+ Typofix.
329
452
-* sb/blame-config-doc (2017-11-06) 1 commit
453
- - config: document blame configuration
454
-
455
- Will merge to 'next'.
456
-
457
---------------------------------------------------
458
-[Stalled]
459
-
460
-* jn/ssh-wrappers (2017-10-30) 6 commits
461
- - fixup! ssh: 'auto' variant to select between 'ssh' and 'simple'
462
- - ssh: 'simple' variant does not support --port
463
- - ssh: 'simple' variant does not support -4/-6
464
- - ssh: 'auto' variant to select between 'ssh' and 'simple'
465
- - connect: split ssh command line options into separate function
466
- - connect: split git:// setup into a separate function
467
- (this branch uses bw/protocol-v1.)
468
-
469
- The ssh-variant 'simple' introduced earlier broke existing
470
- installations by not passing --port/-4/-6 and not diagnosing an
471
- attempt to pass these as an error. Instead, default to
472
- automatically detect how compatible the GIT_SSH/GIT_SSH_COMMAND is
473
- to OpenSSH convention and then error out an invocation to make it
474
- easier to diagnose connection errors.
475
-
476
- Expecting a reroll.
477
- cf. <20171025163243.xmm7szrkwgblpgcc@aiede.mtv.corp.google.com>
478
-
479
-
480
-* js/for-each-ref-remote-name-and-ref (2017-10-11) 3 commits
481
- - for-each-ref: test :remotename and :remoteref
482
- - for-each-ref: let upstream/push optionally report the remote ref name
483
- - for-each-ref: let upstream/push optionally report the remote name
484
-
485
- The "--format=..." option "git for-each-ref" takes learned to show
486
- the name of the 'remote' repository and the ref at the remote side
487
- that is affected for 'upstream' and 'push' via "%(push:remotename)"
488
- and friends.
489
-
490
- Needs a bit more work on the documentation part.
491
-
492
-
493
-* jc/merge-symlink-ours-theirs (2017-09-26) 1 commit
494
- - merge: teach -Xours/-Xtheirs to symbolic link merge
495
-
496
- "git merge -Xours/-Xtheirs" learned to use our/their version when
497
- resolving a conflicting updates to a symbolic link.
498
-
499
- Needs review.
500
-
501
-
502
-* mg/merge-base-fork-point (2017-09-17) 3 commits
503
- - merge-base: find fork-point outside partial reflog
504
- - merge-base: return fork-point outside reflog
505
- - t6010: test actual test output
506
-
507
- "merge-base --fork-point $branch $commit" is used to guess on which
508
- commit among the commits that were once at the tip of the $branch the
509
- $commit was built on top of, and it learns these historical tips from
510
- the reflog of the $branch. When the true fork-point is lost due to
511
- pruning of old reflog entries, the command does not give any output,
512
- because it has no way to guess correctly and does not want to mislead
513
- the user with a wrong guess.
514
-
515
- The command has been updated to give the best but not known to be
516
- correct guess, based on a hope that a merge-base between $commit and a
517
- virtual merge across all the reflog entries that still are available
518
- for $branch may still be a closer to the true fork-point than the
519
- merge-base between $commit and the current tip of the $branch.
520
-
521
- This may have to be offered by an additional option, to allow the
522
- users that are prepared to see a potentially incorrect guess to opt
523
- into the feature, without affecting the current callers that may not
524
- be prepared to accept a guess that is not known to be correct.
525
-
526
- What's the doneness of this one?
527
-
528
-
529
-* cc/perf-run-config (2017-09-24) 9 commits
530
- - perf: store subsection results in "test-results/$GIT_PERF_SUBSECTION/"
531
- - perf/run: show name of rev being built
532
- - perf/run: add run_subsection()
533
- - perf/run: update get_var_from_env_or_config() for subsections
534
- - perf/run: add get_subsections()
535
- - perf/run: add calls to get_var_from_env_or_config()
536
- - perf/run: add GIT_PERF_DIRS_OR_REVS
537
- - perf/run: add get_var_from_env_or_config()
538
- - perf/run: add '--config' option to the 'run' script
539
-
540
- Needs review.
541
-
542
-
543
-* jk/drop-ancient-curl (2017-08-09) 5 commits
544
- - http: #error on too-old curl
545
- - curl: remove ifdef'd code never used with curl >=7.19.4
546
- - http: drop support for curl < 7.19.4
547
- - http: drop support for curl < 7.16.0
548
- - http: drop support for curl < 7.11.1
549
-
550
- Some code in http.c that has bitrot is being removed.
551
-
552
- Expecting a reroll.
553
-
554
-
555
-* mk/use-size-t-in-zlib (2017-08-10) 1 commit
556
- . zlib.c: use size_t for size
557
-
558
- The wrapper to call into zlib followed our long tradition to use
559
- "unsigned long" for sizes of regions in memory, which have been
560
- updated to use "size_t".
561
-
562
- Needs resurrecting by making sure the fix is good and still applies
563
- (or adjusted to today's codebase).
564
-
565
-
566
-* mg/status-in-progress-info (2017-05-10) 2 commits
567
- - status --short --inprogress: spell it as --in-progress
568
- - status: show in-progress info for short status
569
-
570
- "git status" learns an option to report various operations
571
- (e.g. "merging") that the user is in the middle of.
572
-
573
- cf. <xmqqmvakcdqw.fsf@gitster.mtv.corp.google.com>
330
+ Will merge to 'master'.
331
332
576
-* nd/worktree-move (2017-04-20) 6 commits
577
- - worktree remove: new command
578
- - worktree move: refuse to move worktrees with submodules
579
- - worktree move: accept destination as directory
580
- - worktree move: new command
581
- - worktree.c: add update_worktree_location()
582
- - worktree.c: add validate_worktree()
333
+* sb/blame-config-doc (2017-11-06) 1 commit
334
+ (merged to 'next' on 2017-11-07 at 0576cb452f)
335
+ + config: document blame configuration
336
584
- "git worktree" learned move and remove subcommands.
337
+ Description of blame.{showroot,blankboundary,showemail,date}
338
+ configuration variables have been added to "git config --help".
339
586
- Expecting a reroll.
587
- cf. <20170420101024.7593-1-pclouds@gmail.com>
588
- cf. <20170421145916.mknekgqzhxffu7di@sigill.intra.peff.net>
589
- cf. <d0e81b1e-5869-299e-f462-4d43dc997bd1@ramsayjones.plus.com>
340
+ Will merge to 'master'.
341
591
---------------------------------------------------
592
-[Cooking]
342
343
* ad/5580-unc-tests-on-cygwin (2017-11-01) 1 commit
595
- - t5580: add Cygwin support
344
+ (merged to 'next' on 2017-11-07 at 34fc479da2)
345
+ + t5580: add Cygwin support
346
347
UNC paths are also relevant in Cygwin builds and they are now
348
tested just like Mingw builds.
349
600
- Will merge to 'next'.
350
+ Will merge to 'master'.
351
352
353
* ao/merge-verbosity-getenv-just-once (2017-11-01) 1 commit
@@ -608,8 +358,8 @@ of the repositories listed at
358
Will merge to 'next'.
359
360
611
-* bp/read-index-from-skip-verification (2017-11-01) 1 commit
612
- - read_index_from(): speed index loading by skiping verification of the entry order
361
+* bp/read-index-from-skip-verification (2017-11-08) 1 commit
362
+ - read_index_from(): speed index loading by skipping verification of the entry order
363
364
Drop (perhaps overly cautious) sanity check before using the index
365
read from the filesystem at runtime.
@@ -618,13 +368,14 @@ of the repositories listed at
368
369
370
* rs/hex-to-bytes-cleanup (2017-11-01) 3 commits
621
- - sha1_file: use hex_to_bytes()
622
- - http-push: use hex_to_bytes()
623
- - notes: move hex_to_bytes() to hex.c and export it
371
+ (merged to 'next' on 2017-11-07 at fac14770e1)
372
+ + sha1_file: use hex_to_bytes()
373
+ + http-push: use hex_to_bytes()
374
+ + notes: move hex_to_bytes() to hex.c and export it
375
376
Code cleanup.
377
627
- Will merge to 'next'.
378
+ Will merge to 'master'.
379
380
381
* rs/sequencer-rewrite-file-cleanup (2017-11-02) 3 commits
@@ -650,27 +401,27 @@ of the repositories listed at
401
<commit-ish>:<path> that refers to a given blob object.
402
403
653
-* tb/add-renormalize (2017-10-31) 1 commit
404
+* tb/add-renormalize (2017-11-07) 2 commits
405
+ - SQUASH???
406
- add: introduce "--renormalize"
407
408
"git add --renormalize ." is a new and safer way to record the fact
409
that you are correcting the end-of-line convention and other
410
"convert_to_git()" glitches in the in-repository data.
411
660
- Needs review.
661
-
412
413
* ab/mediawiki-name-truncation (2017-11-01) 1 commit
664
- - remote-mediawiki: limit filenames to legal
414
+ (merged to 'next' on 2017-11-07 at b30233d585)
415
+ + remote-mediawiki: limit filenames to legal
416
417
The remote-helper for talking to MediaWiki has been updated to
418
truncate an overlong pagename so that ".mw" suffix can still be
419
added.
420
670
- Will merge to 'next'.
421
+ Will merge to 'master'.
422
423
673
-* ab/mediawiki-namespace (2017-11-03) 7 commits
424
+* ab/mediawiki-namespace (2017-11-08) 7 commits
425
- remote-mediawiki: show progress while fetching namespaces
426
- remote-mediawiki: process namespaces in order
427
- remote-mediawiki: support fetching from (Main) namespace
@@ -682,42 +433,43 @@ of the repositories listed at
433
The remote-helper for talking to MediaWiki has been updated to
434
work with mediawiki namespaces.
435
685
- Are we expecting another round for v4? What is queued above came
686
- as <20171102212518.1601-8-anarcat@debian.org> and marked v3.
687
- cf. <87375w5qdw.fsf@curie.anarc.at>
436
+ Will merge to 'next'.
437
438
439
* js/mingw-full-version-in-resources (2017-11-01) 1 commit
691
- - mingw: include the full version information in the resources
440
+ (merged to 'next' on 2017-11-07 at 3a256b5ddc)
441
+ + mingw: include the full version information in the resources
442
443
MinGW updates.
444
695
- Will merge to 'next'.
445
+ Will merge to 'master'.
446
447
448
* js/wincred-empty-cred (2017-11-01) 2 commits
699
- - wincred: handle empty username/password correctly
700
- - t0302: check helper can handle empty credentials
449
+ (merged to 'next' on 2017-11-07 at 43d3fcc30a)
450
+ + wincred: handle empty username/password correctly
451
+ + t0302: check helper can handle empty credentials
452
453
MinGW updates.
454
704
- Will merge to 'next'.
455
+ Will merge to 'master'.
456
457
458
* bw/diff-opt-impl-to-bitfields (2017-11-01) 8 commits
708
- - diff: make struct diff_flags members lowercase
709
- - diff: remove DIFF_OPT_CLR macro
710
- - diff: remove DIFF_OPT_SET macro
711
- - diff: remove DIFF_OPT_TST macro
712
- - diff: remove touched flags
713
- - diff: add flag to indicate textconv was set via cmdline
714
- - diff: convert flags to be stored in bitfields
715
- - add, reset: use DIFF_OPT_SET macro to set a diff flag
459
+ (merged to 'next' on 2017-11-07 at 8be78206ba)
460
+ + diff: make struct diff_flags members lowercase
461
+ + diff: remove DIFF_OPT_CLR macro
462
+ + diff: remove DIFF_OPT_SET macro
463
+ + diff: remove DIFF_OPT_TST macro
464
+ + diff: remove touched flags
465
+ + diff: add flag to indicate textconv was set via cmdline
466
+ + diff: convert flags to be stored in bitfields
467
+ + add, reset: use DIFF_OPT_SET macro to set a diff flag
468
469
A single-word "unsigned flags" in the diff options is being split
470
into a structure with many bitfields.
471
720
- Will merge to 'next'.
472
+ Will merge to 'master'.
473
474
475
* mh/tidy-ref-update-flags (2017-11-06) 10 commits
@@ -738,25 +490,27 @@ of the repositories listed at
490
491
492
* ao/diff-populate-filespec-lstat-errorpath-fix (2017-10-29) 1 commit
741
- - diff: fix lstat() error handling in diff_populate_filespec()
493
+ (merged to 'next' on 2017-11-07 at b03241e6e5)
494
+ + diff: fix lstat() error handling in diff_populate_filespec()
495
496
After an error from lstat(), diff_populate_filespec() function
497
sometimes still went ahead and used invalid data in struct stat,
498
which has been fixed.
499
747
- Will merge to 'next'.
500
+ Will merge to 'master'.
501
502
503
* av/fsmonitor (2017-11-01) 4 commits
751
- - fsmonitor: delay updating state until after split index is merged
752
- - fsmonitor: document GIT_TRACE_FSMONITOR
753
- - fsmonitor: don't bother pretty-printing JSON from watchman
754
- - fsmonitor: set the PWD to the top of the working tree
504
+ (merged to 'next' on 2017-11-07 at a06178754d)
505
+ + fsmonitor: delay updating state until after split index is merged
506
+ + fsmonitor: document GIT_TRACE_FSMONITOR
507
+ + fsmonitor: don't bother pretty-printing JSON from watchman
508
+ + fsmonitor: set the PWD to the top of the working tree
509
(this branch uses bp/fsmonitor.)
510
511
Various fixes to bp/fsmonitor topic.
512
759
- Will merge to 'next'.
513
+ Will merge to 'master'.
514
515
516
* bc/submitting-patches-in-asciidoc (2017-10-30) 2 commits
@@ -783,16 +537,13 @@ of the repositories listed at
537
cf. <0102015f5e5ee171-f30f4868-886f-47a1-a4e4-b4936afc545d-000000@eu-west-1.amazonses.com>
538
539
786
-* jc/ignore-cr-at-eol (2017-10-27) 2 commits
540
+* jc/ignore-cr-at-eol (2017-11-08) 2 commits
541
- diff: --ignore-cr-at-eol
542
- xdiff: reassign xpparm_t.flags bits
543
544
The "diff" family of commands learned to ignore differences in
545
carriage return at the end of line.
546
793
- Just a lunch-time hack.
794
- Lacks tests, docs and proper log message.
795
-
547
548
* mh/avoid-rewriting-packed-refs (2017-10-30) 2 commits
549
- files-backend: don't rewrite the `packed-refs` file unnecessarily
@@ -805,30 +556,32 @@ of the repositories listed at
556
Will merge to 'next'.
557
558
808
-* cc/git-packet-pm (2017-10-31) 7 commits
809
- - fixup! Git/Packet.pm: extract parts of t0021/rot13-filter.pl for reuse
559
+* cc/git-packet-pm (2017-11-07) 8 commits
560
- Git/Packet.pm: extract parts of t0021/rot13-filter.pl for reuse
561
- t0021/rot13-filter: add capability functions
562
+ - t0021/rot13-filter: refactor checking final lf
563
- t0021/rot13-filter: add packet_initialize()
564
- t0021/rot13-filter: improve error message
565
- t0021/rot13-filter: improve 'if .. elsif .. else' style
566
- t0021/rot13-filter: refactor packet reading functions
567
+ - t0021/rot13-filter: fix list comparison
568
569
Parts of a test to drive the long-running content filter interface
570
has been split into its own module, hopefully to eventually become
571
reusable.
572
821
- Reroll posted but not picked up yet.
822
- cf. <20171105213836.11717-1-chriscool@tuxfamily.org>
573
+ Will merge to 'next'.
574
575
576
* jm/status-ignored-files-list (2017-10-31) 4 commits
826
- (merged to 'next' on 2017-11-01 at fe51978a30)
577
+ (merged to 'next' on 2017-11-07 at 682c74a2cb)
578
+ status: test ignored modes
579
+ status: document options to show matching ignored files
580
+ status: report matching ignored and normal untracked
581
+ status: add option to show ignored files differently
582
583
+ Originally merged to 'next' on 2017-11-01
584
+
585
The set of paths output from "git status --ignored" was tied
586
closely with its "--untracked=<mode>" option, but now it can be
587
controlled more flexibly. Most notably, a directory that is
@@ -840,20 +593,19 @@ of the repositories listed at
593
594
595
* jc/branch-name-sanity (2017-10-14) 3 commits
843
- (merged to 'next' on 2017-10-16 at 174646d1c3)
844
- + branch: forbid refs/heads/HEAD
845
- + branch: split validate_new_branchname() into two
846
- + branch: streamline "attr_only" handling in validate_new_branchname()
596
+ - branch: forbid refs/heads/HEAD
597
+ - branch: split validate_new_branchname() into two
598
+ - branch: streamline "attr_only" handling in validate_new_branchname()
599
600
"git branch" and "git checkout -b" are now forbidden from creating
601
a branch whose name is "HEAD".
602
851
- Will kick back to 'pu'.
603
+ Reported to cause problems when renaming HEAD during a rebase.
604
+ cf. <49563f7c-354e-334e-03a6-c3a40884b6d0@gmail.com>
605
606
607
* ot/mru-on-list (2017-10-01) 1 commit
855
- (merged to 'next' on 2017-10-07 at e6ad4a16a2)
856
- + mru: use double-linked list from list.h
608
+ - mru: use double-linked list from list.h
609
610
The first step to getting rid of mru API and using the
611
doubly-linked list API directly instead.
@@ -875,11 +627,10 @@ of the repositories listed at
627
628
629
* bp/fsmonitor (2017-10-05) 15 commits
878
- (merged to 'next' on 2017-10-05 at 964a029d13)
630
+ (merged to 'next' on 2017-11-07 at b08d44ba60)
631
+ fsmonitor: preserve utf8 filenames in fsmonitor-watchman log
632
+ fsmonitor: read entirety of watchman output
633
+ fsmonitor: MINGW support for watchman integration
882
- (merged to 'next' on 2017-10-02 at cf0c67979c)
634
+ fsmonitor: add a performance test
635
+ fsmonitor: add a sample integration script for Watchman
636
+ fsmonitor: add test cases for fsmonitor extension
@@ -894,25 +645,26 @@ of the repositories listed at
645
+ bswap: add 64 bit endianness helper get_be64
646
(this branch is used by av/fsmonitor.)
647
648
+ Originally merged to 'next' on 2017-10-05
649
+
650
We learned to talk to watchman to speed up "git status" and other
651
operations that need to see which paths have been modified.
652
900
- Waiting for av/fsmonitor topic.
653
+ Will merge to 'master'.
654
655
656
* bw/protocol-v1 (2017-10-17) 11 commits
904
- (merged to 'next' on 2017-10-18 at 835fe09e50)
905
- + Documentation: document Extra Parameters
906
- + ssh: introduce a 'simple' ssh variant
907
- + i5700: add interop test for protocol transition
908
- + http: tell server that the client understands v1
909
- + connect: tell server that the client understands v1
910
- + connect: teach client to recognize v1 server response
911
- + upload-pack, receive-pack: introduce protocol version 1
912
- + daemon: recognize hidden request arguments
913
- + protocol: introduce protocol extension mechanisms
914
- + pkt-line: add packet_write function
915
- + connect: in ref advertisement, shallows are last
657
+ - Documentation: document Extra Parameters
658
+ - ssh: introduce a 'simple' ssh variant
659
+ - i5700: add interop test for protocol transition
660
+ - http: tell server that the client understands v1
661
+ - connect: tell server that the client understands v1
662
+ - connect: teach client to recognize v1 server response
663
+ - upload-pack, receive-pack: introduce protocol version 1
664
+ - daemon: recognize hidden request arguments
665
+ - protocol: introduce protocol extension mechanisms
666
+ - pkt-line: add packet_write function
667
+ - connect: in ref advertisement, shallows are last
668
(this branch is used by jn/ssh-wrappers.)
669
670
A new mechanism to upgrade the wire protocol in place is proposed
@@ -936,26 +688,22 @@ of the repositories listed at
688
689
690
* sb/submodule-recursive-checkout-detach-head (2017-07-28) 2 commits
939
- (merged to 'next' on 2017-10-26 at 30994b4c76)
940
- + Documentation/checkout: clarify submodule HEADs to be detached
941
- + recursive submodules: detach HEAD from new state
691
+ - Documentation/checkout: clarify submodule HEADs to be detached
692
+ - recursive submodules: detach HEAD from new state
693
694
"git checkout --recursive" may overwrite and rewind the history of
695
the branch that happens to be checked out in submodule
696
repositories, which might not be desirable. Detach the HEAD but
697
still allow the recursive checkout to succeed in such a case.
698
948
- Undecided.
949
- This needs justification in a larger picture; it is unclear why
950
- this is better than rejecting recursive checkout, for example.
699
+ Expecting a reroll.
700
701
--------------------------------------------------
702
[Discarded]
703
704
* jk/ui-color-always-to-auto-maint (2017-10-13) 2 commits
956
- (merged to 'next' on 2017-10-13 at bb16e1edc8)
957
- + color: document that "git -c color.*=always" is a bit special
958
- + color: downgrade "always" to "auto" only for on-disk configuration
705
+ . color: document that "git -c color.*=always" is a bit special
706
+ . color: downgrade "always" to "auto" only for on-disk configuration
707
708
It turns out that "git -c color.ui=always cmd" is relied on by many
709
third-party tools as a way to force coloured output no matter what
@@ -993,7 +741,7 @@ of the repositories listed at
741
742
743
* jh/dir-add-exclude-from-blob (2017-10-27) 1 commit
996
- - dir: allow exclusions from blob in addition to file
744
+ . dir: allow exclusions from blob in addition to file
745
746
The code to read exclusion list from a blob, which is used when the
747
.gitignore file is outside a sparse checkout area, has been