On Tue, Jun 22, 2010 at 11:10:44PM +0200, Gianluca Montecchi wrote: > On Tuesday 22 June 2010 22:48:29 Chris Ball wrote: > > It appears that Ubuntu Lucid is still shipping 2.1 with the broken > > locking, so that's problematic. Maybe we can file a bug and get them > > to backport whatever the fix was, if they aren't willing to move to a > > 2.2 beta? > > Also Debian squeeze ship bzr 2.1.2, even if you can get a 2.2b2 version in > experimental.
Beta-verisions are problematic, because of Bzr.version_cmp. It's obvious that 2.12 > 2.5, but is 2.2 >? 2.2r4 > 2.2b2 or whatever are less clear. Currently Bzr.version_cmp just bails, see below. If someone has a robust way to deal with these things, let me know (or just fix Bzr.version_cmp). > Anyway I get some errors with mercurial (because I have not configured it, I > suppose) and some other errors. > ... > ====================================================================== > ERROR: Children list should be revision dependent. > ---------------------------------------------------------------------- > Traceback (most recent call last): > ... > NotImplementedError: Cannot parse non-integer portion "2b2" of Bzr version > "2.2b2" This is the version parsing problem I mentioned above. > ====================================================================== > ERROR: Get should be able to return the previous version. > ---------------------------------------------------------------------- > Traceback (most recent call last): > ... > NotImplementedError: Cannot parse non-integer portion "2b2" of Bzr version > "2.2b2" And here it is again. > ====================================================================== > FAIL: Changed lists should reflect past activity > ---------------------------------------------------------------------- > Traceback (most recent call last): > File "/home/gian/Devel/be/libbe/storage/base.py", line 1024, in test_changed > 'Unexpected new: %s' % new) > File "/home/gian/Devel/be/libbe/storage/base.py", line 575, in failUnless > '(%s) %s' % (self._classname(), msg) > AssertionError: (Hg:1.5.2) Unexpected new: ['modified', 'moved2', 'new', > 'version'] For some reason, your hg is interpreting the modified files: 'modified', 'version' as new files. In fact, it shouldn't be picking up 'version' at all, since that file is only used internally by VCS and its subclasses. > File "/home/gian/Devel/be/libbe/util/subproc.py", line 96, in > libbe.util.subproc.Pipe > Failed example: > p.status > Expected: > 1 > Got: > 0 > ---------------------------------------------------------------------- > File "/home/gian/Devel/be/libbe/util/subproc.py", line 98, in > libbe.util.subproc.Pipe > Failed example: > p.statuses > Expected: > [1, 0] > Got: > [0, 0] > ---------------------------------------------------------------------- > File "/home/gian/Devel/be/libbe/util/subproc.py", line 100, in > libbe.util.subproc.Pipe > Failed example: > p.stderrs # doctest: +ELLIPSIS > Expected: > [...find: ...: Permission denied..., ''] > Got: > ['', ''] All of these are because you have execute permissions on every directory in /etc/. That is moderately concerning ;). Anyhow, perhaps we should replace this test with some sort of python call, to be more OS/distro agnostic. Any ideas? I'm not sure about the other failures yet... -- This email may be signed or encrypted with GPG (http://www.gnupg.org). The GPG signature (if present) will be attached as 'signature.asc'. For more information, see http://en.wikipedia.org/wiki/Pretty_Good_Privacy My public key is at http://www.physics.drexel.edu/~wking/pubkey.txt
pgp6edsK8mySy.pgp
Description: PGP signature
_______________________________________________ Be-devel mailing list [email protected] http://void.printf.net/cgi-bin/mailman/listinfo/be-devel
