Wine and locales

[Dmitry Timoshkov]

"Alexandre Julliard" <julliard at winehq.org> wrote:

> > There is no need to test every API, but there are APIs we need to require
> > mandatory testing:
> >
> > GetSystemDefaultLCID
> > GetUserDefaultLCID
> > GetACP
> 
> I don't see what you want to test here.

I want to test whether locale has really changed and what subsequences happened.
I want to see system/user locale before the switch and after. I'd like also add
GetSystemDefaultLangID, GetSystemDefaultUILanguage, GetUserDefaultLangID and
GetUserDefaultUILanguage to the list. Probably also GetThreadLocale.

> > FindResourceA
> > FindResourceExA
> 
> These ones have been well tested already.

I don't see the tests for them in the CVS, especially part about LANGID
preference order.

> > CompareStringA
> > GetLocaleInfoA
> 
> These take an explicit locale parameter so I don't see why changing
> the system locale would suddenly break them.

You are right.

> > Right, that's why I'm asking to revert patch and commit it simultaneously
> > with a test case and possible fixes for bugs revealed by the tests.
> 
> I see no reason to revert the patch. It's going to make no difference
> for 99% of our users, and for the remaining 1% it's about as likely to
> fix things as it is to break them.

OK, in the worst case perhaps I will need to send a custom version
of kernel32 with the patch reverted to the affected cxoffice users.

-- 
Dmitry.