What's cooking (2013/09 #02)
Junio C Hamano committed
Sep 9, 2013 at 15:33 UTC
188f62fcae648718a1ea293a1aa09f0fdadbcf29
1 file changed
+903
-851
whats-cooking.txt
+903
-851
@@ -1,35 +1,17 @@
1
To: git@vger.kernel.org
2
Bcc: lwn@lwn.net
3
-Subject: What's cooking in git.git (Sep 2013, #01; Tue, 3)
4
-X-master-at: 0a41de8f81d8fa39fa17943cd1ed5e0274fef187
5
-X-next-at: 91af5188d5bdbdf006c601a997f25c618050dcee
3
+Subject: What's cooking in git.git (Sep 2013, #02; Mon, 9)
4
+X-master-at: bb80ee09974667a1db6bbc5e33574ed869b76a88
5
+X-next-at: 117d2e3ab92364a7d246c98cc5b71f2b751600e6
6
7
-What's cooking in git.git (Sep 2013, #01; Tue, 3)
7
+What's cooking in git.git (Sep 2013, #02; Mon, 9)
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
-The tip of 'next' has been rewound. I ejected a handful of topics
15
-that have been cooking there while rebuilding it, but it is not
16
-because I found anything in them problematic, but merely because
17
-they were young and I wanted to give their authors a chance to tweak
18
-with a reroll instead of a set of follow-up patches. Unless I hear
19
-otherwise in a few days, they will be merged back to 'next'.
20
-
21
-By the way, the push that overrides the usual "must fast-forward"
22
-was done using the "force-with-lease" option that has been cooking
23
-in next, like so:
24
-
25
- $ git fetch ko next
26
- $ anchor=$(git rev-parse --verify FETCH_HEAD)
27
- $ for remote in ko repo gph github2
28
- do
29
- git push --force-with-lease=refs/heads/next:$anchor $remote next
30
- done
31
-
32
-I used to do this with just --force.
14
+The second batch of topics are now in 'master'.
15
16
You can find the changes described here in the integration branches
17
of the repositories listed at
@@ -39,392 +21,469 @@ of the repositories listed at
21
--------------------------------------------------
22
[Graduated to "master"]
23
42
-* bc/completion-for-bash-3.0 (2013-08-22) 3 commits
43
- (merged to 'next' on 2013-08-28 at fb87a84)
44
- + contrib/git-prompt.sh: handle missing 'printf -v' more gracefully
45
- + t9902-completion.sh: old Bash still does not support array+=('') notation
46
- + git-completion.bash: use correct Bash/Zsh array length syntax
47
-
48
- Originally merged to 'next' on 2013-08-22
49
-
50
- Some people still use rather old versions of bash, which cannot
51
- grok some constructs like 'printf -v varname' the prompt and
52
- completion code started to use recently.
24
+* aj/p4-symlink-lose-nl (2013-08-12) 1 commit
25
+ (merged to 'next' on 2013-08-28 at 5ce9ceb)
26
+ + git-p4: Fix occasional truncation of symlink contents.
27
28
+ Originally merged to 'next' on 2013-08-14
29
55
-* hv/config-from-blob (2013-08-26) 1 commit
56
- (merged to 'next' on 2013-08-28 at 189c6c1)
57
- + config: do not use C function names as struct members
30
59
- Portability fix.
31
+* ap/remote-hg-tilde-is-home-directory (2013-08-09) 1 commit
32
+ (merged to 'next' on 2013-08-28 at 9bfc430)
33
+ + remote-hg: fix path when cloning with tilde expansion
34
61
- Need to apply to 1.8.4.x maintenance track.
35
+ Originally merged to 'next' on 2013-08-14
36
37
64
-* mb/docs-favor-en-us (2013-08-01) 1 commit
65
- (merged to 'next' on 2013-08-28 at 82910e5)
66
- + Provide some linguistic guidance for the documentation.
38
+* bc/unuse-packfile (2013-08-02) 2 commits
39
+ (merged to 'next' on 2013-08-28 at 9029801)
40
+ + Don't close pack fd when free'ing pack windows
41
+ + sha1_file: introduce close_one_pack() to close packs on fd pressure
42
43
Originally merged to 'next' on 2013-08-06
44
70
- Declare that the official grammar & spelling of the source of this
71
- project is en_US, but strongly discourage patches only to "fix"
72
- existing en_UK strings to avoid unnecessary churns.
45
+ Handle memory pressure and file descriptor pressure separately when
46
+ deciding to release pack windows to honor resource limits.
47
48
75
-* mm/war-on-whatchanged (2013-08-13) 2 commits
76
- (merged to 'next' on 2013-08-28 at 25e9c18)
77
- + whatchanged: document its historical nature
78
- + core-tutorial: trim the section on Inspecting Changes
49
+* da/darwin (2013-08-05) 3 commits
50
+ (merged to 'next' on 2013-08-28 at 6e4c5ae)
51
+ + OS X: Fix redeclaration of die warning
52
+ + Makefile: Fix APPLE_COMMON_CRYPTO with BLK_SHA1
53
+ + imap-send: use Apple's Security framework for base64 encoding
54
80
- Originally merged to 'next' on 2013-08-14
55
+ Originally merged to 'next' on 2013-08-05
56
57
83
-* nd/fetch-pack-shallow-fix (2013-08-25) 1 commit
84
- (merged to 'next' on 2013-08-28 at 6f5ed42)
85
- + fetch-pack: do not remove .git/shallow file when --depth is not specified
58
+* db/http-savecookies (2013-08-05) 2 commits
59
+ (merged to 'next' on 2013-08-28 at 22a8464)
60
+ + t5551: Remove header from curl cookie file
61
+ + http: add http.savecookies option to write out HTTP cookies
62
87
- Originally merged to 'next' on 2013-08-27
63
+ Originally merged to 'next' on 2013-08-05
64
89
- Recent "short-cut clone connectivity check" topic broke a shallow
90
- repository when a fetch operation tries to auto-follow tags.
65
92
- Need to apply to 1.8.4.x maintenance track.
66
+* es/blame-L-more (2013-08-05) 11 commits
67
+ (merged to 'next' on 2013-08-28 at b73cdc9)
68
+ + blame: reject empty ranges -L,+0 and -L,-0
69
+ + t8001/t8002: blame: demonstrate acceptance of bogus -L,+0 and -L,-0
70
+ + blame: reject empty ranges -LX,+0 and -LX,-0
71
+ + t8001/t8002: blame: demonstrate acceptance of bogus -LX,+0 and -LX,-0
72
+ + log: fix -L bounds checking bug
73
+ + t4211: retire soon-to-be unimplementable tests
74
+ + t4211: log: demonstrate -L bounds checking bug
75
+ + blame: fix -L bounds checking bug
76
+ + t8001/t8002: blame: add empty file & partial-line tests
77
+ + t8001/t8002: blame: demonstrate -L bounds checking bug
78
+ + t8001/t8002: blame: decompose overly-large test
79
+ (this branch is used by es/blame-L-twice and es/contacts-blame-L-multi.)
80
81
+ Originally merged to 'next' on 2013-08-06
82
95
-* rj/doc-rev-parse (2013-07-22) 2 commits
96
- (merged to 'next' on 2013-08-28 at 0fce928)
97
- + rev-parse(1): logically group options
98
- + rev-parse: remove restrictions on some options
83
+ More fixes to the code to parse the "-L" option in "log" and "blame".
84
100
- Originally merged to 'next' on 2013-07-22
85
86
+* es/blame-L-twice (2013-08-06) 16 commits
87
+ (merged to 'next' on 2013-08-28 at e83ae2e)
88
+ + line-range: reject -L line numbers less than 1
89
+ + t8001/t8002: blame: add tests of -L line numbers less than 1
90
+ + line-range: teach -L^:RE to search from start of file
91
+ + line-range: teach -L:RE to search from end of previous -L range
92
+ + line-range: teach -L^/RE/ to search from start of file
93
+ + line-range-format.txt: document -L/RE/ relative search
94
+ + log: teach -L/RE/ to search from end of previous -L range
95
+ + blame: teach -L/RE/ to search from end of previous -L range
96
+ + line-range: teach -L/RE/ to search relative to anchor point
97
+ + blame: document multiple -L support
98
+ + t8001/t8002: blame: add tests of multiple -L options
99
+ + blame: accept multiple -L ranges
100
+ + blame: inline one-line function into its lone caller
101
+ + range-set: publish API for re-use by git-blame -L
102
+ + line-range-format.txt: clarify -L:regex usage form
103
+ + git-log.txt: place each -L option variation on its own line
104
+ (this branch is used by es/contacts-blame-L-multi; uses es/blame-L-more.)
105
103
-* rt/doc-merge-file-diff3 (2013-08-09) 1 commit
104
- (merged to 'next' on 2013-08-28 at d6b8006)
105
- + Documentation/git-merge-file: document option "--diff3"
106
+ Originally merged to 'next' on 2013-08-08
107
107
- Originally merged to 'next' on 2013-08-14
108
+ Teaches "git blame" to take more than one -L ranges.
109
110
110
-* sp/doc-smart-http (2013-08-21) 1 commit
111
- (merged to 'next' on 2013-08-28 at 20a6877)
112
- + Document the HTTP transport protocols
111
+* es/contacts-blame-L-multi (2013-08-13) 3 commits
112
+ (merged to 'next' on 2013-08-28 at 24ff8b4)
113
+ + contacts: reduce git-blame invocations
114
+ + contacts: gather all blame sources prior to invoking git-blame
115
+ + contacts: validate hunk length earlier
116
+ (this branch uses es/blame-L-more and es/blame-L-twice.)
117
114
- Originally merged to 'next' on 2013-08-22
118
+ Originally merged to 'next' on 2013-08-14
119
116
---------------------------------------------------
117
-[New Topics]
120
119
-* fc/contrib-bzr-hg-fixes (2013-09-03) 10 commits
120
- - contrib/remote-helpers: quote variable references in redirection targets
121
- - contrib/remote-helpers: style updates for test scripts
122
- - remote-hg: use notes to keep track of Hg revisions
123
- - remote-helpers: cleanup more global variables
124
- - remote-helpers: trivial style fixes
125
- - remote-hg: improve basic test
126
- - remote-hg: add missing &&s in the test
127
- - remote-hg: fix test
128
- - remote-bzr: make bzr branches configurable per-repo
129
- - remote-bzr: fix export of utf-8 authors
121
+* fc/remote-hg-shared-setup (2013-08-11) 2 commits
122
+ (merged to 'next' on 2013-08-28 at 41ee3fe)
123
+ + remote-hg: add shared repo upgrade
124
+ + remote-hg: ensure shared repo is initialized
125
131
- Will merge to 'next'.
126
+ Originally merged to 'next' on 2013-08-14
127
128
134
-* jc/pager-configuration-doc (2013-08-29) 1 commit
135
- - config: rewrite core.pager documentation
129
+* fc/unpack-trees-leakfix (2013-08-13) 1 commit
130
+ (merged to 'next' on 2013-08-28 at 3ad1f21)
131
+ + unpack-trees: plug a memory leak
132
137
- Will merge to 'next'.
133
+ Originally merged to 'next' on 2013-08-14
134
135
140
-* np/pack-v4 (2013-09-03) 24 commits
141
- - Document pack v4 format
142
- - initial pack index v3 support on the read side
143
- - pack v4: add progress display
144
- - pack v4: normalize pack name to properly generate the pack index file name
145
- - pack index v3
146
- - pack v4: relax commit parsing a bit
147
- - pack v4: honor pack.compression config option
148
- - pack v4: load delta candidate for encoding tree objects
149
- - pack v4: tree object delta encoding
150
- - pack v4: object writing
151
- - pack v4: object data copy
152
- - pack v4: object headers
153
- - pack v4: creation code
154
- - pack v4: dictionary table output
155
- - pack v4: tree object encoding
156
- - pack v4: commit object encoding
157
- - pack v4: basic references encoding
158
- - pack v4: move to struct pack_idx_entry and get rid of our own struct idx_entry
159
- - pack v4: split the object list and dictionary creation
160
- - pack v4: add commit object parsing
161
- - pack v4: add tree entry mode support to dictionary entries
162
- - pack v4: scan tree objects
163
- - export packed_object_info()
164
- - pack v4: initial pack dictionary structure and code
136
+* jc/check-x-z (2013-07-11) 4 commits
137
+ (merged to 'next' on 2013-08-28 at a7670ce)
138
+ + check-attr -z: a single -z should apply to both input and output
139
+ + check-ignore -z: a single -z should apply to both input and output
140
+ + check-attr: the name of the character is NUL, not NULL
141
+ + check-ignore: the name of the character is NUL, not NULL
142
143
+ Originally merged to 'next' on 2013-07-15
144
167
-* jk/free-tree-buffer (2013-06-06) 1 commit
168
- - clear parsed flag when we free tree buffers
145
+ "git check-ignore -z" applied the NUL termination to both its input
146
+ (with --stdin) and its output, but "git check-attr -z" ignored the
147
+ option on the output side.
148
170
- Will merge to 'next'.
149
+ This is potentially a backward incompatible fix. Let's see if
150
+ anybody screams before deciding if we want to do anything to help
151
+ existing users (there may be none).
152
153
173
-* jk/has-sha1-file-retry-packed (2013-08-30) 1 commit
174
- - has_sha1_file: re-check pack directory before giving up
154
+* jc/diff-filter-negation (2013-07-19) 6 commits
155
+ (merged to 'next' on 2013-08-28 at 5cc7720)
156
+ + diff: deprecate -q option to diff-files
157
+ + diff: allow lowercase letter to specify what change class to exclude
158
+ + diff: reject unknown change class given to --diff-filter
159
+ + diff: preparse --diff-filter string argument
160
+ + diff: factor out match_filter()
161
+ + diff: pass the whole diff_options to diffcore_apply_filter()
162
+ (this branch is used by jc/hold-diff-remove-q-synonym-for-no-deletion.)
163
176
- Will merge to 'next'.
164
+ Originally merged to 'next' on 2013-07-22
165
166
+ Teach "git diff --diff-filter" to express "I do not want to see
167
+ these classes of changes" more directly by listing only the
168
+ unwanted ones in lowercase (e.g. "--diff-filter=d" will show
169
+ everything but deletion) and deprecate "diff-files -q" which did
170
+ the same thing as "--diff-filter=d".
171
179
-* ab/gitweb-author-initials (2013-08-30) 1 commit
180
- - gitweb: Fix the author initials in blame for non-ASCII names
172
182
- Will merge to 'next'.
173
+* jc/parseopt-command-modes (2013-07-30) 2 commits
174
+ (merged to 'next' on 2013-08-28 at 738b399)
175
+ + tag: use OPT_CMDMODE
176
+ + parse-options: add OPT_CMDMODE()
177
+ (this branch is used by sb/parseopt-boolean-removal.)
178
179
+ Many commands use --dashed-option as a operation mode selector
180
+ (e.g. "git tag --delete") that the user can use at most one
181
+ (e.g. "git tag --delete --verify" is a nonsense) and you cannot
182
+ negate (e.g. "git tag --no-delete" is a nonsense). Make it easier
183
+ for users of parse_options() to enforce these restrictions.
184
185
-* bk/refs-multi-update (2013-08-30) 8 commits
186
- - update-ref: add test cases covering --stdin signature
187
- - update-ref: support multiple simultaneous updates
188
- - refs: add update_refs for multiple simultaneous updates
189
- - refs: add function to repack without multiple refs
190
- - refs: factor delete_ref loose ref step into a helper
191
- - refs: factor update_ref steps into helpers
192
- - refs: report ref type from lock_any_ref_for_update
193
- - reset: rename update_refs to reset_refs
185
186
+* jc/push-cas (2013-08-02) 9 commits
187
+ (merged to 'next' on 2013-08-28 at 784281f)
188
+ + push: teach --force-with-lease to smart-http transport
189
+ + send-pack: fix parsing of --force-with-lease option
190
+ + t5540/5541: smart-http does not support "--force-with-lease"
191
+ + t5533: test "push --force-with-lease"
192
+ + push --force-with-lease: tie it all together
193
+ + push --force-with-lease: implement logic to populate old_sha1_expect[]
194
+ + remote.c: add command line option parser for "--force-with-lease"
195
+ + builtin/push.c: use OPT_BOOL, not OPT_BOOLEAN
196
+ + cache.h: move remote/connect API out of it
197
196
-* fc/at-head (2013-09-03) 2 commits
197
- - Add new @ shortcut for HEAD
198
- - sha1-name: pass len argument to interpret_branch_name()
198
+ Originally merged to 'next' on 2013-08-05
199
200
- Attempt to resurrect "Type @ for HEAD"; the bottom one seems to be
201
- a genuine code improvement, but identifying cases where "@" means
202
- HEAD seems to be harder than it seems---the tip one does not yet
203
- get it right.
200
+ Allow a safer "rewind of the remote tip" push than blind "--force",
201
+ by requiring that the overwritten remote ref to be unchanged since
202
+ the new history to replace it was prepared.
203
204
+ The machinery is more or less ready. The "--force" option is again
205
+ the big red button to override any safety, thanks to J6t's sanity
206
+ (the original round allowed --lockref to defeat --force).
207
206
-* fc/fast-export (2013-09-03) 2 commits
207
- - fast-export: refactor get_tags_and_duplicates()
208
- - fast-export: make extra_refs global
208
+ The logic to choose the default implemented here is fragile
209
+ (e.g. "git fetch" after seeing a failure will update the
210
+ remote-tracking branch and will make the next "push" pass,
211
+ defeating the safety pretty easily). It is suitable only for the
212
+ simplest workflows, and it may hurt users more than it helps them.
213
210
- Code simpification.
214
212
- Will merge to 'next'.
215
+* jc/transport-do-not-use-connect-twice-in-fetch (2013-08-28) 6 commits
216
+ (merged to 'next' on 2013-08-29 at f88a6da)
217
+ + builtin/fetch.c: Fix a sparse warning
218
+ (merged to 'next' on 2013-08-28 at 2ed86df)
219
+ + fetch: work around "transport-take-over" hack
220
+ + fetch: refactor code that fetches leftover tags
221
+ + fetch: refactor code that prepares a transport
222
+ + fetch: rename file-scope global "transport" to "gtransport"
223
+ + t5802: add test for connect helper
224
225
+ Originally merged to 'next' on 2013-08-20
226
215
-* fc/rev-parse-test-updates (2013-09-03) 4 commits
216
- - rev-parse test: use standard test functions for setup
217
- - rev-parse test: use test_cmp instead of "test" builtin
218
- - rev-parse test: use test_must_fail, not "if <command>; then false; fi"
219
- - rev-parse test: modernize quoting and whitespace
227
+ The auto-tag-following code in "git fetch" tries to reuse the same
228
+ transport twice when the serving end does not cooperate and does
229
+ not give tags that point to commits that are asked for as part of
230
+ the primary transfer. Unfortunately, Git-aware transport helper
231
+ interface is not designed to be used more than once, hence this
232
+ does not work over smart-http transfer.
233
221
- Modernize tests.
234
223
- Will merge to 'next'.
235
+* jc/url-match (2013-08-09) 7 commits
236
+ (merged to 'next' on 2013-08-28 at 2a9fd23)
237
+ + builtin/config.c: compilation fix
238
+ + config: "git config --get-urlmatch" parses section.<url>.key
239
+ + builtin/config: refactor collect_config()
240
+ + config: parse http.<url>.<variable> using urlmatch
241
+ + config: add generic callback wrapper to parse section.<url>.key
242
+ + config: add helper to normalize and match URLs
243
+ + http.c: fix parsing of http.sslCertPasswordProtected variable
244
245
+ Originally merged to 'next' on 2013-08-12
246
226
-* fc/t3200-fixes (2013-09-03) 3 commits
227
- - t: branch: fix broken && chains
228
- - t: branch: fix typo
229
- - t: branch: trivial style fix
247
+ Allow section.<urlpattern>.var configuration variables to be
248
+ treated as a "virtual" section.var given a URL, and use the
249
+ mechanism to enhance http.* configuration variables.
250
231
- Will merge to 'next'.
251
+ This is a reroll of Kyle J. McKay's work.
252
253
234
-* fc/trivial (2013-09-03) 5 commits
235
- - branch: use $curr_branch_short more
236
- - add: trivial style cleanup
237
- - reset: trivial style cleanup
238
- - branch: trivial style fix
239
- - reset: trivial refactoring
254
+* jk/cat-file-batch-optim (2013-08-05) 1 commit
255
+ (merged to 'next' on 2013-08-28 at 624e60d)
256
+ + cat-file: only split on whitespace when %(rest) is used
257
241
- Will merge to 'next'.
258
+ Originally merged to 'next' on 2013-08-06
259
260
+ Rework the reverted change to `cat-file --batch-check`.
261
244
-* jc/ref-excludes (2013-09-03) 2 commits
245
- - document --exclude option
246
- - revision: introduce --exclude=<glob> to tame wildcards
262
263
+* jk/fast-import-empty-ls (2013-06-23) 4 commits
264
+ (merged to 'next' on 2013-08-28 at 876a39d)
265
+ + fast-import: allow moving the root tree
266
+ + fast-import: allow ls or filecopy of the root tree
267
+ + fast-import: set valid mode on root tree in "ls"
268
+ + t9300: document fast-import empty path issues
269
249
-* jk/pager-bypass-cat-for-default-pager (2013-09-03) 1 commit
250
- - pager: turn on "cat" optimization for DEFAULT_PAGER
270
+ Originally merged to 'next' on 2013-07-22
271
252
- If a build-time fallback is set to "cat" instead of "less", we
253
- should apply the same "no subprocess or pipe" optimization as we
254
- apply to user-supplied GIT_PAGER=cat.
272
256
- Will merge to 'next'.
273
+* jk/mailmap-incomplete-line (2013-08-28) 1 commit
274
+ (merged to 'next' on 2013-09-04 at e77e887)
275
+ + mailmap: handle mailmap blobs without trailing newlines
276
277
+ Will merge later to 'maint'.
278
259
-* nd/git-dir-pointing-at-gitfile (2013-09-03) 1 commit
260
- - Make setup_git_env() resolve .git file when $GIT_DIR is not specified
279
262
- We made sure that we notice the user-supplied GIT_DIR is actually a
263
- gitfile, but failed to do so when the default ".git" is a gitfile.
280
+* jl/some-submodule-config-are-not-boolean (2013-08-19) 1 commit
281
+ (merged to 'next' on 2013-08-28 at 7fabc3e)
282
+ + avoid segfault on submodule.*.path set to an empty "true"
283
265
- Will merge to 'next'.
284
+ Originally merged to 'next' on 2013-08-20
285
286
268
-* nr/git-cd-to-a-directory (2013-09-03) 1 commit
269
- - git: run in a directory given with -C option
287
+* jl/submodule-mv (2013-08-06) 5 commits
288
+ (merged to 'next' on 2013-08-28 at 4ec6b41)
289
+ + rm: delete .gitmodules entry of submodules removed from the work tree
290
+ + mv: update the path entry in .gitmodules for moved submodules
291
+ + submodule.c: add .gitmodules staging helper functions
292
+ + mv: move submodules using a gitfile
293
+ + mv: move submodules together with their work trees
294
+ (this branch is tangled with nd/magic-pathspec.)
295
271
- Just like "make -C <directory>", make "git -C <directory> ..." to
272
- go there before doing anything else.
296
+ Originally merged to 'next' on 2013-08-08
297
298
+ "git mv A B" when moving a submodule A does "the right thing",
299
+ inclusing relocating its working tree and adjusting the paths in
300
+ the .gitmodules file.
301
275
-* rh/peeling-tag-to-tag (2013-09-03) 2 commits
276
- - peel_onion: do not assume length of x_type globals
277
- - peel_onion(): add support for <rev>^{tag}
302
279
- Make "foo^{tag}" to peel a tag to itself, i.e. no-op., and fail if
280
- "foo" is not a tag.
303
+* km/svn-1.8-serf-only (2013-07-18) 3 commits
304
+ (merged to 'next' on 2013-08-28 at 1119134)
305
+ + Git.pm: revert _temp_cache use of temp_is_locked
306
+ + git-svn: allow git-svn fetching to work using serf
307
+ + Git.pm: add new temp_is_locked function
308
282
---------------------------------------------------
283
-[Stalled]
309
+ Originally merged to 'next' on 2013-07-19
310
285
-* rv/send-email-cache-generated-mid (2013-08-21) 2 commits
286
- - git-send-email: Cache generated message-ids, use them when prompting
287
- - git-send-email: add optional 'choices' parameter to the ask sub
311
+ Subversion 1.8.0 that was recently released breaks older subversion
312
+ clients coming over http/https in various ways.
313
314
290
-* rj/read-default-config-in-show-ref-pack-refs (2013-06-17) 3 commits
291
- - ### DONTMERGE: needs better explanation on what config they need
292
- - pack-refs.c: Add missing call to git_config()
293
- - show-ref.c: Add missing call to git_config()
315
+* mm/no-shell-escape-in-die-message (2013-08-07) 1 commit
316
+ (merged to 'next' on 2013-08-28 at 86b300f)
317
+ + die_with_status: use "printf '%s\n'", not "echo"
318
295
- The changes themselves are probably good, but it is unclear what
296
- basic setting needs to be read for which exact operation.
319
+ Originally merged to 'next' on 2013-08-08
320
298
- Waiting for clarification.
299
- $gmane/228294
321
+ Fixes a minor bug in "git rebase -i" (there could be others, as the
322
+ root cause is pretty generic) where the code feeds a random, data
323
+ dependeant string to 'echo' and expects it to come out literally.
324
325
302
-* jh/shorten-refname (2013-05-07) 4 commits
303
- - t1514: refname shortening is done after dereferencing symbolic refs
304
- - shorten_unambiguous_ref(): Fix shortening refs/remotes/origin/HEAD to origin
305
- - t1514: Demonstrate failure to correctly shorten "refs/remotes/origin/HEAD"
306
- - t1514: Add tests of shortening refnames in strict/loose mode
326
+* ms/fetch-prune-configuration (2013-07-18) 1 commit
327
+ (merged to 'next' on 2013-08-28 at fa6ffc7)
328
+ + fetch: make --prune configurable
329
308
- When remotes/origin/HEAD is not a symbolic ref, "rev-parse
309
- --abbrev-ref remotes/origin/HEAD" ought to show "origin", not
310
- "origin/HEAD", which is fixed with this series (if it is a symbolic
311
- ref that points at remotes/origin/something, then it should show
312
- "origin/something" and it already does).
330
+ Originally merged to 'next' on 2013-07-21
331
314
- Expecting a reroll, as an early part of a larger series.
315
- $gmane/225137
332
+ Allow fetch.prune and remote.*.prune configuration variables to be set,
333
+ and "git fetch" to behave as if "--prune" is given.
334
335
+ "git fetch" that honors remote.*.prune is fine, but I wonder if we
336
+ should somehow make "git push" aware of it as well. Perhaps
337
+ remote.*.prune should not be just a boolean, but a 4-way "none",
338
+ "push", "fetch", "both"?
339
318
-* mg/more-textconv (2013-05-10) 7 commits
319
- - grep: honor --textconv for the case rev:path
320
- - grep: allow to use textconv filters
321
- - t7008: demonstrate behavior of grep with textconv
322
- - cat-file: do not die on --textconv without textconv filters
323
- - show: honor --textconv for blobs
324
- - diff_opt: track whether flags have been set explicitly
325
- - t4030: demonstrate behavior of show with textconv
340
327
- Make "git grep" and "git show" pay attention to --textconv when
328
- dealing with blob objects.
341
+* nd/clone-connectivity-shortcut (2013-07-23) 1 commit
342
+ (merged to 'next' on 2013-08-28 at d227f9f)
343
+ + smart http: use the same connectivity check on cloning
344
330
- I thought this was pretty well designed and executed, but it seems
331
- there are some doubts on the list; kicked back to 'pu'.
345
+ Originally merged to 'next' on 2013-07-30
346
347
334
-* jc/format-patch (2013-04-22) 2 commits
335
- - format-patch: --inline-single
336
- - format-patch: rename "no_inline" field
348
+* nd/gc-lock-against-each-other (2013-08-09) 1 commit
349
+ (merged to 'next' on 2013-08-28 at 46e072b)
350
+ + gc: reject if another gc is running, unless --force is given
351
338
- A new option to send a single patch to the standard output to be
339
- appended at the bottom of a message. I personally have no need for
340
- this, but it was easy enough to cobble together. Tests, docs and
341
- stripping out more MIMEy stuff are left as exercises to interested
342
- parties.
352
+ Originally merged to 'next' on 2013-08-14
353
344
- Not ready for inclusion.
354
346
- Will discard unless we hear from anybody who is interested in
347
- tying its loose ends.
355
+* nd/magic-pathspec (2013-09-05) 50 commits
356
+ (merged to 'next' on 2013-09-05 at b832b8d)
357
+ + add: lift the pathspec magic restriction on "add -p"
358
+ + pathspec: catch prepending :(prefix) on pathspec with short magic
359
+ (merged to 'next' on 2013-08-28 at 305c6e9)
360
+ + rm: do not set a variable twice without intermediate reading.
361
+ + t6131 - skip tests if on case-insensitive file system
362
+ + parse_pathspec: accept :(icase)path syntax
363
+ + pathspec: support :(glob) syntax
364
+ + pathspec: make --literal-pathspecs disable pathspec magic
365
+ + pathspec: support :(literal) syntax for noglob pathspec
366
+ + kill limit_pathspec_to_literal() as it's only used by parse_pathspec()
367
+ + parse_pathspec: preserve prefix length via PATHSPEC_PREFIX_ORIGIN
368
+ + parse_pathspec: make sure the prefix part is wildcard-free
369
+ + rename field "raw" to "_raw" in struct pathspec
370
+ + tree-diff: remove the use of pathspec's raw[] in follow-rename codepath
371
+ + remove match_pathspec() in favor of match_pathspec_depth()
372
+ + remove init_pathspec() in favor of parse_pathspec()
373
+ + remove diff_tree_{setup,release}_paths
374
+ + convert common_prefix() to use struct pathspec
375
+ + convert add_files_to_cache to take struct pathspec
376
+ + convert {read,fill}_directory to take struct pathspec
377
+ + convert refresh_index to take struct pathspec
378
+ + convert report_path_error to take struct pathspec
379
+ + checkout: convert read_tree_some to take struct pathspec
380
+ + convert unmerge_cache to take struct pathspec
381
+ + convert run_add_interactive to use struct pathspec
382
+ + convert read_cache_preload() to take struct pathspec
383
+ + line-log: convert to use parse_pathspec
384
+ + reset: convert to use parse_pathspec
385
+ + add: convert to use parse_pathspec
386
+ + check-ignore: convert to use parse_pathspec
387
+ + archive: convert to use parse_pathspec
388
+ + ls-files: convert to use parse_pathspec
389
+ + rm: convert to use parse_pathspec
390
+ + checkout: convert to use parse_pathspec
391
+ + rerere: convert to use parse_pathspec
392
+ + status: convert to use parse_pathspec
393
+ + commit: convert to use parse_pathspec
394
+ + clean: convert to use parse_pathspec
395
+ + guard against new pathspec magic in pathspec matching code
396
+ + parse_pathspec: support prefixing original patterns
397
+ + parse_pathspec: support stripping/checking submodule paths
398
+ + parse_pathspec: support stripping submodule trailing slashes
399
+ + parse_pathspec: add special flag for max_depth feature
400
+ + convert some get_pathspec() calls to parse_pathspec()
401
+ + parse_pathspec: add PATHSPEC_PREFER_{CWD,FULL} flags
402
+ + parse_pathspec: save original pathspec for reporting
403
+ + add parse_pathspec() that converts cmdline args to struct pathspec
404
+ + pathspec: add copy_pathspec
405
+ + pathspec: i18n-ize error strings in pathspec parsing code
406
+ + move struct pathspec and related functions to pathspec.[ch]
407
+ + clean: remove unused variable "seen"
408
+ (this branch is tangled with jl/submodule-mv.)
409
410
+ Originally merged to 'next' on 2013-07-23
411
350
-* jk/gitweb-utf8 (2013-04-08) 4 commits
351
- - gitweb: Fix broken blob action parameters on blob/commitdiff pages
352
- - gitweb: Don't append ';js=(0|1)' to external links
353
- - gitweb: Make feed title valid utf8
354
- - gitweb: Fix utf8 encoding for blob_plain, blobdiff_plain, commitdiff_plain, and patch
412
+ Use "struct pathspec" interface in more places, instead of array of
413
+ characters, the latter of which cannot express magic pathspecs
414
+ (e.g. ":(icase)makefile" that matches both Makefile and makefile).
415
356
- Various fixes to gitweb.
416
358
- Drew Northup volunteered to take a look into this.
359
- $gmane/226216
417
+* nd/sq-quote-buf (2013-07-30) 3 commits
418
+ (merged to 'next' on 2013-08-28 at d0ffd55)
419
+ + quote: remove sq_quote_print()
420
+ + tar-tree: remove dependency on sq_quote_print()
421
+ + for-each-ref, quote: convert *_quote_print -> *_quote_buf
422
423
+ Originally merged to 'next' on 2013-08-01
424
362
-* jc/show-branch (2013-06-07) 5 commits
363
- - show-branch: use commit slab to represent bitflags of arbitrary width
364
- - show-branch.c: remove "all_mask"
365
- - show-branch.c: abstract out "flags" operation
366
- - show-branch.c: lift all_mask/all_revs to a global static
367
- - show-branch.c: update comment style
425
+ Code simplification as a preparatory step to something larger.
426
369
- Waiting for the final step to lift the hard-limit before sending it out.
427
371
---------------------------------------------------
372
-[Cooking]
428
+* rr/feed-real-path-to-editor (2013-07-29) 1 commit
429
+ (merged to 'next' on 2013-08-28 at 7458fa4)
430
+ + editor: use canonicalized absolute path
431
374
-* cc/replace-with-the-same-type (2013-08-27) 5 commits
375
- - Documentation/replace: add Creating Replacement Objects section
376
- - t6050-replace: add test to clean up all the replace refs
377
- - t6050-replace: test that objects are of the same type
378
- - Documentation/replace: state that objects must be of the same type
379
- - replace: forbid replacing an object with one of a different type
432
+ Originally merged to 'next' on 2013-07-30
433
381
- Using the replace mechanism to swap an object with another object
382
- of a different type can introduce inconsistency (e.g. a tree
383
- expects an object name to refer to a blob, but the blob object can
384
- be mistakenly or maliciously replaced with an object with a
385
- different type). Attempt to forbid such.
434
387
- This may need to be given an escape hatch "--force", though.
435
+* sb/diff-delta-remove-needless-comparison (2013-08-18) 1 commit
436
+ (merged to 'next' on 2013-08-28 at ba09c3f)
437
+ + create_delta_index: simplify condition always evaluating to true
438
439
+ Originally merged to 'next' on 2013-08-20
440
390
-* jx/clean-interactive (2013-08-28) 1 commit
391
- - documentation: clarify notes for clean.requireForce
441
393
- Finishing touches to update the document to adjust to a new option
394
- "git clean" learned recently.
442
+* sb/misc-cleanup (2013-08-09) 3 commits
443
+ (merged to 'next' on 2013-08-28 at bc2045b)
444
+ + rm: remove unneeded null pointer check
445
+ + diff: fix a possible null pointer dereference
446
+ + diff: remove ternary operator evaluating always to true
447
396
- Will merge to 'next'.
448
+ Originally merged to 'next' on 2013-08-14
449
450
399
-* mm/status-without-comment-char (2013-08-29) 5 commits
400
- - status: introduce status.displayCommentChar to disable display of #
401
- - SQUASH: do not fprintf() random strings
402
- - get rid of "git submodule summary --for-status"
403
- - wt-status: use argv_array API
404
- - builtin/stripspace.c: fix broken indentation
451
+* sb/parseopt-boolean-removal (2013-08-07) 9 commits
452
+ (merged to 'next' on 2013-08-28 at 2f3f21f)
453
+ + revert: use the OPT_CMDMODE for parsing, reducing code
454
+ + checkout-index: fix negations of even numbers of -n
455
+ + config parsing options: allow one flag multiple times
456
+ + hash-object: replace stdin parsing OPT_BOOLEAN by OPT_COUNTUP
457
+ + branch, commit, name-rev: ease up boolean conditions
458
+ + checkout: remove superfluous local variable
459
+ + log, format-patch: parsing uses OPT__QUIET
460
+ + Replace deprecated OPT_BOOLEAN by OPT_BOOL
461
+ + Remove deprecated OPTION_BOOLEAN for parsing arguments
462
+ (this branch uses jc/parseopt-command-modes.)
463
406
- Allow "git status" to omit the prefix to make its output a comment
407
- in a commit log editor, which is not necessary for human
408
- consumption.
464
+ Convert most uses of OPT_BOOLEAN/OPTION_BOOLEAN that can use
465
+ OPT_BOOL/OPTION_BOOLEAN which have much saner semantics, and turn
466
+ remaining ones into OPT_SET_INT, OPT_COUNTUP, etc. as necessary.
467
468
411
-* js/xread-in-full (2013-08-20) 1 commit
412
- - stream_to_pack: xread does not guarantee to read all requested bytes
469
+* sg/bash-prompt-lf-in-cwd-test (2013-08-18) 1 commit
470
+ (merged to 'next' on 2013-08-28 at eb449b5)
471
+ + bash prompt: test the prompt with newline in repository path
472
473
Originally merged to 'next' on 2013-08-20
474
416
- A call to xread() was used without a loop around to cope with short
417
- read in the codepath to stream new contents to a pack.
418
-
419
- Will merge to 'next'.
420
-
475
422
-* sb/mailmap-freeing-NULL-is-ok (2013-08-20) 1 commit
423
- - mailmap: remove redundant check for freeing memory
476
+* sp/clip-read-write-to-8mb (2013-08-20) 2 commits
477
+ (merged to 'next' on 2013-08-28 at c9840d7)
478
+ + Revert "compat/clipped-write.c: large write(2) fails on Mac OS X/XNU"
479
+ + xread, xwrite: limit size of IO to 8MB
480
425
- Originally merged to 'next' on 2013-08-20
481
+ Originally merged to 'next' on 2013-08-22
482
427
- Will merge to 'next'.
483
+ Send a large request to read(2)/write(2) as a smaller but still
484
+ reasonably large chunks, which would improve the latency when the
485
+ operation needs to be killed and incidentally works around broken
486
+ 64-bit systems that cannot take a 2GB write or read in one go.
487
488
489
* tg/index-struct-sizes (2013-08-20) 1 commit
@@ -436,799 +495,828 @@ of the repositories listed at
495
The code that reads from a region that mmaps an on-disk index
496
assumed that "int"/"short" are always 32/16 bits.
497
439
- Will merge to 'master' in the second batch.
498
499
+* tr/fd-gotcha-fixes (2013-08-06) 1 commit
500
+ (merged to 'next' on 2013-08-28 at 5305271)
501
+ + t0070: test that git_mkstemps correctly checks return value of open()
502
+
503
+ Originally merged to 'next' on 2013-08-08
504
+
505
+ Finishing touches to an earlier fix already in 'master'.
506
+
507
+
508
+* tr/log-full-diff-keep-true-parents (2013-08-05) 2 commits
509
+ (merged to 'next' on 2013-08-28 at e86ec11)
510
+ + log: use true parents for diff when walking reflogs
511
+ + log: use true parents for diff even when rewriting
512
+
513
+ Originally merged to 'next' on 2013-08-08
514
+
515
+ Output from "git log --full-diff -- <pathspec>" looked strange,
516
+ because comparison was done with the previous ancestor that touched
517
+ the specified <pathspec>, causing the patches for paths outside the
518
+ pathspec to show more than the single commit has changed.
519
+
520
+ Tweak "git reflog -p" for the same reason using the same mechanism.
521
+
522
+--------------------------------------------------
523
+[New Topics]
524
442
-* bc/submodule-status-ignored (2013-09-03) 2 commits
443
- . submodule: don't print status output with ignore=all
444
- . submodule: fix confusing variable name
525
+* js/add-i-mingw (2013-09-04) 1 commit
526
+ - add--interactive: fix external command invocation on Windows
527
446
- Originally merged to 'next' on 2013-08-22
528
+ The implementation of "add -i" has a crippling code to work around
529
+ ActiveState Perl limitation but it by mistake also triggered on Git
530
+ for Windows where MSYS perl is used.
531
448
- On hold; this conflicts and breaks tests when used with
449
- mm/status-without-comment-char.
532
+ Will merge to 'next'.
533
534
452
-* jk/config-int-range-check (2013-08-21) 2 commits
453
- - teach git-config to output large integers
454
- - config: properly range-check integer values
535
+* rh/ishes-doc (2013-09-04) 7 commits
536
+ - glossary: fix and clarify the definition of 'ref'
537
+ - revisions.txt: fix and clarify <rev>^{<type>}
538
+ - glossary: more precise definition of tree-ish (a.k.a. treeish)
539
+ - use 'commit-ish' instead of 'committish'
540
+ - use 'tree-ish' instead of 'treeish'
541
+ - glossary: define commit-ish (a.k.a. committish)
542
+ - glossary: mention 'treeish' as an alternative to 'tree-ish'
543
456
- Originally merged to 'next' on 2013-08-22
544
+ We liberally use "committish" and "commit-ish" (and "treeish" and
545
+ "tree-ish"); as these are non-words, let's unify these terms to
546
+ their dashed form. More importantly, clarify the documentation on
547
+ object peeling using these terms.
548
458
- "git config --int section.var 3g" should somehow diagnose that the
459
- number does not fit in "int" (on 32-bit platforms anyway) but it
460
- did not.
549
+ Will merge to 'next'.
550
551
463
-* jk/duplicate-objects-in-packs (2013-08-24) 6 commits
464
- - default pack.indexDuplicates to false
465
- - index-pack: optionally reject packs with duplicate objects
466
- - test index-pack on packs with recoverable delta cycles
467
- - add tests for indexing packs with delta cycles
468
- - sha1-lookup: handle duplicate keys with GIT_USE_LOOKUP
469
- - test-sha1: add a binary output mode
552
+* jc/commit-is-spelled-with-two-ems (2013-09-05) 2 commits
553
+ (merged to 'next' on 2013-09-05 at 982aef2)
554
+ + typofix: cherry is spelled with two ars
555
+ + typofix: commit is spelled with two ems
556
471
- A packfile that stores the same object more than once is broken and
472
- will be rejected.
557
+ Will merge to 'master'.
558
474
- Will merge to 'next'.
559
560
+* bc/http-backend-allow-405 (2013-09-09) 1 commit
561
+ - http-backend: provide Allow header for 405
562
477
-* mm/mediawiki-dumb-push-fix (2013-09-03) 4 commits
478
- - git-remote-mediawiki: no need to update private ref in non-dumb push
479
- - git-remote-mediawiki: use no-private-update capability on dumb push
480
- - transport-helper: add no-private-update capability
481
- - git-remote-mediawiki: add test and check Makefile targets
563
+ When the webserver responds with "405 Method Not Allowed", it
564
+ should tell the client what methods are allowed with the "Allow"
565
+ header.
566
567
Will merge to 'next'.
568
569
486
-* rt/rebase-p-no-merge-summary (2013-08-21) 1 commit
487
- - rebase --preserve-merges: ignore "merge.log" config
570
+* dw/check-ignore-sans-index (2013-09-05) 2 commits
571
+ - SQUASH??? do not let --no-index squat on short-and-sweet -i for now
572
+ - check-ignore: Add option to ignore index contents
573
489
- Originally merged to 'next' on 2013-08-22
574
+ "git check-ignore" follows the same rule as "git add" and "git
575
+ status" in that the ignore/exclude mechanism does not take effect
576
+ on paths that are already tracked. With "--no-index" option, it
577
+ can be used to diagnose which paths that should have been ignored
578
+ have been mistakenly added to the index.
579
491
- "git rebase -p" internally used the merge machinery, but when
492
- rebasing, there should not be a need for merge summary.
580
+ Will squash the tip in and merge to 'next'.
581
494
- Will merge to 'next'.
582
583
+* hu/cherry-pick-previous-branch (2013-09-09) 1 commit
584
+ - cherry-pick: allow "-" as abbreviation of '@{-1}'
585
497
-* sp/clip-read-write-to-8mb (2013-08-20) 2 commits
498
- (merged to 'next' on 2013-08-28 at c9840d7)
499
- + Revert "compat/clipped-write.c: large write(2) fails on Mac OS X/XNU"
500
- + xread, xwrite: limit size of IO to 8MB
586
+ Just like "git checkout -" knows to check out and "git merge -"
587
+ knows to merge the branch you were previously on, teach "git
588
+ cherry-pick" to understand "-" as the previous branch.
589
502
- Originally merged to 'next' on 2013-08-22
590
+ Will merge to 'next'.
591
504
- Send a large request to read(2)/write(2) as a smaller but still
505
- reasonably large chunks, which would improve the latency when the
506
- operation needs to be killed and incidentally works around broken
507
- 64-bit systems that cannot take a 2GB write or read in one go.
592
509
- Will merge to 'master' in the second batch.
593
+* jh/checkout-auto-tracking (2013-09-09) 5 commits
594
+ - branch.c: Relax unnecessary requirement on upstream's remote ref name
595
+ - t3200: Add test demonstrating minor regression in 41c21f2
596
+ - Refer to branch.<name>.remote/merge when documenting --track
597
+ - t3200: Minor fix when preparing for tracking failure
598
+ - t2024: Fix &&-chaining and a couple of typos
599
600
+ Fix a minor regression in v1.8.3.2 and later that made it
601
+ impossible to base your local work on anything but a local branch
602
+ of the upstream repository you are tracking from.
603
512
-* dw/diff-no-index-doc (2013-08-28) 2 commits
513
- - diff --no-index: describe in a separate paragraph
514
- - diff --no-index: clarify operation when not inside a repository
604
+ Will merge to 'next'.
605
516
- When the user types "git diff" outside a working tree, thinking he
517
- is inside one, the current error message that is a single-liner
518
- "usage: git diff --no-index <path> <path>" may not be sufficient to
519
- make him realize the mistake. Add "Not a git repository" to the
520
- error message when we fell into the "--no-index" mode without an
521
- explicit command line option to instruct us to do so.
606
607
+* jk/remove-remote-helpers-in-python (2013-09-09) 1 commit
608
+ - git_remote_helpers: remove little used Python library
609
524
-* sb/repack-in-c (2013-08-30) 2 commits
525
- - repack: retain the return value of pack-objects
526
- - repack: rewrite the shell script in C
610
+ Remove now disused remote-helpers framework for helpers written in
611
+ Python.
612
528
- Any further reviews?
613
+ Will merge to 'next'.
614
615
531
-* ap/commit-author-mailmap (2013-08-24) 1 commit
532
- - commit: search author pattern against mailmap
616
+* jk/upload-pack-keepalive (2013-09-09) 2 commits
617
+ - upload-pack: bump keepalive default to 5 seconds
618
+ - upload-pack: send keepalive packets during pack computation
619
534
- "git commit --author=$name", when $name is not in the canonical
535
- "A. U. Thor <au.thor@example.xz>" format, looks for a matching name
536
- from existing history, but did not consult mailmap to grab the
537
- preferred author name.
620
+ When running "fetch -q", a long silence while the sender side
621
+ computes the set of objects to send can be mistaken by proxies as
622
+ dropped connection.
623
624
Will merge to 'next'.
625
626
542
-* jk/mailmap-incomplete-line (2013-08-28) 1 commit
543
- - mailmap: handle mailmap blobs without trailing newlines
544
-
545
- Will merge to 'next' and later to 'master' and 'maint'.
546
-
547
-
548
-* jk/write-broken-index-with-nul-sha1 (2013-08-28) 1 commit
549
- - write_index: optionally allow broken null sha1s
627
+* mm/rebase-continue-freebsd-WB (2013-09-09) 1 commit
628
+ - rebase: fix run_specific_rebase's use of "return" on FreeBSD
629
551
- Earlier we started rejecting an attempt to add 0{40} object name to
552
- the index and to tree objects, but it sometimes is necessary to
553
- allow so to be able to use tools like filter-branch to correct such
554
- broken tree objects.
630
+ Work around a bug in FreeBSD shell that caused a regression to "git
631
+ rebase" in v1.8.4. It would be lovely to hear from FreeBSD folks a
632
+ success report to make sure we didn't miss any other use of a bare
633
+ "return" from dot-sourced script.
634
635
Will merge to 'next'.
636
637
559
-* kk/tests-with-no-perl (2013-08-24) 4 commits
560
- - reset test: modernize style
561
- - t/t7106-reset-unborn-branch.sh: Add PERL prerequisite
562
- - add -i test: use skip_all instead of repeated PERL prerequisite
563
- - Make test "using invalid commit with -C" more strict
638
+* ss/doclinks (2013-09-06) 1 commit
639
+ - Documentation: make AsciiDoc links always point to HTML files
640
565
- Some tests were not skipped under NO_PERL build.
641
+ When we converted many documents that were traditionally text-only
642
+ to be formatted to AsciiDoc, we did not update links that point at
643
+ them to refer to the formatted HTML files.
644
645
Will merge to 'next'.
646
647
570
-* mm/fast-import-feature-doc (2013-08-25) 1 commit
571
- - Documentation/fast-import: clarify summary for `feature` command
648
+* uh/git-svn-serf-fix (2013-09-06) 1 commit
649
+ - git-svn: fix termination issues for remote svn connections
650
573
- Will merge to 'next'.
651
+ "git-svn" used with SVN 1.8.0 when talking over https:// connection
652
+ dumped core due to a bug in the serf library that SVN uses. Work
653
+ it around on our side, even though the SVN side is being fixed.
654
655
+--------------------------------------------------
656
+[Stalled]
657
576
-* mm/remote-helpers-doc (2013-08-26) 1 commit
577
- - Documentation/remote-helpers: document common use-case for private ref
658
+* rv/send-email-cache-generated-mid (2013-08-21) 2 commits
659
+ - git-send-email: Cache generated message-ids, use them when prompting
660
+ - git-send-email: add optional 'choices' parameter to the ask sub
661
579
- Will merge to 'next'.
662
663
+* rj/read-default-config-in-show-ref-pack-refs (2013-06-17) 3 commits
664
+ - ### DONTMERGE: needs better explanation on what config they need
665
+ - pack-refs.c: Add missing call to git_config()
666
+ - show-ref.c: Add missing call to git_config()
667
582
-* mn/doc-pack-heu-remove-dead-pastebin (2013-08-23) 1 commit
583
- - remove dead pastebin link from pack-heuristics document
668
+ The changes themselves are probably good, but it is unclear what
669
+ basic setting needs to be read for which exact operation.
670
585
- Will merge to 'next'.
671
+ Waiting for clarification.
672
+ $gmane/228294
673
674
588
-* ta/user-manual (2013-08-27) 11 commits
589
- - "git prune" is safe
590
- - Remove irrelevant reference from "Tying it all together"
591
- - Remove unnecessary historical note from "Object storage format"
592
- - Improve section "Merging multiple trees"
593
- - Improve section "Manipulating branches"
594
- - Simplify "How to make a commit"
595
- - Fix some typos and improve wording
596
- - Use "git merge" instead of "git pull ."
597
- - Use current output for "git repack"
598
- - Use current "detached HEAD" message
599
- - Call it "Git User Manual" and remove reference to very old Git version
675
+* jh/shorten-refname (2013-05-07) 4 commits
676
+ - t1514: refname shortening is done after dereferencing symbolic refs
677
+ - shorten_unambiguous_ref(): Fix shortening refs/remotes/origin/HEAD to origin
678
+ - t1514: Demonstrate failure to correctly shorten "refs/remotes/origin/HEAD"
679
+ - t1514: Add tests of shortening refnames in strict/loose mode
680
601
- Update the user's manual to more recent versions of Git.
681
+ When remotes/origin/HEAD is not a symbolic ref, "rev-parse
682
+ --abbrev-ref remotes/origin/HEAD" ought to show "origin", not
683
+ "origin/HEAD", which is fixed with this series (if it is a symbolic
684
+ ref that points at remotes/origin/something, then it should show
685
+ "origin/something" and it already does).
686
687
+ Expecting a reroll, as an early part of a larger series.
688
+ $gmane/225137
689
604
-* tb/precompose-autodetect-fix (2013-08-27) 1 commit
605
- - Set core.precomposeunicode to true on e.g. HFS+
690
607
- Will merge to 'next'.
691
+* mg/more-textconv (2013-05-10) 7 commits
692
+ - grep: honor --textconv for the case rev:path
693
+ - grep: allow to use textconv filters
694
+ - t7008: demonstrate behavior of grep with textconv
695
+ - cat-file: do not die on --textconv without textconv filters
696
+ - show: honor --textconv for blobs
697
+ - diff_opt: track whether flags have been set explicitly
698
+ - t4030: demonstrate behavior of show with textconv
699
700
+ Make "git grep" and "git show" pay attention to --textconv when
701
+ dealing with blob objects.
702
610
-* tf/gitweb-ss-tweak (2013-08-20) 4 commits
611
- - gitweb: make search help link less ugly
612
- - gitweb: omit the repository owner when it is unset
613
- - gitweb: vertically centre contents of page footer
614
- - gitweb: ensure OPML text fits inside its box
703
+ I thought this was pretty well designed and executed, but it seems
704
+ there are some doubts on the list; kicked back to 'pu'.
705
616
- Originally merged to 'next' on 2013-08-22
706
618
- Tweak Gitweb CSS to layout some elements better.
707
+* jc/format-patch (2013-04-22) 2 commits
708
+ - format-patch: --inline-single
709
+ - format-patch: rename "no_inline" field
710
620
- Will merge to 'next'.
711
+ A new option to send a single patch to the standard output to be
712
+ appended at the bottom of a message. I personally have no need for
713
+ this, but it was easy enough to cobble together. Tests, docs and
714
+ stripping out more MIMEy stuff are left as exercises to interested
715
+ parties.
716
717
+ Not ready for inclusion.
718
623
-* es/rebase-i-respect-core-commentchar (2013-08-18) 1 commit
624
- - rebase -i: fix cases ignoring core.commentchar
719
+ Will discard unless we hear from anybody who is interested in
720
+ tying its loose ends.
721
626
- Originally merged to 'next' on 2013-08-20
722
628
- Will merge to 'next'.
723
+* jk/gitweb-utf8 (2013-04-08) 4 commits
724
+ - gitweb: Fix broken blob action parameters on blob/commitdiff pages
725
+ - gitweb: Don't append ';js=(0|1)' to external links
726
+ - gitweb: Make feed title valid utf8
727
+ - gitweb: Fix utf8 encoding for blob_plain, blobdiff_plain, commitdiff_plain, and patch
728
729
+ Various fixes to gitweb.
730
631
-* jx/branch-vv-always-compare-with-upstream (2013-08-26) 2 commits
632
- - status: always show tracking branch even no change
633
- - branch: report invalid tracking branch as gone
731
+ Drew Northup volunteered to take a look into this.
732
+ $gmane/226216
733
635
- "git branch -v -v" (and "git status") did not distinguish among a
636
- branch that does not build on any other branch, a branch that is in
637
- sync with the branch it builds on, and a branch that is configured
638
- to build on some other branch that no longer exists.
734
640
- Am I waiting for another reroll?
735
+* jc/show-branch (2013-06-07) 5 commits
736
+ - show-branch: use commit slab to represent bitflags of arbitrary width
737
+ - show-branch.c: remove "all_mask"
738
+ - show-branch.c: abstract out "flags" operation
739
+ - show-branch.c: lift all_mask/all_revs to a global static
740
+ - show-branch.c: update comment style
741
742
+ Waiting for the final step to lift the hard-limit before sending it out.
743
643
-* nd/fetch-into-shallow (2013-08-28) 7 commits
644
- - Add testcase for needless objects during a shallow fetch
645
- - list-objects: mark more commits as edges in mark_edges_uninteresting
646
- - list-objects: reduce one argument in mark_edges_uninteresting
647
- - upload-pack: delegate rev walking in shallow fetch to pack-objects
648
- - shallow: add setup_temporary_shallow()
649
- - shallow: only add shallow graft points to new shallow file
650
- - move setup_alternate_shallow and write_shallow_commits to shallow.c
744
+--------------------------------------------------
745
+[Cooking]
746
652
- When there is no sufficient overlap between old and new history
653
- during a fetch into a shallow repository, we unnecessarily sent
654
- objects the sending side knows the receiving end has.
747
+* fc/contrib-bzr-hg-fixes (2013-09-03) 10 commits
748
+ - contrib/remote-helpers: quote variable references in redirection targets
749
+ - contrib/remote-helpers: style updates for test scripts
750
+ - remote-hg: use notes to keep track of Hg revisions
751
+ - remote-helpers: cleanup more global variables
752
+ - remote-helpers: trivial style fixes
753
+ - remote-hg: improve basic test
754
+ - remote-hg: add missing &&s in the test
755
+ - remote-hg: fix test
756
+ - remote-bzr: make bzr branches configurable per-repo
757
+ - remote-bzr: fix export of utf-8 authors
758
759
Will merge to 'next'.
760
761
659
-* sb/diff-delta-remove-needless-comparison (2013-08-18) 1 commit
660
- (merged to 'next' on 2013-08-28 at ba09c3f)
661
- + create_delta_index: simplify condition always evaluating to true
762
+* jc/pager-configuration-doc (2013-08-29) 1 commit
763
+ (merged to 'next' on 2013-09-05 at 3169083)
764
+ + config: rewrite core.pager documentation
765
663
- Originally merged to 'next' on 2013-08-20
766
+ It was unclear in the documentation how various configurations and
767
+ environment variables determine which pager is eventually used.
768
769
Will merge to 'master'.
770
771
668
-* sg/bash-prompt-lf-in-cwd-test (2013-08-18) 1 commit
669
- (merged to 'next' on 2013-08-28 at eb449b5)
670
- + bash prompt: test the prompt with newline in repository path
772
+* np/pack-v4 (2013-09-03) 24 commits
773
+ - Document pack v4 format
774
+ - initial pack index v3 support on the read side
775
+ - pack v4: add progress display
776
+ - pack v4: normalize pack name to properly generate the pack index file name
777
+ - pack index v3
778
+ - pack v4: relax commit parsing a bit
779
+ - pack v4: honor pack.compression config option
780
+ - pack v4: load delta candidate for encoding tree objects
781
+ - pack v4: tree object delta encoding
782
+ - pack v4: object writing
783
+ - pack v4: object data copy
784
+ - pack v4: object headers
785
+ - pack v4: creation code
786
+ - pack v4: dictionary table output
787
+ - pack v4: tree object encoding
788
+ - pack v4: commit object encoding
789
+ - pack v4: basic references encoding
790
+ - pack v4: move to struct pack_idx_entry and get rid of our own struct idx_entry
791
+ - pack v4: split the object list and dictionary creation
792
+ - pack v4: add commit object parsing
793
+ - pack v4: add tree entry mode support to dictionary entries
794
+ - pack v4: scan tree objects
795
+ - export packed_object_info()
796
+ - pack v4: initial pack dictionary structure and code
797
672
- Originally merged to 'next' on 2013-08-20
798
+ Will replace with a newer version of the series.
799
674
- Will merge to 'master'.
800
801
+* jk/free-tree-buffer (2013-06-06) 1 commit
802
+ (merged to 'next' on 2013-09-09 at 3576189)
803
+ + clear parsed flag when we free tree buffers
804
677
-* jl/some-submodule-config-are-not-boolean (2013-08-19) 1 commit
678
- (merged to 'next' on 2013-08-28 at 7fabc3e)
679
- + avoid segfault on submodule.*.path set to an empty "true"
805
+ Will merge to 'master'.
806
681
- Originally merged to 'next' on 2013-08-20
807
683
- Will merge to 'master'.
808
+* jk/has-sha1-file-retry-packed (2013-08-30) 1 commit
809
+ (merged to 'next' on 2013-09-09 at fc42e9b)
810
+ + has_sha1_file: re-check pack directory before giving up
811
812
+ When an object is not found after checking the packfiles and then
813
+ loose object directory, read_sha1_file() re-checks the packfiles to
814
+ prevent racing with a concurrent repacker; teach the same logic to
815
+ has_sha1_file().
816
686
-* jc/ls-files-killed-optim (2013-08-23) 4 commits
687
- - dir.c::test_one_path(): work around directory_exists_in_index_icase() breakage
688
- - t3010: update to demonstrate "ls-files -k" optimization pitfalls
689
- - ls-files -k: a directory only can be killed if the index has a non-directory
690
- - dir.c: use the cache_* macro to access the current index
817
+ Will merge to 'master'.
818
692
- Originally merged to 'next' on 2013-08-27
819
694
- "git ls-files -k" needs to crawl only the part of the working tree
695
- that may overlap the paths in the index to find killed files, but
696
- shared code with the logic to find all the untracked files, which
697
- made it unnecessarily inefficient.
820
+* ab/gitweb-author-initials (2013-08-30) 1 commit
821
+ (merged to 'next' on 2013-09-09 at ecb924d)
822
+ + gitweb: Fix the author initials in blame for non-ASCII names
823
699
- Will merge to 'next'.
824
+ Will merge to 'master'.
825
826
702
-* es/blame-L-twice (2013-08-06) 16 commits
703
- (merged to 'next' on 2013-08-28 at e83ae2e)
704
- + line-range: reject -L line numbers less than 1
705
- + t8001/t8002: blame: add tests of -L line numbers less than 1
706
- + line-range: teach -L^:RE to search from start of file
707
- + line-range: teach -L:RE to search from end of previous -L range
708
- + line-range: teach -L^/RE/ to search from start of file
709
- + line-range-format.txt: document -L/RE/ relative search
710
- + log: teach -L/RE/ to search from end of previous -L range
711
- + blame: teach -L/RE/ to search from end of previous -L range
712
- + line-range: teach -L/RE/ to search relative to anchor point
713
- + blame: document multiple -L support
714
- + t8001/t8002: blame: add tests of multiple -L options
715
- + blame: accept multiple -L ranges
716
- + blame: inline one-line function into its lone caller
717
- + range-set: publish API for re-use by git-blame -L
718
- + line-range-format.txt: clarify -L:regex usage form
719
- + git-log.txt: place each -L option variation on its own line
720
- (this branch is used by es/contacts-blame-L-multi; uses es/blame-L-more.)
827
+* bk/refs-multi-update (2013-09-09) 8 commits
828
+ - update-ref: add test cases covering --stdin signature
829
+ - update-ref: support multiple simultaneous updates
830
+ - refs: add update_refs for multiple simultaneous updates
831
+ - refs: add function to repack without multiple refs
832
+ - refs: factor delete_ref loose ref step into a helper
833
+ - refs: factor update_ref steps into helpers
834
+ - refs: report ref type from lock_any_ref_for_update
835
+ - reset: rename update_refs to reset_refs
836
722
- Originally merged to 'next' on 2013-08-08
837
+ Give "update-refs" a "--stdin" option to read multiple update
838
+ requests and perform them in an all-or-none fashion.
839
724
- Teaches "git blame" to take more than one -L ranges.
840
+ Comments?
841
726
- Will merge to 'master' in the second batch.
842
843
+* fc/at-head (2013-09-09) 3 commits
844
+ - SQUASH??? fixup
845
+ - Add new @ shortcut for HEAD
846
+ - sha1-name: pass len argument to interpret_branch_name()
847
729
-* mm/no-shell-escape-in-die-message (2013-08-07) 1 commit
730
- (merged to 'next' on 2013-08-28 at 86b300f)
731
- + die_with_status: use "printf '%s\n'", not "echo"
848
+ Attempt to resurrect "Type @ for HEAD"; the bottom one seems to be
849
+ a genuine code improvement, but identifying cases where "@" means
850
+ HEAD were harder than it should have been. I think the result of
851
+ squashing the tip one in covers all the necessary cases.
852
733
- Originally merged to 'next' on 2013-08-08
853
+ Will squash the tip in and merge to 'next'.
854
735
- Fixes a minor bug in "git rebase -i" (there could be others, as the
736
- root cause is pretty generic) where the code feeds a random, data
737
- dependeant string to 'echo' and expects it to come out literally.
855
739
- Will merge to 'master'.
856
+* fc/fast-export (2013-09-03) 2 commits
857
+ (merged to 'next' on 2013-09-09 at 8d5d396)
858
+ + fast-export: refactor get_tags_and_duplicates()
859
+ + fast-export: make extra_refs global
860
861
+ Code simpification.
862
742
-* tr/fd-gotcha-fixes (2013-08-06) 1 commit
743
- (merged to 'next' on 2013-08-28 at 5305271)
744
- + t0070: test that git_mkstemps correctly checks return value of open()
863
+ Will merge to 'master'.
864
746
- Originally merged to 'next' on 2013-08-08
865
748
- Finishing touches to an earlier fix already in 'master'.
866
+* fc/rev-parse-test-updates (2013-09-03) 4 commits
867
+ (merged to 'next' on 2013-09-09 at 92c51ef)
868
+ + rev-parse test: use standard test functions for setup
869
+ + rev-parse test: use test_cmp instead of "test" builtin
870
+ + rev-parse test: use test_must_fail, not "if <command>; then false; fi"
871
+ + rev-parse test: modernize quoting and whitespace
872
+
873
+ Modernize tests.
874
875
Will merge to 'master'.
876
877
753
-* ap/remote-hg-tilde-is-home-directory (2013-08-09) 1 commit
754
- (merged to 'next' on 2013-08-28 at 9bfc430)
755
- + remote-hg: fix path when cloning with tilde expansion
878
+* fc/t3200-fixes (2013-09-03) 3 commits
879
+ (merged to 'next' on 2013-09-09 at 3626363)
880
+ + t: branch: fix broken && chains
881
+ + t: branch: fix typo
882
+ + t: branch: trivial style fix
883
757
- Originally merged to 'next' on 2013-08-14
884
+ Will merge to 'master'.
885
+
886
+
887
+* fc/trivial (2013-09-08) 5 commits
888
+ (merged to 'next' on 2013-09-09 at a8ad2e1)
889
+ + pull: use $curr_branch_short more
890
+ + add: trivial style cleanup
891
+ + reset: trivial style cleanup
892
+ + branch: trivial style fix
893
+ + reset: trivial refactoring
894
895
Will merge to 'master'.
896
897
762
-* es/rebase-i-no-abbrev (2013-08-25) 3 commits
763
- - rebase -i: fix short SHA-1 collision
764
- - t3404: rebase -i: demonstrate short SHA-1 collision
765
- - t3404: make tests more self-contained
898
+* jc/ref-excludes (2013-09-03) 2 commits
899
+ - document --exclude option
900
+ - revision: introduce --exclude=<glob> to tame wildcards
901
767
- Originally merged to 'next' on 2013-08-26
902
+ People often wished a way to tell "git log --branches" (and "git
903
+ log --remotes --not --branches") to exclude some local branches
904
+ from the expansion of "--branches" (similarly for "--tags", "--all"
905
+ and "--glob=<pattern>"). Now they have one.
906
907
Will merge to 'next'.
908
909
772
-* fc/remote-hg-shared-setup (2013-08-11) 2 commits
773
- (merged to 'next' on 2013-08-28 at 41ee3fe)
774
- + remote-hg: add shared repo upgrade
775
- + remote-hg: ensure shared repo is initialized
910
+* jk/pager-bypass-cat-for-default-pager (2013-09-03) 1 commit
911
+ (merged to 'next' on 2013-09-09 at c9cfbaa)
912
+ + pager: turn on "cat" optimization for DEFAULT_PAGER
913
777
- Originally merged to 'next' on 2013-08-14
914
+ If a build-time fallback is set to "cat" instead of "less", we
915
+ should apply the same "no subprocess or pipe" optimization as we
916
+ apply to user-supplied GIT_PAGER=cat.
917
918
Will merge to 'master'.
919
920
782
-* jc/transport-do-not-use-connect-twice-in-fetch (2013-08-28) 6 commits
783
- (merged to 'next' on 2013-08-29 at f88a6da)
784
- + builtin/fetch.c: Fix a sparse warning
785
- (merged to 'next' on 2013-08-28 at 2ed86df)
786
- + fetch: work around "transport-take-over" hack
787
- + fetch: refactor code that fetches leftover tags
788
- + fetch: refactor code that prepares a transport
789
- + fetch: rename file-scope global "transport" to "gtransport"
790
- + t5802: add test for connect helper
791
-
792
- Originally merged to 'next' on 2013-08-20
921
+* nd/git-dir-pointing-at-gitfile (2013-09-03) 1 commit
922
+ (merged to 'next' on 2013-09-09 at 5b0a4bf)
923
+ + Make setup_git_env() resolve .git file when $GIT_DIR is not specified
924
794
- The auto-tag-following code in "git fetch" tries to reuse the same
795
- transport twice when the serving end does not cooperate and does
796
- not give tags that point to commits that are asked for as part of
797
- the primary transfer. Unfortunately, Git-aware transport helper
798
- interface is not designed to be used more than once, hence this
799
- does not work over smart-http transfer.
925
+ We made sure that we notice the user-supplied GIT_DIR is actually a
926
+ gitfile, but failed to do so when the default ".git" is a gitfile.
927
801
- Will merge to 'master' in the second batch.
928
+ Will merge to 'master'.
929
930
804
-* ks/p4-view-spec (2013-09-03) 2 commits
805
- - git p4: implement view spec wildcards with "p4 where"
806
- - git p4 test: sanitize P4CHARSET
931
+* nr/git-cd-to-a-directory (2013-09-09) 1 commit
932
+ - git: run in a directory given with -C option
933
808
- Replaced with a reroll that was whitespace damaged.
934
+ Just like "make -C <directory>", make "git -C <directory> ..." to
935
+ go there before doing anything else.
936
810
- Waiting for the final Ack.
937
+ Will merge to 'next'.
938
939
813
-* nd/gc-lock-against-each-other (2013-08-09) 1 commit
814
- (merged to 'next' on 2013-08-28 at 46e072b)
815
- + gc: reject if another gc is running, unless --force is given
940
+* rh/peeling-tag-to-tag (2013-09-03) 2 commits
941
+ - peel_onion: do not assume length of x_type globals
942
+ - peel_onion(): add support for <rev>^{tag}
943
817
- Originally merged to 'next' on 2013-08-14
944
+ Make "foo^{tag}" to peel a tag to itself, i.e. no-op., and fail if
945
+ "foo" is not a tag.
946
819
- Will merge to 'master'.
947
+ Will merge to 'next'.
948
949
822
-* sb/misc-cleanup (2013-08-09) 3 commits
823
- (merged to 'next' on 2013-08-28 at bc2045b)
824
- + rm: remove unneeded null pointer check
825
- + diff: fix a possible null pointer dereference
826
- + diff: remove ternary operator evaluating always to true
950
+* cc/replace-with-the-same-type (2013-09-09) 8 commits
951
+ - Doc: 'replace' merge and non-merge commits
952
+ - t6050-replace: use some long option names
953
+ - replace: allow long option names
954
+ - Documentation/replace: add Creating Replacement Objects section
955
+ - t6050-replace: add test to clean up all the replace refs
956
+ - t6050-replace: test that objects are of the same type
957
+ - Documentation/replace: state that objects must be of the same type
958
+ - replace: forbid replacing an object with one of a different type
959
828
- Originally merged to 'next' on 2013-08-14
960
+ Using the replace mechanism to swap an object with another object
961
+ of a different type can introduce inconsistency (e.g. a tree
962
+ expects an object name to refer to a blob, but the blob object can
963
+ be mistakenly or maliciously replaced with an object with a
964
+ different type). Attempt to forbid such.
965
830
- Will merge to 'master'.
966
+ Will merge to 'next'.
967
968
833
-* aj/p4-symlink-lose-nl (2013-08-12) 1 commit
834
- (merged to 'next' on 2013-08-28 at 5ce9ceb)
835
- + git-p4: Fix occasional truncation of symlink contents.
969
+* jx/clean-interactive (2013-08-28) 1 commit
970
+ (merged to 'next' on 2013-09-09 at 477fec6)
971
+ + documentation: clarify notes for clean.requireForce
972
837
- Originally merged to 'next' on 2013-08-14
973
+ Finishing touches to update the document to adjust to a new option
974
+ "git clean" learned recently.
975
976
Will merge to 'master'.
977
978
842
-* es/contacts-blame-L-multi (2013-08-13) 3 commits
843
- (merged to 'next' on 2013-08-28 at 24ff8b4)
844
- + contacts: reduce git-blame invocations
845
- + contacts: gather all blame sources prior to invoking git-blame
846
- + contacts: validate hunk length earlier
847
- (this branch uses es/blame-L-more and es/blame-L-twice.)
979
+* mm/status-without-comment-char (2013-09-06) 6 commits
980
+ - status: add missing blank line after list of "other" files
981
+ - tests: don't set status.displayCommentPrefix file-wide
982
+ - status: disable display of '#' comment prefix by default
983
+ - submodule summary: ignore --for-status option
984
+ - wt-status: use argv_array API
985
+ - builtin/stripspace.c: fix broken indentation
986
849
- Originally merged to 'next' on 2013-08-14
987
+ Allow "git status" to omit the prefix to make its output a comment
988
+ in a commit log editor, which is not necessary for human
989
+ consumption.
990
851
- Will merge to 'master' in the second batch.
991
+ We may want to tighten the output to omit unnecessary trailing
992
+ blank lines, but that does not have to be in the scope of this
993
+ series.
994
995
+ Will merge to 'next'.
996
854
-* fc/unpack-trees-leakfix (2013-08-13) 1 commit
855
- (merged to 'next' on 2013-08-28 at 3ad1f21)
856
- + unpack-trees: plug a memory leak
997
858
- Originally merged to 'next' on 2013-08-14
998
+* js/xread-in-full (2013-08-20) 1 commit
999
+ (merged to 'next' on 2013-09-04 at 5bfb049)
1000
+ + stream_to_pack: xread does not guarantee to read all requested bytes
1001
860
- Will merge to 'master'.
1002
+ Originally merged to 'next' on 2013-08-20
1003
+
1004
+ A call to xread() was used without a loop around to cope with short
1005
+ read in the codepath to stream new contents to a pack.
1006
1007
+ Will merge to 'master' in the third batch.
1008
863
-* nd/push-no-thin (2013-08-13) 1 commit
864
- - push: respect --no-thin
1009
866
- Originally merged to 'next' on 2013-08-14
1010
+* sb/mailmap-freeing-NULL-is-ok (2013-08-20) 1 commit
1011
+ (merged to 'next' on 2013-09-04 at c831015)
1012
+ + mailmap: remove redundant check for freeing memory
1013
868
- Will merge to 'next'.
1014
+ Originally merged to 'next' on 2013-08-20
1015
1016
+ Will merge to 'master' in the third batch.
1017
871
-* sh/pull-rebase-preserve (2013-08-13) 1 commit
872
- - pull: Allow pull to preserve merges when rebasing.
1018
874
- Originally merged to 'next' on 2013-08-14
1019
+* bc/submodule-status-ignored (2013-09-04) 2 commits
1020
+ - submodule: don't print status output with ignore=all
1021
+ - submodule: fix confusing variable name
1022
+
1023
+ Originally merged to 'next' on 2013-08-22
1024
1025
Will merge to 'next'.
1026
1027
879
-* es/blame-L-more (2013-08-05) 11 commits
880
- (merged to 'next' on 2013-08-28 at b73cdc9)
881
- + blame: reject empty ranges -L,+0 and -L,-0
882
- + t8001/t8002: blame: demonstrate acceptance of bogus -L,+0 and -L,-0
883
- + blame: reject empty ranges -LX,+0 and -LX,-0
884
- + t8001/t8002: blame: demonstrate acceptance of bogus -LX,+0 and -LX,-0
885
- + log: fix -L bounds checking bug
886
- + t4211: retire soon-to-be unimplementable tests
887
- + t4211: log: demonstrate -L bounds checking bug
888
- + blame: fix -L bounds checking bug
889
- + t8001/t8002: blame: add empty file & partial-line tests
890
- + t8001/t8002: blame: demonstrate -L bounds checking bug
891
- + t8001/t8002: blame: decompose overly-large test
892
- (this branch is used by es/blame-L-twice and es/contacts-blame-L-multi.)
1028
+* jk/config-int-range-check (2013-09-09) 5 commits
1029
+ (merged to 'next' on 2013-09-09 at 9ab779d)
1030
+ + git-config: always treat --int as 64-bit internally
1031
+ + config: make numeric parsing errors more clear
1032
+ + config: set errno in numeric git_parse_* functions
1033
+ + config: properly range-check integer values
1034
+ + config: factor out integer parsing from range checks
1035
894
- Originally merged to 'next' on 2013-08-06
1036
+ Originally merged to 'next' on 2013-08-22
1037
896
- More fixes to the code to parse the "-L" option in "log" and "blame".
1038
+ "git config --int section.var 3g" should somehow diagnose that the
1039
+ number does not fit in "int" (on 32-bit platforms anyway) but it
1040
+ did not.
1041
898
- Will merge to 'master' in the second batch.
1042
+ Will merge to 'master'.
1043
1044
901
-* jk/cat-file-batch-optim (2013-08-05) 1 commit
902
- (merged to 'next' on 2013-08-28 at 624e60d)
903
- + cat-file: only split on whitespace when %(rest) is used
1045
+* jk/duplicate-objects-in-packs (2013-09-04) 5 commits
1046
+ (merged to 'next' on 2013-09-09 at 72f2c3d)
1047
+ + t5308: check that index-pack --strict detects duplicate objects
1048
+ + test index-pack on packs with recoverable delta cycles
1049
+ + add tests for indexing packs with delta cycles
1050
+ + sha1-lookup: handle duplicate keys with GIT_USE_LOOKUP
1051
+ + test-sha1: add a binary output mode
1052
905
- Originally merged to 'next' on 2013-08-06
1053
+ A packfile that stores the same object more than once is broken and
1054
+ will be rejected.
1055
907
- Rework the reverted change to `cat-file --batch-check`.
1056
+ Will merge to 'master'.
1057
909
- Will merge to 'master' in the second batch.
1058
1059
+* mm/mediawiki-dumb-push-fix (2013-09-03) 4 commits
1060
+ (merged to 'next' on 2013-09-05 at f8313f4)
1061
+ + git-remote-mediawiki: no need to update private ref in non-dumb push
1062
+ + git-remote-mediawiki: use no-private-update capability on dumb push
1063
+ + transport-helper: add no-private-update capability
1064
+ + git-remote-mediawiki: add test and check Makefile targets
1065
912
-* jn/post-receive-utf8 (2013-08-05) 3 commits
913
- - hooks/post-receive-email: set declared encoding to utf-8
914
- - hooks/post-receive-email: force log messages in UTF-8
915
- - hooks/post-receive-email: use plumbing instead of git log/show
1066
+ Will merge to 'master'.
1067
917
- Originally merged to 'next' on 2013-08-20
1068
919
- Update post-receive-email script to make sure the message contents
920
- and pathnames are encoded consistently in UTF-8.
1069
+* rt/rebase-p-no-merge-summary (2013-08-21) 1 commit
1070
+ (merged to 'next' on 2013-09-04 at d8d89ee)
1071
+ + rebase --preserve-merges: ignore "merge.log" config
1072
922
- I have a feeling that it is a lost cause to solve the issue the
923
- topic tries to address in general, because the patch text can have
924
- payload in any encodings that are different from either the
925
- pathnames or the log message. Patches that touch paths that use an
926
- encoding that conflicts with the encoding of the payload and/or the
927
- log message could be transferred with core.quotepath set and patch
928
- generated as all binary, but that would be pretty much useless.
1073
+ Originally merged to 'next' on 2013-08-22
1074
930
- Will merge to 'next'.
1075
+ "git rebase -p" internally used the merge machinery, but when
1076
+ rebasing, there should not be a need for merge summary.
1077
1078
+ Will merge to 'master' in the third batch.
1079
933
-* sb/parseopt-boolean-removal (2013-08-07) 9 commits
934
- (merged to 'next' on 2013-08-28 at 2f3f21f)
935
- + revert: use the OPT_CMDMODE for parsing, reducing code
936
- + checkout-index: fix negations of even numbers of -n
937
- + config parsing options: allow one flag multiple times
938
- + hash-object: replace stdin parsing OPT_BOOLEAN by OPT_COUNTUP
939
- + branch, commit, name-rev: ease up boolean conditions
940
- + checkout: remove superfluous local variable
941
- + log, format-patch: parsing uses OPT__QUIET
942
- + Replace deprecated OPT_BOOLEAN by OPT_BOOL
943
- + Remove deprecated OPTION_BOOLEAN for parsing arguments
944
- (this branch uses jc/parseopt-command-modes.)
1080
946
- Convert most uses of OPT_BOOLEAN/OPTION_BOOLEAN that can use
947
- OPT_BOOL/OPTION_BOOLEAN which have much saner semantics, and turn
948
- remaining ones into OPT_SET_INT, OPT_COUNTUP, etc. as necessary.
1081
+* dw/diff-no-index-doc (2013-08-28) 2 commits
1082
+ - diff --no-index: describe in a separate paragraph
1083
+ - diff --no-index: clarify operation when not inside a repository
1084
950
- Will merge to 'master'.
1085
+ When the user types "git diff" outside a working tree, thinking he
1086
+ is inside one, the current error message that is a single-liner
1087
+ "usage: git diff --no-index <path> <path>" may not be sufficient to
1088
+ make him realize the mistake. Add "Not a git repository" to the
1089
+ error message when we fell into the "--no-index" mode without an
1090
+ explicit command line option to instruct us to do so.
1091
1092
+ Will merge to 'next'.
1093
953
-* jc/parseopt-command-modes (2013-07-30) 2 commits
954
- (merged to 'next' on 2013-08-28 at 738b399)
955
- + tag: use OPT_CMDMODE
956
- + parse-options: add OPT_CMDMODE()
957
- (this branch is used by sb/parseopt-boolean-removal.)
1094
959
- Many commands use --dashed-option as a operation mode selector
960
- (e.g. "git tag --delete") that the user can use at most one
961
- (e.g. "git tag --delete --verify" is a nonsense) and you cannot
962
- negate (e.g. "git tag --no-delete" is a nonsense). Make it easier
963
- for users of parse_options() to enforce these restrictions.
1095
+* sb/repack-in-c (2013-08-30) 2 commits
1096
+ - repack: retain the return value of pack-objects
1097
+ - repack: rewrite the shell script in C
1098
965
- Will merge to 'master'.
1099
+ Any further reviews?
1100
1101
968
-* tr/log-full-diff-keep-true-parents (2013-08-05) 2 commits
969
- (merged to 'next' on 2013-08-28 at e86ec11)
970
- + log: use true parents for diff when walking reflogs
971
- + log: use true parents for diff even when rewriting
1102
+* ap/commit-author-mailmap (2013-08-24) 1 commit
1103
+ (merged to 'next' on 2013-09-09 at 79d5a20)
1104
+ + commit: search author pattern against mailmap
1105
973
- Originally merged to 'next' on 2013-08-08
1106
+ "git commit --author=$name", when $name is not in the canonical
1107
+ "A. U. Thor <au.thor@example.xz>" format, looks for a matching name
1108
+ from existing history, but did not consult mailmap to grab the
1109
+ preferred author name.
1110
975
- Output from "git log --full-diff -- <pathspec>" looked strange,
976
- because comparison was done with the previous ancestor that touched
977
- the specified <pathspec>, causing the patches for paths outside the
978
- pathspec to show more than the single commit has changed.
1111
+ Will merge to 'master'.
1112
980
- Tweak "git reflog -p" for the same reason using the same mechanism.
1113
982
- Will merge to 'master' in the second batch.
1114
+* jk/write-broken-index-with-nul-sha1 (2013-08-28) 1 commit
1115
+ (merged to 'next' on 2013-09-09 at 6953f27)
1116
+ + write_index: optionally allow broken null sha1s
1117
1118
+ Earlier we started rejecting an attempt to add 0{40} object name to
1119
+ the index and to tree objects, but it sometimes is necessary to
1120
+ allow so to be able to use tools like filter-branch to correct such
1121
+ broken tree objects.
1122
985
-* bc/unuse-packfile (2013-08-02) 2 commits
986
- (merged to 'next' on 2013-08-28 at 9029801)
987
- + Don't close pack fd when free'ing pack windows
988
- + sha1_file: introduce close_one_pack() to close packs on fd pressure
1123
+ Will merge to 'master'.
1124
990
- Originally merged to 'next' on 2013-08-06
1125
992
- Handle memory pressure and file descriptor pressure separately when
993
- deciding to release pack windows to honor resource limits.
1126
+* kk/tests-with-no-perl (2013-08-24) 4 commits
1127
+ (merged to 'next' on 2013-09-09 at 67510b1)
1128
+ + reset test: modernize style
1129
+ + t/t7106-reset-unborn-branch.sh: Add PERL prerequisite
1130
+ + add -i test: use skip_all instead of repeated PERL prerequisite
1131
+ + Make test "using invalid commit with -C" more strict
1132
995
- Will merge to 'master'.
1133
+ Some tests were not skipped under NO_PERL build.
1134
1135
+ Will merge to 'master'.
1136
998
-* da/darwin (2013-08-05) 3 commits
999
- (merged to 'next' on 2013-08-28 at 6e4c5ae)
1000
- + OS X: Fix redeclaration of die warning
1001
- + Makefile: Fix APPLE_COMMON_CRYPTO with BLK_SHA1
1002
- + imap-send: use Apple's Security framework for base64 encoding
1137
1004
- Originally merged to 'next' on 2013-08-05
1138
+* mm/fast-import-feature-doc (2013-08-25) 1 commit
1139
+ (merged to 'next' on 2013-09-05 at 83802e2)
1140
+ + Documentation/fast-import: clarify summary for `feature` command
1141
1142
Will merge to 'master'.
1143
1144
1009
-* nd/sq-quote-buf (2013-07-30) 3 commits
1010
- (merged to 'next' on 2013-08-28 at d0ffd55)
1011
- + quote: remove sq_quote_print()
1012
- + tar-tree: remove dependency on sq_quote_print()
1013
- + for-each-ref, quote: convert *_quote_print -> *_quote_buf
1145
+* mm/remote-helpers-doc (2013-08-26) 1 commit
1146
+ (merged to 'next' on 2013-09-05 at c181b35)
1147
+ + Documentation/remote-helpers: document common use-case for private ref
1148
1015
- Originally merged to 'next' on 2013-08-01
1149
+ Will merge to 'master'.
1150
1017
- Code simplification as a preparatory step to something larger.
1151
1019
- Will merge to 'master'.
1152
+* mn/doc-pack-heu-remove-dead-pastebin (2013-08-23) 1 commit
1153
+ (merged to 'next' on 2013-09-05 at 5caecec)
1154
+ + remove dead pastebin link from pack-heuristics document
1155
1156
+ Will merge to 'master'.
1157
1022
-* jc/url-match (2013-08-09) 7 commits
1023
- (merged to 'next' on 2013-08-28 at 2a9fd23)
1024
- + builtin/config.c: compilation fix
1025
- + config: "git config --get-urlmatch" parses section.<url>.key
1026
- + builtin/config: refactor collect_config()
1027
- + config: parse http.<url>.<variable> using urlmatch
1028
- + config: add generic callback wrapper to parse section.<url>.key
1029
- + config: add helper to normalize and match URLs
1030
- + http.c: fix parsing of http.sslCertPasswordProtected variable
1158
1032
- Originally merged to 'next' on 2013-08-12
1159
+* ta/user-manual (2013-08-27) 11 commits
1160
+ - "git prune" is safe
1161
+ - Remove irrelevant reference from "Tying it all together"
1162
+ - Remove unnecessary historical note from "Object storage format"
1163
+ - Improve section "Merging multiple trees"
1164
+ - Improve section "Manipulating branches"
1165
+ - Simplify "How to make a commit"
1166
+ - Fix some typos and improve wording
1167
+ - Use "git merge" instead of "git pull ."
1168
+ - Use current output for "git repack"
1169
+ - Use current "detached HEAD" message
1170
+ - Call it "Git User Manual" and remove reference to very old Git version
1171
1034
- Allow section.<urlpattern>.var configuration variables to be
1035
- treated as a "virtual" section.var given a URL, and use the
1036
- mechanism to enhance http.* configuration variables.
1172
+ Update the user's manual to more recent versions of Git.
1173
1038
- This is a reroll of Kyle J. McKay's work.
1174
+ Will merge to 'next'.
1175
1040
- Will merge to 'master' in the second batch.
1176
1177
+* tb/precompose-autodetect-fix (2013-08-27) 1 commit
1178
+ (merged to 'next' on 2013-09-09 at 9bfdac1)
1179
+ + Set core.precomposeunicode to true on e.g. HFS+
1180
1043
-* jl/submodule-mv (2013-08-06) 5 commits
1044
- (merged to 'next' on 2013-08-28 at 4ec6b41)
1045
- + rm: delete .gitmodules entry of submodules removed from the work tree
1046
- + mv: update the path entry in .gitmodules for moved submodules
1047
- + submodule.c: add .gitmodules staging helper functions
1048
- + mv: move submodules using a gitfile
1049
- + mv: move submodules together with their work trees
1050
- (this branch uses nd/magic-pathspec.)
1181
+ On MacOS X, we detected if the filesystem needs the "pre-composed
1182
+ unicode strings" workaround, but did not automatically enable it.
1183
+ Now we do.
1184
1052
- Originally merged to 'next' on 2013-08-08
1185
+ Will merge to 'master'.
1186
1054
- "git mv A B" when moving a submodule A does "the right thing",
1055
- inclusing relocating its working tree and adjusting the paths in
1056
- the .gitmodules file.
1187
1058
- Will merge to 'master' in the second batch.
1188
+* tf/gitweb-ss-tweak (2013-08-20) 4 commits
1189
+ (merged to 'next' on 2013-09-04 at 774bfbe)
1190
+ + gitweb: make search help link less ugly
1191
+ + gitweb: omit the repository owner when it is unset
1192
+ + gitweb: vertically centre contents of page footer
1193
+ + gitweb: ensure OPML text fits inside its box
1194
1195
+ Originally merged to 'next' on 2013-08-22
1196
1061
-* nd/clone-connectivity-shortcut (2013-07-23) 1 commit
1062
- (merged to 'next' on 2013-08-28 at d227f9f)
1063
- + smart http: use the same connectivity check on cloning
1197
+ Tweak Gitweb CSS to layout some elements better.
1198
1065
- Originally merged to 'next' on 2013-07-30
1199
+ Will merge to 'master' in the third batch.
1200
1067
- Will merge to 'master' in the second batch.
1201
1202
+* es/rebase-i-respect-core-commentchar (2013-08-18) 1 commit
1203
+ (merged to 'next' on 2013-09-04 at 8c1ce68)
1204
+ + rebase -i: fix cases ignoring core.commentchar
1205
1070
-* db/http-savecookies (2013-08-05) 2 commits
1071
- (merged to 'next' on 2013-08-28 at 22a8464)
1072
- + t5551: Remove header from curl cookie file
1073
- + http: add http.savecookies option to write out HTTP cookies
1206
+ Originally merged to 'next' on 2013-08-20
1207
1075
- Originally merged to 'next' on 2013-08-05
1208
+ "rebase -i" forgot that the comment character can be configurable
1209
+ while reading its insn sheet.
1210
1077
- Will merge to 'master' in the second batch.
1211
+ Will merge to 'master' in the third batch.
1212
1213
1080
-* rr/feed-real-path-to-editor (2013-07-29) 1 commit
1081
- (merged to 'next' on 2013-08-28 at 7458fa4)
1082
- + editor: use canonicalized absolute path
1214
+* jx/branch-vv-always-compare-with-upstream (2013-08-26) 2 commits
1215
+ - status: always show tracking branch even no change
1216
+ - branch: report invalid tracking branch as gone
1217
1084
- Originally merged to 'next' on 2013-07-30
1218
+ "git branch -v -v" (and "git status") did not distinguish among a
1219
+ branch that does not build on any other branch, a branch that is in
1220
+ sync with the branch it builds on, and a branch that is configured
1221
+ to build on some other branch that no longer exists.
1222
1086
- Will merge to 'master'.
1223
+ Will merge to 'next'.
1224
1225
1089
-* jk/fast-import-empty-ls (2013-06-23) 4 commits
1090
- (merged to 'next' on 2013-08-28 at 876a39d)
1091
- + fast-import: allow moving the root tree
1092
- + fast-import: allow ls or filecopy of the root tree
1093
- + fast-import: set valid mode on root tree in "ls"
1094
- + t9300: document fast-import empty path issues
1226
+* nd/fetch-into-shallow (2013-08-28) 7 commits
1227
+ (merged to 'next' on 2013-09-09 at 87a3b99)
1228
+ + Add testcase for needless objects during a shallow fetch
1229
+ + list-objects: mark more commits as edges in mark_edges_uninteresting
1230
+ + list-objects: reduce one argument in mark_edges_uninteresting
1231
+ + upload-pack: delegate rev walking in shallow fetch to pack-objects
1232
+ + shallow: add setup_temporary_shallow()
1233
+ + shallow: only add shallow graft points to new shallow file
1234
+ + move setup_alternate_shallow and write_shallow_commits to shallow.c
1235
1096
- Originally merged to 'next' on 2013-07-22
1236
+ When there is no sufficient overlap between old and new history
1237
+ during a fetch into a shallow repository, we unnecessarily sent
1238
+ objects the sending side knows the receiving end has.
1239
1240
Will merge to 'master'.
1241
1242
1101
-* jc/diff-filter-negation (2013-07-19) 6 commits
1102
- (merged to 'next' on 2013-08-28 at 5cc7720)
1103
- + diff: deprecate -q option to diff-files
1104
- + diff: allow lowercase letter to specify what change class to exclude
1105
- + diff: reject unknown change class given to --diff-filter
1106
- + diff: preparse --diff-filter string argument
1107
- + diff: factor out match_filter()
1108
- + diff: pass the whole diff_options to diffcore_apply_filter()
1109
- (this branch is used by jc/hold-diff-remove-q-synonym-for-no-deletion.)
1243
+* jc/ls-files-killed-optim (2013-08-23) 4 commits
1244
+ (merged to 'next' on 2013-09-04 at 20c2304)
1245
+ + dir.c::test_one_path(): work around directory_exists_in_index_icase() breakage
1246
+ + t3010: update to demonstrate "ls-files -k" optimization pitfalls
1247
+ + ls-files -k: a directory only can be killed if the index has a non-directory
1248
+ + dir.c: use the cache_* macro to access the current index
1249
1111
- Originally merged to 'next' on 2013-07-22
1250
+ Originally merged to 'next' on 2013-08-27
1251
1113
- Teach "git diff --diff-filter" to express "I do not want to see
1114
- these classes of changes" more directly by listing only the
1115
- unwanted ones in lowercase (e.g. "--diff-filter=d" will show
1116
- everything but deletion) and deprecate "diff-files -q" which did
1117
- the same thing as "--diff-filter=d".
1252
+ "git ls-files -k" needs to crawl only the part of the working tree
1253
+ that may overlap the paths in the index to find killed files, but
1254
+ shared code with the logic to find all the untracked files, which
1255
+ made it unnecessarily inefficient.
1256
1119
- Will merge to 'master' in the second batch.
1257
+ Will merge to 'master' in the third batch.
1258
1259
1122
-* nd/magic-pathspec (2013-07-23) 48 commits
1123
- (merged to 'next' on 2013-08-28 at 305c6e9)
1124
- + rm: do not set a variable twice without intermediate reading.
1125
- + t6131 - skip tests if on case-insensitive file system
1126
- + parse_pathspec: accept :(icase)path syntax
1127
- + pathspec: support :(glob) syntax
1128
- + pathspec: make --literal-pathspecs disable pathspec magic
1129
- + pathspec: support :(literal) syntax for noglob pathspec
1130
- + kill limit_pathspec_to_literal() as it's only used by parse_pathspec()
1131
- + parse_pathspec: preserve prefix length via PATHSPEC_PREFIX_ORIGIN
1132
- + parse_pathspec: make sure the prefix part is wildcard-free
1133
- + rename field "raw" to "_raw" in struct pathspec
1134
- + tree-diff: remove the use of pathspec's raw[] in follow-rename codepath
1135
- + remove match_pathspec() in favor of match_pathspec_depth()
1136
- + remove init_pathspec() in favor of parse_pathspec()
1137
- + remove diff_tree_{setup,release}_paths
1138
- + convert common_prefix() to use struct pathspec
1139
- + convert add_files_to_cache to take struct pathspec
1140
- + convert {read,fill}_directory to take struct pathspec
1141
- + convert refresh_index to take struct pathspec
1142
- + convert report_path_error to take struct pathspec
1143
- + checkout: convert read_tree_some to take struct pathspec
1144
- + convert unmerge_cache to take struct pathspec
1145
- + convert run_add_interactive to use struct pathspec
1146
- + convert read_cache_preload() to take struct pathspec
1147
- + line-log: convert to use parse_pathspec
1148
- + reset: convert to use parse_pathspec
1149
- + add: convert to use parse_pathspec
1150
- + check-ignore: convert to use parse_pathspec
1151
- + archive: convert to use parse_pathspec
1152
- + ls-files: convert to use parse_pathspec
1153
- + rm: convert to use parse_pathspec
1154
- + checkout: convert to use parse_pathspec
1155
- + rerere: convert to use parse_pathspec
1156
- + status: convert to use parse_pathspec
1157
- + commit: convert to use parse_pathspec
1158
- + clean: convert to use parse_pathspec
1159
- + guard against new pathspec magic in pathspec matching code
1160
- + parse_pathspec: support prefixing original patterns
1161
- + parse_pathspec: support stripping/checking submodule paths
1162
- + parse_pathspec: support stripping submodule trailing slashes
1163
- + parse_pathspec: add special flag for max_depth feature
1164
- + convert some get_pathspec() calls to parse_pathspec()
1165
- + parse_pathspec: add PATHSPEC_PREFER_{CWD,FULL} flags
1166
- + parse_pathspec: save original pathspec for reporting
1167
- + add parse_pathspec() that converts cmdline args to struct pathspec
1168
- + pathspec: add copy_pathspec
1169
- + pathspec: i18n-ize error strings in pathspec parsing code
1170
- + move struct pathspec and related functions to pathspec.[ch]
1171
- + clean: remove unused variable "seen"
1172
- (this branch is used by jl/submodule-mv.)
1260
+* es/rebase-i-no-abbrev (2013-08-25) 3 commits
1261
+ (merged to 'next' on 2013-09-04 at 6027805)
1262
+ + rebase -i: fix short SHA-1 collision
1263
+ + t3404: rebase -i: demonstrate short SHA-1 collision
1264
+ + t3404: make tests more self-contained
1265
1174
- Originally merged to 'next' on 2013-07-23
1266
+ Originally merged to 'next' on 2013-08-26
1267
1176
- Use "struct pathspec" interface in more places, instead of array of
1177
- characters, the latter of which cannot express magic pathspecs
1178
- (e.g. ":(icase)makefile" that matches both Makefile and makefile).
1268
+ The commit object names in the insn sheet that was prepared at the
1269
+ beginning of "rebase -i" session can become ambiguous as the
1270
+ rebasing progresses and the repository gains more commits. Make
1271
+ sure the internal record is kept with full 40-hex object names.
1272
1180
- Will merge to 'master' in the second batch.
1273
+ Will merge to 'master' in the third batch.
1274
1275
1183
-* jc/check-x-z (2013-07-11) 4 commits
1184
- (merged to 'next' on 2013-08-28 at a7670ce)
1185
- + check-attr -z: a single -z should apply to both input and output
1186
- + check-ignore -z: a single -z should apply to both input and output
1187
- + check-attr: the name of the character is NUL, not NULL
1188
- + check-ignore: the name of the character is NUL, not NULL
1276
+* ks/p4-view-spec (2013-09-03) 2 commits
1277
+ - git p4: implement view spec wildcards with "p4 where"
1278
+ - git p4 test: sanitize P4CHARSET
1279
1190
- Originally merged to 'next' on 2013-07-15
1280
+ Replaced with a reroll that was whitespace damaged.
1281
1192
- "git check-ignore -z" applied the NUL termination to both its input
1193
- (with --stdin) and its output, but "git check-attr -z" ignored the
1194
- option on the output side.
1282
+ Waiting for the final Ack.
1283
1196
- This is potentially a backward incompatible fix. Let's see if
1197
- anybody screams before deciding if we want to do anything to help
1198
- existing users (there may be none).
1284
1200
- Will merge to 'master'.
1285
+* nd/push-no-thin (2013-08-13) 1 commit
1286
+ (merged to 'next' on 2013-09-04 at faa8c02)
1287
+ + push: respect --no-thin
1288
1289
+ Originally merged to 'next' on 2013-08-14
1290
1203
-* ms/fetch-prune-configuration (2013-07-18) 1 commit
1204
- (merged to 'next' on 2013-08-28 at fa6ffc7)
1205
- + fetch: make --prune configurable
1291
+ "git push --no-thin" was a no-op by mistake.
1292
1207
- Originally merged to 'next' on 2013-07-21
1293
+ Will merge to 'master' in the third batch.
1294
1209
- Allow fetch.prune and remote.*.prune configuration variables to be set,
1210
- and "git fetch" to behave as if "--prune" is given.
1295
1212
- "git fetch" that honors remote.*.prune is fine, but I wonder if we
1213
- should somehow make "git push" aware of it as well. Perhaps
1214
- remote.*.prune should not be just a boolean, but a 4-way "none",
1215
- "push", "fetch", "both"?
1296
+* sh/pull-rebase-preserve (2013-09-04) 1 commit
1297
+ (merged to 'next' on 2013-09-04 at 32a93bb)
1298
+ + pull: allow pull to preserve merges when rebasing
1299
1217
- Will merge to 'master' in the second batch.
1300
+ Originally merged to 'next' on 2013-08-14
1301
1302
+ "git pull --rebase" always flattened the history; pull.rebase can
1303
+ now be set to "preserve" to invoke "rebase --preserve-merges".
1304
1220
-* km/svn-1.8-serf-only (2013-07-18) 3 commits
1221
- (merged to 'next' on 2013-08-28 at 1119134)
1222
- + Git.pm: revert _temp_cache use of temp_is_locked
1223
- + git-svn: allow git-svn fetching to work using serf
1224
- + Git.pm: add new temp_is_locked function
1305
+ Will merge to 'master' in the third batch.
1306
1226
- Originally merged to 'next' on 2013-07-19
1307
1228
- Subversion 1.8.0 that was recently released breaks older subversion
1229
- clients coming over http/https in various ways.
1308
+* jn/post-receive-utf8 (2013-08-05) 3 commits
1309
+ (merged to 'next' on 2013-09-04 at 3a3f480)
1310
+ + hooks/post-receive-email: set declared encoding to utf-8
1311
+ + hooks/post-receive-email: force log messages in UTF-8
1312
+ + hooks/post-receive-email: use plumbing instead of git log/show
1313
1231
- Will merge to 'master'.
1314
+ Originally merged to 'next' on 2013-08-20
1315
+
1316
+ Update post-receive-email script to make sure the message contents
1317
+ and pathnames are encoded consistently in UTF-8.
1318
+
1319
+ Will merge to 'master' in the third batch.
1320
1321
1322
* tr/merge-recursive-index-only (2013-07-07) 3 commits
@@ -1239,37 +1327,6 @@ of the repositories listed at
1327
Will hold, until we get any user.
1328
1329
1242
-* jc/push-cas (2013-08-02) 9 commits
1243
- (merged to 'next' on 2013-08-28 at 784281f)
1244
- + push: teach --force-with-lease to smart-http transport
1245
- + send-pack: fix parsing of --force-with-lease option
1246
- + t5540/5541: smart-http does not support "--force-with-lease"
1247
- + t5533: test "push --force-with-lease"
1248
- + push --force-with-lease: tie it all together
1249
- + push --force-with-lease: implement logic to populate old_sha1_expect[]
1250
- + remote.c: add command line option parser for "--force-with-lease"
1251
- + builtin/push.c: use OPT_BOOL, not OPT_BOOLEAN
1252
- + cache.h: move remote/connect API out of it
1253
-
1254
- Originally merged to 'next' on 2013-08-05
1255
-
1256
- Allow a safer "rewind of the remote tip" push than blind "--force",
1257
- by requiring that the overwritten remote ref to be unchanged since
1258
- the new history to replace it was prepared.
1259
-
1260
- The machinery is more or less ready. The "--force" option is again
1261
- the big red button to override any safety, thanks to J6t's sanity
1262
- (the original round allowed --lockref to defeat --force).
1263
-
1264
- The logic to choose the default implemented here is fragile
1265
- (e.g. "git fetch" after seeing a failure will update the
1266
- remote-tracking branch and will make the next "push" pass,
1267
- defeating the safety pretty easily). It is suitable only for the
1268
- simplest workflows, and it may hurt users more than it helps them.
1269
-
1270
- Will merge to 'master' in the second batch.
1271
-
1272
-
1330
* jc/pull-training-wheel (2013-07-19) 1 commit
1331
(merged to 'next' on 2013-08-28 at c39bd15)
1332
+ pull: require choice between rebase/merge on non-fast-forward pull
@@ -1280,14 +1337,10 @@ of the repositories listed at
1337
integrate from where) refuse with "it does not fast forward; choose
1338
between 'pull --merge' and 'pull --rebase'".
1339
1283
- See
1284
-
1285
- http://thread.gmane.org/gmane.comp.version-control.git/225146/focus=225326
1286
- http://thread.gmane.org/gmane.comp.version-control.git/225146/focus=229221
1287
-
1288
- for full discussion.
1340
+ Will revert the merge, wait for the discussion to settle and
1341
+ produce a replacement to queue again.
1342
1290
- Will merge to 'master' in the second batch.
1343
+ http://thread.gmane.org/gmane.comp.version-control.git/233554/focus=234375
1344
1345
1346
* jc/reflog-doc (2013-06-19) 1 commit
@@ -1328,6 +1381,5 @@ of the repositories listed at
1381
1382
* jc/hold-diff-remove-q-synonym-for-no-deletion (2013-07-19) 1 commit
1383
- diff: remove "diff-files -q" in a version of Git in a distant future
1331
- (this branch uses jc/diff-filter-negation.)
1384
1385
Will cook in 'next' until a distant future.