What's cooking (2016/03 #01)

Junio C Hamano committed Mar 2, 2016 at 12:57 UTC 482ffba9a87503b6e585d2c6a5c61bf3f14f7334
1 file changed +230 -342
whats-cooking.txt
+230 -342
@@ -1,10 +1,10 @@
1 To: git@vger.kernel.org
2 Bcc: lwn@lwn.net
3 -Subject: What's cooking in git.git (Feb 2016, #08; Fri, 26)
3 +Subject: What's cooking in git.git (Mar 2016, #01; Wed, 2)
4 X-master-at: f02fbc4f9433937ee0463d0342d6d7d97e1f6f1e
5 -X-next-at: f92f5fe29a68d654894c6a10c5f5b3f0ce4bb1a5
5 +X-next-at: 391b917453d3b3cca00e679379036ed0e6d85125
6
7 -What's cooking in git.git (Feb 2016, #08; Fri, 26)
7 +What's cooking in git.git (Mar 2016, #01; Wed, 2)
8 --------------------------------------------------
9
10 Here are the topics that have been cooking. Commits prefixed with
@@ -12,23 +12,16 @@ Here are the topics that have been cooking. Commits prefixed with
12 '+' are in 'next'. The ones marked with '.' do not appear in any of
13 the integration branches, but I am still holding onto them.
14
15 -v2.8-rc0 has been tagged; this is a preview for the next release.
15 +v2.8-rc1 at around the end of this week.
16
17 -In previous cycles, I often left many topics in 'next' when tagging
18 -this zero-th preview, but eventually merged them before the final.
19 -I decided to do things a bit differently for this cycle: a topic,
20 -once it hits 'next', will not be rewound and only refined and
21 -corrected with incremental updates, so the only effect such a late
22 -merge to 'master' before the final is that some topics are not as
23 -widely tested on 'master' before the final one is tagged.
24 -
25 -So this -rc0 is deliberately aggressive in that it includes all
26 -topics that have been cooking in 'next' that I think we can fix bugs
27 -that might still lurking in them before the final (it merges 25
28 -topics since the last batch to 'master'). The topics not merged to
29 -this preview, on the other hand, will not be considered for 2.8
30 -final, even though I might later succumb to the temptation to pick
31 -up ones that are in 'next' as of today ;-)
17 +Again, the topics that have not been cooked sufficiently in 'next'
18 +at this point will not be considered for 2.8 final, even though I
19 +might later succumb to the temptation to pick up ones that look
20 +trivially correct. Those who have their topics merged to 'master'
21 +since v2.7 are expected to focus on responding to regressions and
22 +remaining bugs in their topics in 'master', and strongly encouraged
23 +to actively hunt for regressions and remaining bugs there, not in a
24 +random shiny new feature, during the pre-release period.
25
26 You can find the changes described here in the integration branches of the
27 repositories listed at
@@ -36,313 +29,167 @@ repositories listed at
29 http://git-blame.blogspot.com/p/git-public-repositories.html
30
31 --------------------------------------------------
39 -[Graduated to "master"]
32 +[New Topics]
33
41 -* ak/extract-argv0-last-dir-sep (2016-02-19) 1 commit
42 - (merged to 'next' on 2016-02-24 at f590eb0)
43 - + exec_cmd.c: use find_last_dir_sep() for code simplification
34 +* nd/i18n-2.8.0 (2016-02-29) 4 commits
35 + (merged to 'next' on 2016-03-01 at cdf4675)
36 + + trailer.c: mark strings for translation
37 + + ref-filter.c: mark strings for translation
38 + + builtin/clone.c: mark strings for translation
39 + + builtin/checkout.c: mark strings for translation
40
45 - Code simplification.
41 + Will merge to 'master' by 2.8-rc1
42
43
48 -* ak/git-strip-extension-from-dashed-command (2016-02-21) 1 commit
49 - (merged to 'next' on 2016-02-24 at 8a84f91)
50 - + git.c: simplify stripping extension of a file in handle_builtin()
51 -
52 - Code simplification.
53 -
54 -
55 -* dt/initial-ref-xn-commit-doc (2016-02-25) 1 commit
56 - (merged to 'next' on 2016-02-25 at e32d0f6)
57 - + refs: document transaction semantics
44 +* mm/readme-markdown (2016-02-27) 1 commit
45 + (merged to 'next' on 2016-03-01 at 81f3858)
46 + + README.md: don't take 'commandname' literally
47
48 + The top-level README file has been updated to be more appropriate
49 + for the sign on the front door to welcome new acquaintances to Git
50 + by toning down inside jokes and making it into MarkDown.
51
60 -* fa/merge-recursive-no-rename (2016-02-24) 7 commits
61 - (merged to 'next' on 2016-02-25 at 8f18cc1)
62 - + t3034: test deprecated interface
63 - + t3034: test option to disable renames
64 - + t3034: add rename threshold tests
65 - + merge-recursive: find-renames resets threshold
66 - + merge-strategies.txt: fix typo
67 - (merged to 'next' on 2016-02-18 at dd6125b)
68 - + merge-recursive: more consistent interface
69 - + merge-recursive: option to disable renames
52 + Will merge to 'master' after 2.8 final
53
71 - "git merge-recursive" learned "--no-renames" option to disable its
72 - rename detection logic.
54
55 +* gf/fetch-pack-direct-object-fetch (2016-03-01) 1 commit
56 + - fetch-pack: fix object_id of exact sha1
57
75 -* jg/credential-cache-chdir-to-sockdir (2016-02-23) 3 commits
76 - (merged to 'next' on 2016-02-25 at ad35811)
77 - + credential-cache--daemon: change to the socket dir on startup
78 - + credential-cache--daemon: disallow relative socket path
79 - + credential-cache--daemon: refactor check_socket_directory
80 -
81 - The "credential-cache" daemon process used to run in whatever
82 - directory it happened to start in, but this made umount(2)ing the
83 - filesystem that houses the repository harder; now the process
84 - chdir()s to the directory that house its own socket on startup.
85 -
86 -
87 -* jk/epipe-in-async (2016-02-25) 4 commits
88 - (merged to 'next' on 2016-02-25 at 4482ddf)
89 - + t5504: handle expected output from SIGPIPE death
90 - + test_must_fail: report number of unexpected signal
91 - + fetch-pack: ignore SIGPIPE in sideband demuxer
92 - + write_or_die: handle EPIPE in async threads
93 -
94 - Handling of errors while writing into our internal asynchronous
95 - process has been made more robust, which reduces flakiness in our
96 - tests.
97 -
98 -
99 -* jk/grep-binary-workaround-in-test (2016-02-23) 2 commits
100 - (merged to 'next' on 2016-02-25 at 0e96ae8)
101 - + t9200: avoid grep on non-ASCII data
102 - + t8005: avoid grep on non-ASCII data
103 -
104 - Recent versions of GNU grep are pickier when their input contains
105 - arbitrary binary data, which some of our tests uses. Rewrite the
106 - tests to sidestep the problem.
58 + Fetching of history by naming a commit object name directly didn't
59 + work across remote-curl transport.
60
61 + Will merge to 'next'.
62
109 -* jk/more-comments-on-textconv (2016-02-22) 1 commit
110 - (merged to 'next' on 2016-02-24 at 4ccfdc5)
111 - + diff: clarify textconv interface
63
113 - The memory ownership rule of fill_textconv() API, which was a bit
114 - tricky, has been documented a bit better.
64 +* jk/add-i-highlight (2016-02-28) 1 commit
65 + - add--interactive: allow custom diff highlighting programs
66
67 + Will merge to 'next'.
68
117 -* jk/no-diff-emit-common (2016-02-22) 3 commits
118 - (merged to 'next' on 2016-02-24 at 72626ea)
119 - + xdiff: drop XDL_EMIT_COMMON
120 - + merge-tree: drop generate_common strategy
121 - + merge-one-file: use empty blob for add/add base
69
123 - "git merge-tree" used to mishandle "both sides added" conflict with
124 - its own "create a fake ancestor file that has the common parts of
125 - what both sides have added and do a 3-way merge" logic; this has
126 - been updated to use the usual "3-way merge with an empty blob as
127 - the fake common ancestor file" approach used in the rest of the
128 - system.
70 +* jk/config-get-urlmatch (2016-02-28) 3 commits
71 + - Documentation/git-config: fix --get-all description
72 + - Documentation/git-config: use bulleted list for exit codes
73 + - config: fail if --get-urlmatch finds no value
74
75 + "git config --get-urlmatch", unlike other variants of the "git
76 + config --get" family, did not signal error with its exit status
77 + when there was no matching configuration.
78
131 -* jk/tighten-alloc (2016-02-22) 22 commits
132 - (merged to 'next' on 2016-02-24 at 78b3251)
133 - + ewah: convert to REALLOC_ARRAY, etc
134 - + convert ewah/bitmap code to use xmalloc
135 - + diff_populate_gitlink: use a strbuf
136 - + transport_anonymize_url: use xstrfmt
137 - + git-compat-util: drop mempcpy compat code
138 - + sequencer: simplify memory allocation of get_message
139 - + test-path-utils: fix normalize_path_copy output buffer size
140 - + fetch-pack: simplify add_sought_entry
141 - + fast-import: simplify allocation in start_packfile
142 - + write_untracked_extension: use FLEX_ALLOC helper
143 - + prepare_{git,shell}_cmd: use argv_array
144 - + use st_add and st_mult for allocation size computation
145 - + convert trivial cases to FLEX_ARRAY macros
146 - + use xmallocz to avoid size arithmetic
147 - + convert trivial cases to ALLOC_ARRAY
148 - + convert manual allocations to argv_array
149 - + argv-array: add detach function
150 - + add helpers for allocating flex-array structs
151 - + harden REALLOC_ARRAY and xcalloc against size_t overflow
152 - + tree-diff: catch integer overflow in combine_diff_path allocation
153 - + add helpers for detecting size_t overflow
154 - + reflog_expire_cfg: NUL-terminate pattern field
155 -
156 - Update various codepaths to avoid manually-counted malloc().
157 -
158 -
159 -* js/config-set-in-non-repository (2016-02-25) 1 commit
160 - (merged to 'next' on 2016-02-25 at a675267)
161 - + git config: report when trying to modify a non-existing repo config
162 -
163 - "git config section.var value" to set a value in per-repository
164 - configuration file failed when it was run outside any repository,
165 - but didn't say the reason correctly.
166 -
167 -
168 -* kn/ref-filter-atom-parsing (2016-02-17) 11 commits
169 - (merged to 'next' on 2016-02-18 at 9606218)
170 - + ref-filter: introduce objectname_atom_parser()
171 - + ref-filter: introduce contents_atom_parser()
172 - + ref-filter: introduce remote_ref_atom_parser()
173 - + ref-filter: align: introduce long-form syntax
174 - + ref-filter: introduce align_atom_parser()
175 - + ref-filter: introduce parse_align_position()
176 - + ref-filter: introduce color_atom_parser()
177 - + ref-filter: introduce parsing functions for each valid atom
178 - + ref-filter: introduce struct used_atom
179 - + ref-filter: bump 'used_atom' and related code to the top
180 - + ref-filter: use string_list_split over strbuf_split
181 -
182 - The ref-filter's format-parsing code has been refactored, in
183 - preparation for "branch --format" and friends.
184 -
185 -
186 -* ls/config-origin (2016-02-22) 4 commits
187 - (merged to 'next' on 2016-02-24 at 239fcc2)
188 - + config: add '--show-origin' option to print the origin of a config value
189 - + config: add 'origin_type' to config_source struct
190 - + rename git_config_from_buf to git_config_from_mem
191 - + t: do not hide Git's exit code in tests using 'nul_to_q'
192 -
193 - The configuration system has been taught to phrase where it found a
194 - bad configuration variable in a better way in its error messages.
195 - "git config" learnt a new "--show-origin" option to indicate where
196 - the values come from.
197 -
198 -
199 -* ls/makefile-cflags-developer-tweak (2016-02-25) 1 commit
200 - (merged to 'next' on 2016-02-25 at 14bb065)
201 - + add DEVELOPER makefile knob to check for acknowledged warnings
202 -
203 - There is a new DEVELOPER knob that enables many compiler warning
204 - options in the Makefile.
205 -
206 -
207 -* ma/update-hooks-sample-typofix (2016-02-25) 1 commit
208 - (merged to 'next' on 2016-02-25 at be49c79)
209 - + templates/hooks: fix minor typo in the sample update-hook
210 -
211 -
212 -* mg/work-tree-tests (2016-02-20) 1 commit
213 - (merged to 'next' on 2016-02-25 at ca8cbb0)
214 - + tests: rename work-tree tests to *work-tree*
215 -
216 - Traditionally, the tests that try commands that work on the
217 - contents in the working tree were named with "worktree" in their
218 - filenames, but with the recent addition of "git worktree"
219 - subcommand, whose tests are also named similarly, it has become
220 - harder to tell them apart. The traditional tests have been renamed
221 - to use "work-tree" instead in an attempt to differentiate them.
222 -
223 -
224 -* mm/push-default-warning (2016-02-25) 1 commit
225 - (merged to 'next' on 2016-02-25 at 753352f)
226 - + push: remove "push.default is unset" warning message
227 -
228 - Across the transition at around Git version 2.0, the user used to
229 - get a pretty loud warning when running "git push" without setting
230 - push.default configuration variable. We no longer warn, given that
231 - the transition is over long time ago.
232 -
233 -
234 -* mm/push-simple-doc (2016-02-23) 1 commit
235 - (merged to 'next' on 2016-02-25 at dc4f57c)
236 - + Documentation/git-push: document that 'simple' is the default
237 -
238 - The documentation did not clearly state that the 'simple' mode is
239 - now the default for "git push" when push.default configuration is
240 - not set.
241 -
242 -
243 -* mm/readme-markdown (2016-02-25) 5 commits
244 - (merged to 'next' on 2016-02-25 at b8bda29)
245 - + README.md: move down historical explanation about the name
246 - + README.md: don't call git stupid in the title
247 - + README.md: move the link to git-scm.com up
248 - + README.md: add hyperlinks on filenames
249 - + README: use markdown syntax
79 + Will merge to 'next'.
80
251 - README has been renamed to README.md and its contents got tweaked
252 - slightly to make it easier on the eyes.
81
82 +* jk/rev-parse-local-env-vars (2016-02-29) 2 commits
83 + - rev-parse: let some options run outside repository
84 + - t1515: add tests for rev-parse out-of-repo helpers
85
255 -* ps/config-error (2016-02-22) 15 commits
256 - (merged to 'next' on 2016-02-25 at b378629)
257 - + config: rename git_config_set_or_die to git_config_set
258 - + config: rename git_config_set to git_config_set_gently
259 - + compat: die when unable to set core.precomposeunicode
260 - + sequencer: die on config error when saving replay opts
261 - + init-db: die on config errors when initializing empty repo
262 - + clone: die on config error in cmd_clone
263 - + remote: die on config error when manipulating remotes
264 - + remote: die on config error when setting/adding branches
265 - + remote: die on config error when setting URL
266 - + submodule--helper: die on config error when cloning module
267 - + submodule: die on config error when linking modules
268 - + branch: die on config error when editing branch description
269 - + branch: die on config error when unsetting upstream
270 - + branch: report errors in tracking branch setup
271 - + config: introduce set_or_die wrappers
86 + The "--local-env-vars" and "--resolve-git-dir" options of "git
87 + rev-parse" failed to work outside a repository when the command's
88 + option parsing was rewritten in 1.8.5 era.
89
273 - Many codepaths forget to check return value from git_config_set();
274 - the function is made to die() to make sure we do not proceed when
275 - setting a configuration variable failed.
90 + Will merge to 'next'.
91
92
278 -* ps/plug-xdl-merge-leak (2016-02-23) 1 commit
279 - (merged to 'next' on 2016-02-25 at 523609e)
280 - + xdiff/xmerge: fix memory leak in xdl_merge
93 +* sg/completion-updates (2016-02-28) 21 commits
94 + - completion: cache the path to the repository
95 + - completion: extract repository discovery from __gitdir()
96 + - completion: don't guard git executions with __gitdir()
97 + - completion: consolidate silencing errors from git commands
98 + - completion: don't use __gitdir() for git commands
99 + - completion: respect 'git -C <path>'
100 + - completion: fix completion after 'git -C <path>'
101 + - completion: don't offer commands when 'git --opt' needs an argument
102 + - rev-parse: add '--absolute-git-dir' option
103 + - completion: list short refs from a remote given as a URL
104 + - completion: don't list 'HEAD' when trying refs completion outside of a repo
105 + - completion: list refs from remote when remote's name matches a directory
106 + - completion: respect 'git --git-dir=<path>' when listing remote refs
107 + - completion: fix most spots not respecting 'git --git-dir=<path>'
108 + - completion: ensure that the repository path given on the command line exists
109 + - completion tests: add tests for the __git_refs() helper function
110 + - completion tests: check __gitdir()'s output in the error cases
111 + - completion tests: consolidate getting path of current working directory
112 + - completion tests: make the $cur variable local to the test helper functions
113 + - completion tests: don't add test cruft to the test repository
114 + - completion: improve __git_refs()'s in-code documentation
115
116 + Will be rerolled.
117 + ($gmane/287839)
118
283 -* sb/submodule-fetch-nontip (2016-02-24) 1 commit
284 - (merged to 'next' on 2016-02-25 at 1e412f0)
285 - + submodule: try harder to fetch needed sha1 by direct fetching sha1
119
287 - When "git submodule update" did not result in fetching the commit
288 - object in the submodule that is referenced by the superproject, the
289 - command learned to retry another fetch, specifically asking for
290 - that commit that may not be connected to the refs it usually
291 - fetches.
120 +* tb/avoid-gcc-on-darwin-10-6 (2016-02-28) 1 commit
121 + (merged to 'next' on 2016-03-01 at e8dd08a)
122 + + config.mak.uname: use clang for Mac OS X 10.6
123
124 + Out of maintenance gcc on OSX 10.6 fails to compile the code in
125 + 'master'; work it around by using clang by default on the platform.
126
294 -* sb/submodule-module-list-fix (2016-02-24) 1 commit
295 - (merged to 'next' on 2016-02-25 at 21be607)
296 - + submodule helper list: respect correct path prefix
127 + Will merge to 'master' by 2.8-rc1
128
298 - A helper function "git submodule" uses since v2.7.0 to list the
299 - modules that match the pathspec argument given to its subcommands
300 - (e.g. "submodule add <repo> <path>") has been fixed.
129
130 +* jk/tighten-alloc (2016-02-29) 1 commit
131 + (merged to 'next' on 2016-03-01 at f4df936)
132 + + compat/mingw: brown paper bag fix for 50a6c8e
133
303 -* tb/conversion (2016-02-23) 7 commits
304 - (merged to 'next' on 2016-02-25 at 1bb466d)
305 - + convert.c: correct attr_action()
306 - (merged to 'next' on 2016-02-12 at 6faf27b)
307 - + convert.c: simplify text_stat
308 - + convert.c: refactor crlf_action
309 - + convert.c: use text_eol_is_crlf()
310 - + convert.c: remove input_crlf_action()
311 - + convert.c: remove unused parameter 'path'
312 - + t0027: add tests for get_stream_filter()
134 + Will merge to 'master' by 2.8-rc1
135
314 - Code simplification.
136
137 +* js/pthread-exit-emu-windows (2016-03-02) 1 commit
138 + (merged to 'next' on 2016-03-02 at 391b917)
139 + + Mark win32's pthread_exit() as NORETURN
140
317 -* tg/git-remote (2016-02-16) 4 commits
318 - (merged to 'next' on 2016-02-18 at 091d43e)
319 - + remote: use remote_is_configured() for add and rename
320 - + remote: actually check if remote exits
321 - + remote: simplify remote_is_configured()
322 - + remote: use parse_config_key
141 + Will merge to 'master' by 2.8-rc1
142
324 - The internal API to interact with "remote.*" configuration
325 - variables has been streamlined.
143
327 ---------------------------------------------------
328 -[New Topics]
144 +* jx/http-no-proxy (2016-02-29) 1 commit
145 + - http: honor no_http env variable to bypass proxy
146
330 -* jk/credential-clear-config (2016-02-26) 1 commit
331 - - credential: let empty credential specs reset helper list
147 + Needs review.
148
333 - The credential.helper configuration variable is cumulative and
334 - there is no good way to override it from the command line. As
335 - a special case, giving an empty string as its value now serves
336 - as the signal to clear the values specified in various files.
149
150 +* mm/lockfile-error-message (2016-03-01) 2 commits
151 + - lockfile: improve error message when lockfile exists
152 + - lockfile: mark strings for translation
153
339 -* jk/submodule-c-credential (2016-02-25) 1 commit
340 - - git: submodule honor -c credential.* from command line
341 - (this branch uses mg/httpd-tests-update-for-apache-2.4.)
154 + Will merge to 'next'.
155
343 - "git -c credential.<var>=<value> submodule" can now be used to
344 - propagate configuration variables related to credential helper
345 - down to the submodules.
156 +
157 +* sb/submodule-parallel-fetch (2016-03-01) 1 commit
158 + (merged to 'next' on 2016-03-01 at b47ab6e)
159 + + run-command: do not pass child process data into callbacks
160 + (this branch is used by sb/submodule-init and sb/submodule-parallel-update.)
161 +
162 + Simplify the two callback functions that are triggered when the
163 + child process terminates to avoid misuse of the child-process
164 + structure that has already been cleaned up.
165 +
166 + Will merge to 'master' by 2.8-rc1
167 +
168 +
169 +* ss/exc-flag-is-a-collection-of-bits (2016-03-01) 1 commit
170 + - dir: store EXC_FLAG_* values in unsigned integers
171 +
172 + Will merge to 'next'.
173 +
174 +
175 +* jc/bundle (2016-03-02) 4 commits
176 + - bundle v3: the beginning
177 + - bundle: keep a copy of bundle file name in the in-core bundle header
178 + - bundle: plug resource leak
179 + - bundle doc: 'verify' is not about verifying the bundle
180 +
181 + The beginning of "split bundle", which could be one of the
182 + ingredients to allow "git clone" traffic off of the core server
183 + network to CDN.
184 +
185 +
186 +* ss/receive-pack-parse-options (2016-03-01) 1 commit
187 + - builtin/receive-pack.c: use parse_options API
188 +
189 + The command line argument parser for "receive-pack" has been
190 + rewritten to use parse-options.
191 +
192 + Will merge to 'next'.
193
194 --------------------------------------------------
195 [Stalled]
@@ -418,10 +265,44 @@ repositories listed at
265 --------------------------------------------------
266 [Cooking]
267
268 +* jk/credential-clear-config (2016-02-26) 1 commit
269 + - credential: let empty credential specs reset helper list
270 +
271 + The credential.helper configuration variable is cumulative and
272 + there is no good way to override it from the command line. As
273 + a special case, giving an empty string as its value now serves
274 + as the signal to clear the values specified in various files.
275 +
276 + Will merge to 'next'.
277 +
278 +
279 +* jk/submodule-c-credential (2016-03-01) 6 commits
280 + - git: submodule honor -c credential.* from command line
281 + - quote: implement sq_quotef()
282 + - submodule: fix segmentation fault in submodule--helper clone
283 + - submodule: fix submodule--helper clone usage
284 + - submodule: check argc count for git submodule--helper clone
285 + - submodule: don't pass empty string arguments to submodule--helper clone
286 + (this branch uses mg/httpd-tests-update-for-apache-2.4.)
287 +
288 + "git -c credential.<var>=<value> submodule" can now be used to
289 + propagate configuration variables related to credential helper
290 + down to the submodules.
291 +
292 + Will merge to 'next'.
293 +
294 +
295 * mg/httpd-tests-update-for-apache-2.4 (2016-02-25) 1 commit
422 - - t/lib-httpd: load mod_unixd
296 + (merged to 'next' on 2016-03-01 at d2f7e8c)
297 + + t/lib-httpd: load mod_unixd
298 (this branch is used by jk/submodule-c-credential.)
299
300 + The way the test scripts configure the Apache web server has been
301 + updated to work also for Apache 2.4 running on RedHat derived
302 + distros.
303 +
304 + Will merge to 'master' by 2.8-rc1
305 +
306
307 * nd/shallow-deepen (2016-02-23) 25 commits
308 - fetch, upload-pack: --deepen=N extends shallow boundary by N commits
@@ -461,7 +342,9 @@ repositories listed at
342 Needs review.
343
344
464 -* jk/pack-idx-corruption-safety (2016-02-25) 3 commits
345 +* jk/pack-idx-corruption-safety (2016-02-27) 4 commits
346 + (merged to 'next' on 2016-03-01 at 49e08d3)
347 + + sha1_file.c: mark strings for translation
348 (merged to 'next' on 2016-02-26 at ef0d6de)
349 + use_pack: handle signed off_t overflow
350 + nth_packed_object_offset: bounds-check extended offset
@@ -471,7 +354,7 @@ repositories listed at
354 idx file has been made more carefully check the validity of the
355 data in the idx.
356
474 - Will merge to 'master'.
357 + Will merge to 'master' by 2.8-rc1
358
359
360 * mm/diff-renames-default (2016-02-25) 5 commits
@@ -485,7 +368,7 @@ repositories listed at
368 The end-user facing Porcelain level commands like "diff" and "log"
369 now enables the rename detection by default.
370
488 - Will merge to 'master'.
371 + Will merge to 'master' after 2.8 final
372
373
374 * mp/upload-pack-use-embedded-args (2016-02-25) 1 commit
@@ -496,18 +379,17 @@ repositories listed at
379 the command line to run pack-objects instead of using a separate
380 array of strings.
381
499 - Will merge to 'master'.
382 + Will merge to 'master' after 2.8 final
383
384
502 -* sb/submodule-init (2016-02-25) 2 commits
385 +* sb/submodule-init (2016-03-01) 2 commits
386 - submodule: port init from shell to C
387 - submodule: port resolve_relative_url from shell to C
505 - (this branch uses sb/submodule-parallel-update; is tangled with dt/refs-backend-lmdb.)
388 + (this branch uses sb/submodule-parallel-fetch and sb/submodule-parallel-update.)
389
390 Update of "git submodule" to move pieces of logic to C continues.
391
509 - Comments?
510 - ($gmane/286092)
392 + Needs review.
393
394
395 * az/p4-bare-no-rebase (2016-02-19) 1 commit
@@ -558,68 +440,74 @@ repositories listed at
440 ($gmane/286137)
441
442
561 -* sb/submodule-parallel-update (2016-02-25) 11 commits
443 +* sb/submodule-parallel-update (2016-03-01) 10 commits
444 - clone: allow an explicit argument for parallel submodule clones
445 - submodule update: expose parallelism to the user
446 + - submodule helper: remove double 'fatal: ' prefix
447 - git submodule update: have a dedicated helper for cloning
565 - - run_processes_parallel: correctly terminate callbacks with an LF
448 - run_processes_parallel: rename parameters for the callbacks
567 - - run-command: expose default_{start_failure, task_finished}
449 - run_processes_parallel: treat output of children as byte array
450 - submodule update: direct error message to stderr
451 - fetching submodules: respect `submodule.fetchJobs` config option
452 - submodule-config: drop check against NULL
453 - submodule-config: keep update strategy around
573 - (this branch is used by dt/refs-backend-lmdb and sb/submodule-init.)
454 + (this branch is used by sb/submodule-init; uses sb/submodule-parallel-fetch.)
455
456 A major part of "git submodule update" has been ported to C to take
457 advantage of the recently added framework to run download tasks in
458 parallel.
459
579 - Seems to break tests when merged to the tip of 'pu'.
580 -
581 -
582 -* dt/refs-backend-lmdb (2016-02-25) 34 commits
583 - - SQUASH??? Minimum compilation band-aid
584 - - tests: add ref-storage argument
585 - - refs: tests for lmdb backend
586 - - refs: add LMDB refs storage backend
587 - - refs: break out resolve_ref_unsafe_submodule
588 - - config: read ref storage config on startup
589 - - refs: register ref storage backends
590 - - svn: learn ref-storage argument
591 - - clone: allow ref storage backend to be set for clone
592 - - refs: check submodules' ref storage config
593 - - init: allow alternate ref strorage to be set for new repos
594 - - refs: always handle non-normal refs in files backend
595 - - refs: resolve symbolic refs first
596 - - refs: on symref reflog expire, lock symref not referrent
597 - - refs: don't dereference on rename
598 - - refs: allow log-only updates
599 - - refs: move duplicate check to common code
600 - - refs: make lock generic
601 - - refs: handle non-normal ref renames
602 - - refs: add method to rename refs
603 - - refs: add methods to init refs db
604 - - refs: add method for delete_refs
605 - - refs: add method for initial ref transaction commit
606 - - refs: add methods for reflog
607 - - refs: add do_for_each_per_worktree_ref
608 - - refs: reduce the visibility of do_for_each_ref()
609 - - refs: add method for do_for_each_ref
610 - - refs: add methods for misc ref operations
611 - - refs: add a backend method structure with transaction functions
612 - - refs: move resolve_ref_unsafe into common code
613 - - files-backend: break out ref reading
614 - - refs: move for_each_*ref* functions into common code
615 - - refs: move head_ref{,_submodule} to the common code
616 - - Merge branch 'sb/submodule-parallel-update' into dt/refs-backend-lmdb
617 - (this branch uses sb/submodule-parallel-update; is tangled with sb/submodule-init.)
618 -
619 - Building on top of a few refs-backend preparatory series, LMDB
620 - based refs backend has been plugged into the system.
621 -
622 - ($gmane/286572)
460 + Will merge to 'next'.
461 +
462 +
463 +* dt/refs-backend-lmdb (2016-02-25) 45 commits
464 + . SQUASH??? Minimum compilation band-aid
465 + . tests: add ref-storage argument
466 + . refs: tests for lmdb backend
467 + . refs: add LMDB refs storage backend
468 + . refs: break out resolve_ref_unsafe_submodule
469 + . config: read ref storage config on startup
470 + . refs: register ref storage backends
471 + . svn: learn ref-storage argument
472 + . clone: allow ref storage backend to be set for clone
473 + . refs: check submodules' ref storage config
474 + . init: allow alternate ref strorage to be set for new repos
475 + . refs: always handle non-normal refs in files backend
476 + . refs: resolve symbolic refs first
477 + . refs: on symref reflog expire, lock symref not referrent
478 + . refs: don't dereference on rename
479 + . refs: allow log-only updates
480 + . refs: move duplicate check to common code
481 + . refs: make lock generic
482 + . refs: handle non-normal ref renames
483 + . refs: add method to rename refs
484 + . refs: add methods to init refs db
485 + . refs: add method for delete_refs
486 + . refs: add method for initial ref transaction commit
487 + . refs: add methods for reflog
488 + . refs: add do_for_each_per_worktree_ref
489 + . refs: reduce the visibility of do_for_each_ref()
490 + . refs: add method for do_for_each_ref
491 + . refs: add methods for misc ref operations
492 + . refs: add a backend method structure with transaction functions
493 + . refs: move resolve_ref_unsafe into common code
494 + . files-backend: break out ref reading
495 + . refs: move for_each_*ref* functions into common code
496 + . refs: move head_ref{,_submodule} to the common code
497 + . Merge branch 'sb/submodule-parallel-update' into dt/refs-backend-lmdb
498 + . clone: allow an explicit argument for parallel submodule clones
499 + . submodule update: expose parallelism to the user
500 + . git submodule update: have a dedicated helper for cloning
501 + . run_processes_parallel: correctly terminate callbacks with an LF
502 + . run_processes_parallel: rename parameters for the callbacks
503 + . run-command: expose default_{start_failure, task_finished}
504 + . run_processes_parallel: treat output of children as byte array
505 + . submodule update: direct error message to stderr
506 + . fetching submodules: respect `submodule.fetchJobs` config option
507 + . submodule-config: drop check against NULL
508 + . submodule-config: keep update strategy around
509 +
510 + A reroll exists, but it seems that will further be rerolled.
511
512
513 * ad/cygwin-wants-rename (2015-08-07) 1 commit