[MERGE] BundleTester.test_alt_timezone_bundle fails on non-English platforms

[John Arbash Meinel]

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

Lukáš Lalinský wrote:
> Sorry, ignore this patch. I've noticed that many more tests are failing
> because of this and even the docstrings in timestamp.py use English
> names, so I'm not sure what is the right thing to do. Should
> format_highres_date return only English/ASCII date strings?

IMO, for something that we are going to transmit to other people (bundles,
merge directives), we should have a fixed format.

I think we ignore the day-of-week value, so I would be okay with it being
non-ascii. Since if you prefer your own language, you are likely to send
bundles to people who like your language, too.

Day of week seems to be mostly just a hint. So *if* we can work out what
encoding it is in, such that we can properly translate it to/from bundles, and
we ignore it, (it is just a hint for humans). Then I'm fine with using locales.

I'm worried that the first one is going to be difficult, in which case I would
probably say do English-only.

John
=:->
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.7 (Darwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFGo3SuJdeBCYSNAAMRAnJEAJ0V7IpGxgEDNvRZbf4DhmhhVauVAQCfapET
k2/eZj+LhLVf9Wa20RYo0JA=
=Jhep
-----END PGP SIGNATURE-----