Code review comment for lp:~vila/bzr/822571-bzr-home-unicode

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 8/8/2011 11:15 AM, Vincent Ladeuil wrote:
> Vincent Ladeuil has proposed merging
> lp:~vila/bzr/822571-bzr-home-unicode into lp:bzr.
>
> Requested reviews: bzr-core (bzr-core) Related bugs: Bug #822571 in
> Bazaar:
> "bzrlib.tests.blackbox.test_version.TestVersionUnicodeOutput.test_unicode_bzr_home
>
>
fails" https://bugs.launchpad.net/bzr/+bug/822571
>
> For more details, see:
> https://code.launchpad.net/~vila/bzr/822571-bzr-home-unicode/+merge/70706
>
>
>
Ad-hoc fix for the failing test.
>
> It's a bit surprising that we failed this way, I don't fully
> understand what triggered the failure overall.
>
> The actual failure is related to Martin's addition of the fdatasync
> config option which requires getting the path to the config file and
> as such interprets BZR_HOME. It may also be related to Jelmer's
> change about separating get_transport for paths and urls.
>
> There may be a deeper issue at work here but not worth investigating
> now IMHO, just something to stay aware about.

Should we be forcing "utf-8" rather than _fs_enc? I would tend to do

1) _fs_enc if set (not None) and not ASCII

2) fallback to 'mbcs' on Windows

3) fallback to 'utf-8' everywhere else

John
=:->

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (Cygwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAk4/4EAACgkQJdeBCYSNAANjXwCfQ4Z4kFfCFKfqPS09/AlFmc/Z
CgMAoMmxtnj4I/NUb9dJMGMUPu+13zXH
=fg9D
-----END PGP SIGNATURE-----