What's cooking (2009/08 #06)
Junio C Hamano committed
Aug 30, 2009 at 23:58 UTC
f626cdc06b65b35297281ec2f71a39ea9269abf2
1 file changed
+194
-163
whats-cooking.txt
+194
-163
@@ -1,9 +1,9 @@
1
To: git@vger.kernel.org
2
-Subject: What's cooking in git.git (Aug 2009, #05; Wed, 26)
3
-X-master-at: 68ea4741643d9e7c4bdac7cbbe6292edc69430ef
4
-X-next-at: 4fc77847c70f00e94586150f808d0418ca64bbb8
2
+Subject: What's cooking in git.git (Aug 2009, #06; Sun, 30)
3
+X-master-at: f324cb50be3d88c6fb6123c85da749695a749f60
4
+X-next-at: 5fc6248250e68c3b723efb9b868d8fc27499065a
5
6
-What's cooking in git.git (Aug 2009, #05; Wed, 26)
6
+What's cooking in git.git (Aug 2009, #06; Sun, 30)
7
--------------------------------------------------
8
9
Here are the topics that have been cooking. Commits prefixed with '-' are
@@ -20,29 +20,38 @@ During 1.6.5 cycle, 'next' will hold topics meant for 1.6.5 and 1.7.0.
20
--------------------------------------------------
21
[Graduated to "master"]
22
23
-* aj/fix-read-tree-from-scratch (2009-08-17) 1 commit
24
- (merged to 'next' on 2009-08-20 at 7a04133)
25
- + read-tree: Fix regression with creation of a new index file.
23
+* wl/insta-mongoose (2009-08-21) 1 commit
24
+ (merged to 'next' on 2009-08-25 at da1d566)
25
+ + Add support for the Mongoose web server.
26
27
-* jc/maint-checkout-index-to-prefix (2009-08-16) 1 commit
28
- (merged to 'next' on 2009-08-20 at 2f6aea2)
29
- + check_path(): allow symlinked directories to checkout-index --prefix
27
+* as/maint-graph-interesting-fix (2009-08-21) 2 commits.
28
+ (merged to 'next' on 2009-08-25 at 9d5e215)
29
+ + Add tests for rev-list --graph with options that simplify history
30
+ + graph API: fix bug in graph_is_interesting()
31
31
-* jl/submodule-summary-diff-files (2009-08-15) 2 commits
32
- (merged to 'next' on 2009-08-15 at 165bd8e)
33
- + Documentaqtion/git-submodule.txt: Typofix
34
- (merged to 'next' on 2009-08-14 at a702e78)
35
- + git submodule summary: add --files option
32
+* jc/maint-unpack-objects-strict (2009-08-13) 1 commit.
33
+ (merged to 'next' on 2009-08-23 at 38eb750)
34
+ + Fix "unpack-objects --strict"
35
37
-* lh/short-decorate (2009-08-15) 1 commit
38
- (merged to 'next' on 2009-08-18 at b8c1d96)
39
- + git-log: allow --decorate[=short|full]
36
+* jh/submodule-foreach (2009-08-20) 9 commits
37
+ (merged to 'next' on 2009-08-20 at 671bea4)
38
+ + git clone: Add --recursive to automatically checkout (nested) submodules
39
+ + t7407: Use 'rev-parse --short' rather than bash's substring expansion notation
40
+ (merged to 'next' on 2009-08-18 at f4a881d)
41
+ + git submodule status: Add --recursive to recurse into nested submodules
42
+ + git submodule update: Introduce --recursive to update nested submodules
43
+ + git submodule foreach: Add --recursive to recurse into nested submodules
44
+ + git submodule foreach: test access to submodule name as '$name'
45
+ + Add selftest for 'git submodule foreach'
46
+ + git submodule: Cleanup usage string and add option parsing to cmd_foreach()
47
+ + git submodule foreach: Provide access to submodule name, as '$name'
48
41
-* oa/stash-na (2009-08-11) 1 commit
42
- (merged to 'next' on 2009-08-14 at 12c2e2b)
43
- + git stash: Give friendlier errors when there is nothing to apply
44
---------------------------------------------------
45
-[New Topics]
49
+* lt/block-sha1 (2009-08-17) 4 commits
50
+ (merged to 'next' on 2009-08-18 at 67a1ce8)
51
+ + remove ARM and Mozilla SHA1 implementations
52
+ + block-sha1: guard gcc extensions with __GNUC__
53
+ + make sure byte swapping is optimal for git
54
+ + block-sha1: make the size member first in the context struct
55
56
* np/maint-1.6.3-deepen (2009-08-24) 1 commit.
57
(merged to 'next' on 2009-08-25 at 8e383d4)
@@ -52,19 +61,100 @@ During 1.6.5 cycle, 'next' will hold topics meant for 1.6.5 and 1.7.0.
61
(merged to 'next' on 2009-08-25 at 5f29625)
62
+ checkout: do not imply "-f" on unborn branches
63
55
-* mr/gitweb-snapshot (2009-08-25) 3 commits
56
- - gitweb: add t9501 tests for checking HTTP status codes
57
- - gitweb: split test suite into library and tests
58
- - gitweb: improve snapshot error handling
64
+* mm/reset-report (2009-08-21) 2 commits
65
+ (merged to 'next' on 2009-08-25 at f2a4424)
66
+ + reset: make the reminder output consistent with "checkout"
67
+ + Rename REFRESH_SAY_CHANGED to REFRESH_IN_PORCELAIN.
68
+
69
+* jc/shortstatus (2009-08-15) 11 commits
70
+ (merged to 'next' on 2009-08-15 at 7e40766)
71
+ + git commit --dry-run -v: show diff in color when asked
72
+ + Documentation/git-commit.txt: describe --dry-run
73
+ (merged to 'next' on 2009-08-12 at 53bda17)
74
+ + wt-status: collect untracked files in a separate "collect" phase
75
+ + Make git_status_config() file scope static to builtin-commit.c
76
+ + wt-status: move wt_status_colors[] into wt_status structure
77
+ + wt-status: move many global settings to wt_status structure
78
+ + commit: --dry-run
79
+ (merged to 'next' on 2009-08-06 at fe8cb94)
80
+ + status: show worktree status of conflicted paths separately
81
+ + wt-status.c: rework the way changes to the index and work tree are summarized
82
+ + diff-index: keep the original index intact
83
+ + diff-index: report unmerged new entries
84
+ (this branch is used by jc/1.7.0-status.)
85
+
86
+--------------------------------------------------
87
+[New Topics]
88
+
89
+* jc/upload-pack-hook (2009-08-28) 2 commits
90
+ - upload-pack: feed "kind [clone|fetch]" to post-upload-pack hook
91
+ - upload-pack: add a trigger for post-upload-pack hook
92
+
93
+I do not know if the distinction between fetching some but not all refs
94
+and fetching full set of refs into an empty repository is something worth
95
+making, so in that sense the tip commit is somewhat iffy.
96
+
97
+One reason this series makes me somewhat uneasy is that Tom, the original
98
+starter of the discussion went dark after sending a proposed patch. Maybe
99
+he has been too busy, but I have been hoping that GitHub as a stakeholder
100
+has somebody who monitors the list when he is not available.
101
+
102
+Does anybody from GitHub have any input? Is there something that needs to
103
+be improved to fill GitHub's needs? Does GitHub want to stick to its own
104
+fork, and were all these discussions for improvements unwanted?
105
+
106
+* jk/clone-b (2009-08-26) 1 commit
107
+ (merged to 'next' on 2009-08-30 at 10a68d1)
108
+ + clone: add --branch option to select a different HEAD
109
+
110
+* pk/import-dirs (2009-08-24) 1 commit
111
+ - Add script for importing bits-and-pieces to Git.
112
+
113
+This version makes me suspect that the author might regret the choice of
114
+the import format that does not allow escaping of paths, nor does not
115
+allow leading blanks for readability without changing semantics, both of
116
+which make it somewhat limiting and error prone. These issues will be
117
+hard to rectify without breaking the backward compatibility, for a tool
118
+that could otherwise turn out to be useful.
119
+
120
+As a contrib/ material, I probably shouldn't be too worried about these
121
+issues, but I am keeping this out of 'next' for now, just in case the
122
+author chooses to polish the usability of the tool for general audience.
123
+
124
+It is a different story if the submission was just throwing out a one-time
125
+hack in the open in the hope that some other people might find it useful,
126
+but without any intention of maintaining it. But then I do not have a
127
+strong reason to keep this in my tree, either. The mailing list archive
128
+is a more suitable storage media for such a patch.
129
+
130
--------------------------------------------------
131
[Stalled]
132
133
+* jh/notes (2009-08-27) 12 commits.
134
+ - Add '%N'-format for pretty-printing commit notes
135
+ - Add flags to get_commit_notes() to control the format of the note string
136
+ - notes.c: Implement simple memory pooling of leaf nodes
137
+ - Selftests verifying semantics when loading notes trees with various fanouts
138
+ - Teach the notes lookup code to parse notes trees with various fanout schemes
139
+ - t3302-notes-index-expensive: Speed up create_repo()
140
+ - fast-import: Add support for importing commit notes
141
+ - Teach "-m <msg>" and "-F <file>" to "git notes edit"
142
+ - Add an expensive test for git-notes
143
+ - Speed up git notes lookup
144
+ - Add a script to edit/inspect notes
145
+ - Introduce commit notes
146
+
147
+I heard the cvs-helper series depends on this one. It seems that the
148
+fan-out strategy is being rethought?
149
+
150
* js/stash-dwim (2009-07-27) 1 commit.
151
(merged to 'next' on 2009-08-16 at 67896c4)
152
+ Make 'git stash -k' a short form for 'git stash save --keep-index'
153
(this branch is used by tr/reset-checkout-patch.)
154
67
-* tr/reset-checkout-patch (2009-08-18) 8 commits.
155
+* tr/reset-checkout-patch (2009-08-27) 9 commits.
156
+ (merged to 'next' on 2009-08-27 at d314281)
157
+ + Make test case number unique
158
(merged to 'next' on 2009-08-18 at e465bb3)
159
+ tests: disable interactive hunk selection tests if perl is not available
160
(merged to 'next' on 2009-08-16 at 67896c4)
@@ -83,13 +173,6 @@ option "git stash --opt" to "git stash save --opt", to keep it flexible
173
and safe at the same time. I think it is a sane thing to do, but nothing
174
has happened lately.
175
86
-* jn/gitweb-blame (2009-08-06) 3 commits
87
- - gitweb: Create links leading to 'blame_incremental' using JavaScript
88
- - gitweb: Incremental blame (WIP)
89
- - gitweb: Add optional "time to generate page" info in footer
90
-
91
-Ajax-y blame WIP
92
-
176
* db/vcs-helper (2009-08-09) 17 commits
177
- Allow helpers to request marks for fast-import
178
- Allow helpers to report in "list" command that the ref is unchanged
@@ -113,140 +196,85 @@ Ajax-y blame WIP
196
+ Add support for external programs for handling native fetches
197
(this branch is used by jh/cvs-helper.)
198
116
-There was a discussion that suggests that the use of colon ':' before vcs
117
-helper name needs to be corrected. Nothing happened since.
199
+We had a few messages on what the list consensus was with this series. My
200
+impression, after going back to the archive, is that there wasn't.
201
119
-* je/send-email-no-subject (2009-08-05) 1 commit
120
- - send-email: confirm on empty mail subjects
202
+* jn/gitweb-blame (2009-08-06) 3 commits
203
+ - gitweb: Create links leading to 'blame_incremental' using JavaScript
204
+ - gitweb: Incremental blame (WIP)
205
+ - gitweb: Add optional "time to generate page" info in footer
206
+
207
+Ajax-y blame WIP
208
122
-This seems to break t9001. Near the tip of 'pu' I have a iffy
123
-workaround.
209
--------------------------------------------------
210
[Cooking]
211
127
-* mm/reset-report (2009-08-21) 2 commits
128
- (merged to 'next' on 2009-08-25 at f2a4424)
129
- + reset: make the reminder output consistent with "checkout"
130
- + Rename REFRESH_SAY_CHANGED to REFRESH_IN_PORCELAIN.
131
-
132
-* wl/insta-mongoose (2009-08-21) 1 commit
133
- (merged to 'next' on 2009-08-25 at da1d566)
134
- + Add support for the Mongoose web server.
135
-
136
-* lt/approxidate (2009-08-22) 2 commits
212
+* je/send-email-no-subject (2009-08-05) 1 commit
213
+ (merged to 'next' on 2009-08-30 at b6455c2)
214
+ + send-email: confirm on empty mail subjects
215
+
216
+The existing tests to covers the positive case (i.e. as long as the user
217
+says "yes" to the "do you really want to send this message that lacks
218
+subject", the message is sent) of this feature, but the feature itself
219
+needs its own test to verify the negative case (i.e. does it correctly
220
+stop if the user says "no"?)
221
+
222
+* lt/approxidate (2009-08-30) 6 commits
223
+ (merged to 'next' on 2009-08-30 at e016e3d)
224
+ + fix approxidate parsing of relative months and years
225
+ + tests: add date printing and parsing tests
226
+ + refactor test-date interface
227
+ + Add date formatting and parsing functions relative to a given time
228
(merged to 'next' on 2009-08-26 at 62853f9)
229
+ Further 'approxidate' improvements
230
+ Improve on 'approxidate'
231
232
Fixes a few "reasonably formatted but thus-far misparsed" date strings.
142
-As Nico suggested, we would need a test to prevent regression to existing
143
-support for date strings that are "reasonably formatted".
233
+With tests by Peff, this should be ready for -rc0.
234
145
-* jc/mailinfo-scissors (2009-08-25) 2 commits
146
- - Documentation: describe the scissors mark support of "git am"
147
- - Teach mailinfo to ignore everything before -- >8 -- mark
235
+* mr/gitweb-snapshot (2009-08-25) 3 commits
236
+ (merged to 'next' on 2009-08-30 at e4edd0b)
237
+ + gitweb: add t9501 tests for checking HTTP status codes
238
+ + gitweb: split test suite into library and tests
239
+ + gitweb: improve snapshot error handling
240
+
241
+* jc/mailinfo-scissors (2009-08-26) 5 commits
242
+ (merged to 'next' on 2009-08-30 at 5fc6248)
243
+ + mailinfo.scissors: new configuration
244
+ + am/mailinfo: Disable scissors processing by default
245
+ + Documentation: describe the scissors mark support of "git am"
246
+ + Teach mailinfo to ignore everything before -- >8 -- mark
247
+ + builtin-mailinfo.c: fix confusing internal API to mailinfo()
248
+
249
+I didn't pick up the patch to simplify the definition of scissors. I do
250
+not have strong opinion on it either way, but the list would hopefully
251
+decide it before too long.
252
253
* tf/diff-whitespace-incomplete-line (2009-08-23) 2 commits.
254
(merged to 'next' on 2009-08-26 at 4fc7784)
255
+ xutils: Fix xdl_recmatch() on incomplete lines
256
+ xutils: Fix hashing an incomplete line with whitespaces at the end
257
154
-* cc/sequencer-rebase-i (2009-08-21) 17 commits.
258
+Will merge.
259
+
260
+* cc/sequencer-rebase-i (2009-08-28) 15 commits
261
- rebase -i: use "git sequencer--helper --cherry-pick"
262
- sequencer: add "--cherry-pick" option to "git sequencer--helper"
157
- - sequencer: add "do_commit()" and related functions
263
+ - sequencer: add "do_commit()" and related functions working on "next_commit"
264
- pick: libify "pick_help_msg()"
159
- - revert: libify pick
265
+ - revert: libify cherry-pick and revert functionnality
266
- rebase -i: use "git sequencer--helper --fast-forward"
267
- sequencer: let "git sequencer--helper" callers set "allow_dirty"
268
- sequencer: add "--fast-forward" option to "git sequencer--helper"
269
- sequencer: add "do_fast_forward()" to perform a fast forward
270
- rebase -i: use "git sequencer--helper --reset-hard"
271
- sequencer: add "--reset-hard" option to "git sequencer--helper"
166
- - sequencer: add comments about reset_almost_hard()
272
- sequencer: add "reset_almost_hard()" and related functions
273
- rebase -i: use "git sequencer--helper --make-patch"
169
- - sequencer: free memory used in "make_patch" function
274
- sequencer: add "make_patch" function to save a patch
275
- sequencer: add "builtin-sequencer--helper.c"
276
173
-Migrating "rebase -i" bit by bit to C. I am inclined to agree with Dscho
174
-that maybe this approach forces the migration to follow the structure of
175
-the shell script too much, and could force a suboptimal end result, but
176
-we'll see.
177
-
178
-* as/maint-graph-interesting-fix (2009-08-21) 2 commits.
179
- (merged to 'next' on 2009-08-25 at 9d5e215)
180
- + Add tests for rev-list --graph with options that simplify history
181
- + graph API: fix bug in graph_is_interesting()
182
-
183
-Looked sane.
184
-
185
-* jc/shortstatus (2009-08-15) 11 commits
186
- (merged to 'next' on 2009-08-15 at 7e40766)
187
- + git commit --dry-run -v: show diff in color when asked
188
- + Documentation/git-commit.txt: describe --dry-run
189
- (merged to 'next' on 2009-08-12 at 53bda17)
190
- + wt-status: collect untracked files in a separate "collect" phase
191
- + Make git_status_config() file scope static to builtin-commit.c
192
- + wt-status: move wt_status_colors[] into wt_status structure
193
- + wt-status: move many global settings to wt_status structure
194
- + commit: --dry-run
195
- (merged to 'next' on 2009-08-06 at fe8cb94)
196
- + status: show worktree status of conflicted paths separately
197
- + wt-status.c: rework the way changes to the index and work tree are summarized
198
- + diff-index: keep the original index intact
199
- + diff-index: report unmerged new entries
200
- (this branch is used by jc/1.7.0-status.)
201
-
202
-Will cook for a bit more and then merge.
203
-
204
-* jc/maint-unpack-objects-strict (2009-08-13) 1 commit.
205
- (merged to 'next' on 2009-08-23 at 38eb750)
206
- + Fix "unpack-objects --strict"
207
-
208
-Will merge.
209
-
210
-* jh/submodule-foreach (2009-08-20) 9 commits
211
- (merged to 'next' on 2009-08-20 at 671bea4)
212
- + git clone: Add --recursive to automatically checkout (nested) submodules
213
- + t7407: Use 'rev-parse --short' rather than bash's substring expansion notation
214
- (merged to 'next' on 2009-08-18 at f4a881d)
215
- + git submodule status: Add --recursive to recurse into nested submodules
216
- + git submodule update: Introduce --recursive to update nested submodules
217
- + git submodule foreach: Add --recursive to recurse into nested submodules
218
- + git submodule foreach: test access to submodule name as '$name'
219
- + Add selftest for 'git submodule foreach'
220
- + git submodule: Cleanup usage string and add option parsing to cmd_foreach()
221
- + git submodule foreach: Provide access to submodule name, as '$name'
222
-
223
-Will merge.
224
-
225
-* jh/notes (2009-07-29) 8 commits.
226
- - t3302-notes-index-expensive: Speed up create_repo()
227
- - fast-import: Add support for importing commit notes
228
- - First draft of notes tree parser with support for fanout subtrees
229
- - Teach "-m <msg>" and "-F <file>" to "git notes edit"
230
- - Add an expensive test for git-notes
231
- - Speed up git notes lookup
232
- - Add a script to edit/inspect notes
233
- - Introduce commit notes
234
-
235
-The cvs-helper series depends on this one. I have a recollection
236
-that some people were not happy about the fan-out of the notes tree
237
-layout, but has the issue been resolved to a concensus?
238
-
239
-* ne/rev-cache (2009-08-21) 6 commits
240
- . support for path name caching in rev-cache
241
- . full integration of rev-cache into git, completed test suite
242
- . administrative functions for rev-cache, start of integration into git
243
- . support for non-commit object caching in rev-cache
244
- . basic revision cache system, no integration or features
245
- . man page and technical discussion for rev-cache
246
-
247
-Updated but seems to break upload-pack tests when merged to 'pu'; given
248
-what this series touches, breakages in that area are expected.
249
-May discard if a working reroll comes, to give it a fresh start.
277
+Migrating "rebase -i" bit by bit to C.
278
279
* jh/cvs-helper (2009-08-18) 7 commits
280
- More fixes to the git-remote-cvs installation procedure
@@ -258,31 +286,22 @@ May discard if a working reroll comes, to give it a fresh start.
286
- Basic build infrastructure for Python scripts
287
(this branch uses db/vcs-helper.)
288
261
-Builds on db/vcs-helper (which is stalled, so this cannot move).
262
-
263
-The testing of Python part seemed to be still fragile even with the latest
264
-fix on one of my boches with an earlier round already installed, but I
265
-didn't look very deeply before removing the older installation.
289
+Builds on db/vcs-helper (which is stalled, so this cannot move further at
290
+the moment). There is a re-roll planned, so I did not pick up test fixes
291
+from Brandon myself.
292
267
-* sr/gfi-options (2009-08-24) 4 commits
293
+* sr/gfi-options (2009-08-27) 6 commits
294
- fast-import: test the new option command
295
- fast-import: add option command
296
+ - fast-import: test the new feature command
297
+ - fast-import: add feature command
298
- fast-import: put marks reading in it's own function
271
- - fast-import: put option parsing code in seperate functions
272
-
273
-Will merge to 'next' shortly.
274
-
275
-* lt/block-sha1 (2009-08-17) 4 commits
276
- (merged to 'next' on 2009-08-18 at 67a1ce8)
277
- + remove ARM and Mozilla SHA1 implementations
278
- + block-sha1: guard gcc extensions with __GNUC__
279
- + make sure byte swapping is optimal for git
280
- + block-sha1: make the size member first in the context struct
299
+ - fast-import: put option parsing code in separate functions
300
282
-May merge soon; Solaris performance patches that was discussed
283
-earlier can happen on 'master', as the series is usable as-is.
301
+Re-rolled, based on an off-list discussion I was/am not aware of.
302
+Looked ready for 'next'.
303
285
-* nd/sparse (2009-08-20) 20 commits
304
+* nd/sparse (2009-08-20) 19 commits
305
- sparse checkout: inhibit empty worktree
306
- Add tests for sparse checkout
307
- read-tree: add --no-sparse-checkout to disable sparse checkout support
@@ -302,11 +321,7 @@ earlier can happen on 'master', as the series is usable as-is.
321
- Introduce "skip-worktree" bit in index, teach Git to get/set this bit
322
- Add test-index-version
323
- update-index: refactor mark_valid() in preparation for new options
305
- (merged to 'next' on 2009-08-20 at ea167d7)
306
- + Prevent diff machinery from examining assume-unchanged entries on worktree
324
308
-The first one was an independent fix; the rest has been replaced with the
309
-"return of no-checkout" series.
325
--------------------------------------------------
326
[For 1.7.0]
327
@@ -315,7 +330,6 @@ The first one was an independent fix; the rest has been replaced with the
330
+ git status: not "commit --dry-run" anymore
331
+ git stat -s: short status output
332
+ git stat: the beginning of "status that is not a dry-run of commit"
318
- (this branch uses jc/shortstatus.)
333
334
With this, "git status" is no longer "git commit --preview".
335
@@ -323,7 +337,11 @@ With this, "git status" is no longer "git commit --preview".
337
(merged to 'next' on 2009-08-22 at 5106de8)
338
+ send-email: make --no-chain-reply-to the default
339
326
-* jc/1.7.0-diff-whitespace-only-status (2009-05-23) 2 commits.
340
+* jc/1.7.0-diff-whitespace-only-status (2009-08-30) 4 commits.
341
+ (merged to 'next' on 2009-08-30 at 0623572)
342
+ + diff.c: fix typoes in comments
343
+ (merged to 'next' on 2009-08-27 at 81fb2bd)
344
+ + Make test case number unique
345
(merged to 'next' on 2009-08-02 at 9c08420)
346
+ diff: Rename QUIET internal option to QUICK
347
+ diff: change semantics of "ignore whitespace" options
@@ -336,6 +354,7 @@ we could argue that it is a bugfix.
354
(merged to 'next' on 2009-08-02 at 38b82fe)
355
+ Refuse deleting the current branch via push
356
+ Refuse updating the current branch in a non-bare repository via push
357
+
358
--------------------------------------------------
359
[I have been too busy to purge these]
360
@@ -374,3 +393,15 @@ Maybe some people care about this. I dunno.
393
. config: allow false and true values for branch.autosetuprebase
394
395
Has been ejected from 'pu' for some time, expecting a reroll.
396
+
397
+* ne/rev-cache (2009-08-21) 6 commits
398
+ . support for path name caching in rev-cache
399
+ . full integration of rev-cache into git, completed test suite
400
+ . administrative functions for rev-cache, start of integration into git
401
+ . support for non-commit object caching in rev-cache
402
+ . basic revision cache system, no integration or features
403
+ . man page and technical discussion for rev-cache
404
+
405
+Updated but seems to break upload-pack tests when merged to 'pu'; given
406
+what this series touches, breakages in that area are expected.
407
+May discard if a working reroll comes, to give it a fresh start.