Greg Ward <greg-hg@gerg.ca> [Tue, 09 Mar 2010 21:53:16 -0500] rev 10636
rebase: improve help text
- add a paragraph about default dest/source changesets
- option help: say "changeset" not "revision"
- option help: explain -b/--base better
- clarify that -a/--abort and -c/--continue are different from the
other options
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Wed, 10 Mar 2010 22:27:42 +0100] rev 10635
serve: fix port config
Thomas Arendsen Hein <thomas@intevation.de> [Wed, 10 Mar 2010 22:07:38 +0100] rev 10634
merge with other head, will show bug in hg serve
Thomas Arendsen Hein <thomas@intevation.de> [Wed, 10 Mar 2010 22:05:41 +0100] rev 10633
test-serve: Show if port config and option are correctly used
this will break in
d3f27d15c9cb
Thomas Arendsen Hein <thomas@intevation.de> [Wed, 10 Mar 2010 21:35:38 +0100] rev 10632
drop (default: 8000), non-zero default is automatically added to help
Thomas Arendsen Hein <thomas@intevation.de> [Wed, 10 Mar 2010 21:21:15 +0100] rev 10631
make expression shorter, now the line fits into 80 chars
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Wed, 10 Mar 2010 21:14:24 +0100] rev 10630
serve: fix options recording, trailing whitespace
Bryan O'Sullivan <bos@serpentine.com> [Wed, 10 Mar 2010 10:51:37 -0800] rev 10629
serve: allow --port=0 to specify "server chooses the port number"
Sune Foldager <cryo@cyanite.org> [Wed, 10 Mar 2010 12:38:33 +0100] rev 10628
rebase: improve output of hg pull --rebase (
issue2072)
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Wed, 10 Mar 2010 09:52:16 +0100] rev 10627
shrink-revlog: factor out postorder algorithm
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Wed, 10 Mar 2010 09:51:07 +0100] rev 10626
shrink-revlog: factor out suboptimal computation
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Wed, 10 Mar 2010 09:48:15 +0100] rev 10625
shrink-revlog: remove branchsort algorithm (it behaves poorly)
Greg Ward <greg-hg@gerg.ca> [Tue, 09 Mar 2010 21:25:37 -0500] rev 10624
shrink-revlog: add accounting of suboptimal nodes to the new algorithms.
Greg Ward <greg-hg@gerg.ca> [Tue, 09 Mar 2010 21:22:01 -0500] rev 10623
shrink-revlog: add "reverse postorder" and "postorder reverse" toposorts.
Based on a patch by Benoit Boissinot, adapted to the pluggable sort
algorithm design. toposort_reversepostorder() is a very good
performer; it's designed to recreate what the revlog would have looked
like if Mercurial had parent deltas now. toposort_postorderreverse()
is unstable and very inconsistent, but perhaps with some work it could
be made better.
Greg Ward <greg-hg@gerg.ca> [Thu, 25 Feb 2010 09:16:39 -0500] rev 10622
shrink-revlog: add --sort option for user-selectable toposort algorithm.
Greg Ward <greg-hg@gerg.ca> [Tue, 09 Mar 2010 21:30:19 -0500] rev 10621
shrink-revlog: rename some local variables for consistency.
Greg Ward <greg-hg@gerg.ca> [Tue, 09 Mar 2010 21:13:39 -0500] rev 10620
shrink-revlog: instrument sort code to record statistics.
Notably, count "suboptimal" nodes (predecessor is not first parent)
and (with -v) report them at the end of the run.
Patrick Mezard <pmezard@gmail.com> [Tue, 09 Mar 2010 22:18:37 +0100] rev 10619
Merge with crew-stable
Patrick Mezard <pmezard@gmail.com> [Tue, 09 Mar 2010 22:11:43 +0100] rev 10618
convert: handle svn tree with empty roots (
issue2079)
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Tue, 09 Mar 2010 20:47:35 +0100] rev 10617
merge with stable
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Tue, 09 Mar 2010 20:38:23 +0100] rev 10616
bundle: fix bundle generation for empty changegroup
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Tue, 09 Mar 2010 19:04:18 +0100] rev 10615
patch/diff: move diff related code next to each other
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Tue, 09 Mar 2010 18:31:57 +0100] rev 10614
remove header handling out of mdiff.bunidiff, rename it
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Tue, 09 Mar 2010 18:27:02 +0100] rev 10613
remove contrib/hgdiff, it makes refactoring diff/patch harder
it can always be revived if needed
Dirkjan Ochtman <dirkjan@ochtman.nl> [Tue, 09 Mar 2010 16:03:57 +0100] rev 10612
demandimport: blacklist _ssl (
issue1964)
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Mon, 08 Mar 2010 19:43:24 +0100] rev 10611
patch/diff: move patch.export() to cmdutil.export()
This API change will allow us to break a cycle between patch and cmdutil.
Steve Borho <steve@borho.org> [Mon, 08 Mar 2010 11:12:48 -0600] rev 10610
wix: pre-check the 'I approve this license' checkbox
This is simply a convenience to the user, they can click next until the
installer goes away. Optimally, the checkbox would be hidden or removed,
but that requires rewriting the UI.
Matt Mackall <mpm@selenic.com> [Sun, 07 Mar 2010 22:04:35 -0600] rev 10609
Merge with stable
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Mon, 08 Mar 2010 00:01:12 +0100] rev 10608
addremove: avoid fetching data again and again
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Mon, 08 Mar 2010 00:00:03 +0100] rev 10607
addremove: add progress support for similarity search
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Sun, 07 Mar 2010 23:59:28 +0100] rev 10606
addremove: atomically update the dirstate
In case someone hits ^C during the similarity search, properly rollback
Greg Ward <greg-hg@gerg.ca> [Sun, 07 Mar 2010 13:10:23 -0500] rev 10605
inotify: expose the same dirstate.status() interface as dirstate.
The real dirstate.status() does not have any keyword args. For
consistency, wrappers or subclasses should not have them either.
Christian Ebert <blacktrash@gmx.net> [Sun, 07 Mar 2010 13:03:03 +0100] rev 10604
keyword: remove spurious locks, improve handling of wlock
- kwcommitctx is inside the wlock of repo.commit: no lock
- _kwfwrite only needs wlock
wlock outside try block, so we don't need to import lock.release
_kwfwrite should even be safer that way, as we moved the status
call inside the try-except block.
Thanks to Benoit Boissinot for help.
Christian Ebert <blacktrash@gmx.net> [Sun, 07 Mar 2010 13:03:03 +0100] rev 10603
keyword: use status' names instead of index
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Sun, 07 Mar 2010 17:44:43 +0100] rev 10602
hgweb/graph: edge should be same color as the destination
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Sat, 06 Mar 2010 13:29:54 +0100] rev 10601
templatefilters: store the agescales in reverseorder directly
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Sat, 06 Mar 2010 12:45:14 +0100] rev 10600
hgweb: separate generation of entries and sorting (cleanup)
Brendan Cully <brendan@kublai.com> [Fri, 05 Mar 2010 18:08:23 -0800] rev 10599
Merge stable
Greg Ward <greg-hg@gerg.ca> [Wed, 20 Jan 2010 20:23:36 -0500] rev 10598
demandimport: add __main__ to the blacklist (because of setuptools)
Paul Molodowitch <pm@stanfordalumni.org> [Thu, 17 Dec 2009 15:53:28 -0800] rev 10597
bookmarks: add invalidate() to bookmark_repo
Cached bookmark information would not be reset when repo.invalidate()
was called, meaning (for instanced) guis would be left with out-of-date
bookmark info. This patch fixes that by overriding bookmark_repo.invalidate()
Yuya Nishihara <yuya@tcha.org> [Tue, 12 Jan 2010 22:49:10 +0900] rev 10596
branch: help misuse of hg branch to switch branches
Maybe beginners tend to use 'hg branch EXISTING_BRANCH' instead of
'hg update EXISTING_BRANCH' as me.
If so, let's help them to use 'update'.
Also removed '(use --force to override)' because it can mislead them.
Lee Cantey <lcantey@gmail.com> [Sat, 06 Mar 2010 21:29:02 -0800] rev 10595
Make Readme.html reference current OS X release
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Sat, 06 Mar 2010 13:32:14 +0100] rev 10594
progress: clear progress before writing to stderr
Matt Mackall <mpm@selenic.com> [Fri, 05 Mar 2010 18:26:13 -0600] rev 10593
Added signature for changeset
ff2704a8ded3
Matt Mackall <mpm@selenic.com> [Fri, 05 Mar 2010 18:26:12 -0600] rev 10592
Added tag 1.5 for changeset
ff2704a8ded3
Matt Mackall <mpm@selenic.com> [Fri, 05 Mar 2010 17:24:52 -0600] rev 10591
mq: drop -Q in favor of --mq only
Sune Foldager <cryo@cyanite.org> [Fri, 05 Mar 2010 12:27:23 +0100] rev 10590
mq: correct test output
Sune Foldager <cryo@cyanite.org> [Fri, 05 Mar 2010 10:49:56 +0100] rev 10589
mq: disallow : in patch names (
issue2065)
Sune Foldager <cryo@cyanite.org> [Thu, 04 Mar 2010 14:23:54 +0100] rev 10588
mq: disallow # (and : on Windows) in patch names (
issue2065)
Sune Foldager <cryo@cyanite.org> [Thu, 04 Mar 2010 13:11:29 +0100] rev 10587
Merge with stable
Sune Foldager <cryo@cyanite.org> [Thu, 04 Mar 2010 13:10:48 +0100] rev 10586
win32/hgwebdir_wsgi: clarify documentation and clean up script a bit
Martin Geisler <mg@lazybytes.net> [Thu, 04 Mar 2010 10:05:12 +0100] rev 10585
Merge with hg-i18n
Jens Bäckman <jens.backman@gmail.com> [Thu, 04 Mar 2010 07:34:22 +0100] rev 10584
i18n-sv: translated bisect, dates, locate, manifest
FUJIWARA Katsunori <foozy@lares.dti.ne.jp> [Thu, 04 Mar 2010 11:56:53 +0900] rev 10583
i18n-ja: synchronized with
c50a3d7154d2
Wagner Bruna <wbruna@softwareexpress.com.br> [Wed, 03 Mar 2010 18:46:22 -0300] rev 10582
i18n-pt_BR: synchronized with
f142fa3c0a8c
Steve Borho <steve@borho.org> [Wed, 03 Mar 2010 17:37:50 -0600] rev 10581
test-fetch: use relative URL to avoid echoing TMP path
Sune Foldager <cryo@cyanite.org> [Wed, 03 Mar 2010 22:32:51 +0100] rev 10580
fetch: fix newheads calculation (
issue2047)
Thomas Arendsen Hein <thomas@intevation.de> [Wed, 03 Mar 2010 22:04:59 +0100] rev 10579
Make annotate --follow an alias for -f/--file to behave like in older versions
Since
98a0421b9e52 annotate follows copies/renames by default, but the output
of e.g. "annotate --follow --number" should not change without some
deprecation time.
Martin Geisler <mg@lazybytes.net> [Wed, 03 Mar 2010 20:35:20 +0100] rev 10578
win32/hgwebdir_wsgi: clarify copyright license
Adrian Buehlmann <adrian@cadifra.com> [Wed, 03 Mar 2010 14:50:35 +0100] rev 10577
store: only add new entries to the fncache file
Newly added fncache entries were not added to the in-memory cache,
making it possible for 'hg convert' to cause duplicates in
.hg/store/fncache.
Duplicates in the fncache file are harmless, but excessive numbers
of duplicates from large converted repositories may slow down
execution speed considerably.