Questions after testtools merge

[Martin (gzlist)]

On 30/12/2009, Martin (gzlist) <gzlist at googlemail.com> wrote:
>
> Haven't dug into this further yet, but I suspect that bazaar is
> corrupting the environment block when doing the codepage-switching
> attempts. Why testtools does this when the previous testing code
> didn't I have no idea.

Okay, so, the behaviour change seems to be that testtools doesn't call
tearDown if TestSkipped is raised. I don't know if it's a deliberate
difference, but that's what caused the regression. Well,
bb.test_non_ascii is dodgy anyway, but if tearDown is not called then
osutils._cached_user_encoding is left as some bogus value that then
gets used to trample over various things later on. Setting
os.environ['BZR_EMAIL'] prior to the skip check is a worry as well
though.

Martin