Brendan Cully <brendan@kublai.com> [Thu, 09 Nov 2006 14:44:17 -0800] rev 3629
zsh: typo in commit completion
Thomas Arendsen Hein <thomas@intevation.de> [Thu, 09 Nov 2006 21:53:45 +0100] rev 3628
issue228: Fix repositories at the filesystem root (/ or C:\)
Thanks to Robert Shaw and other people on the list for the suggestions.
Brendan Cully <brendan@kublai.com> [Wed, 08 Nov 2006 11:28:33 -0800] rev 3627
Export changelog dict in localrepo.commit
Matt Mackall <mpm@selenic.com> [Wed, 08 Nov 2006 13:20:08 -0600] rev 3626
tests: add -i switch
Prompt to accept changes for tests where output changed
Matt Mackall <mpm@selenic.com> [Wed, 08 Nov 2006 13:20:08 -0600] rev 3625
tests: add -R switch
Restarts at first .err file if one exists, otherwise runs all tests
Matt Mackall <mpm@selenic.com> [Wed, 08 Nov 2006 13:20:08 -0600] rev 3624
tests: sort test list if running all tests
Thomas Arendsen Hein <thomas@intevation.de> [Tue, 07 Nov 2006 22:35:07 +0100] rev 3623
Fix accessing a repository via -R/--repository through a symlink.
Sometimes the repository root was compared to os.getcwd(), which always uses
the canonical path without symbolic links in it.
This would changes self.root of the localrepo objects to always use os.sep as
the directory separator, which is implicitly assumed in some places, but may
not be the case if somebody uses -R foo/repo on windows.
Thomas Arendsen Hein <thomas@intevation.de> [Tue, 07 Nov 2006 22:18:34 +0100] rev 3622
Merge with upstream
Matt Mackall <mpm@selenic.com> [Tue, 07 Nov 2006 14:04:15 -0600] rev 3621
Fix up parent display in logging
For changesets with a first parent nullid, but not rev 0, we ought to
display a null parent.
Matt Mackall <mpm@selenic.com> [Wed, 01 Nov 2006 21:47:55 -0600] rev 3620
Clarify untracked file merge message