Wine and locales

[Shachar Shemesh]

Filip Navara wrote:

> Shachar Shemesh wrote:
>
>> What I don't get is - how can depending on the thread/user locale 
>> ever fail?
>
>
> Maybe I haven't expressed myself clearly. What I mean is that it does 
> the comparsion of the strings first with the thread locale and if that 
> returns negative result (string s not equal)  then it tries to compare 
> the strings again using the system locale.

If that's the case, that's frightening. If the strings are not the same 
according to user locale (thread locale, more precisely), I *want* them 
to be reported different. So what if under another locale they may be 
the same?

>> Only if you have the order off the top of your head, and it is NOT 
>> the one outlined in "find_entry" (dlls/ntdll/resource.c). I.e. - if 
>> you think the one in find_entry is ok, don't waste your time on 
>> writing test cases.
>
>
> I think "find_entry" is almost correct. Only honouring the the default 
> UI lanaguage is missing (as introduced in W2K).

Yes, that's what I'm trying to work on here.

> I think it belongs between step 3. and 4. and that the user default UI 
> language (GetUserDefaultUILanguage) is tried first and system default 
> UI language (GetSystemDefaultUILanguage) just after that.

A. I don't think we are going to have both. No point, and no place to 
initialize it from. I may reconsider when Wine has a multi-user 
installation option.
B. The docs seems to suggest that the thread locale doesn't affect 
resource loading at all any more. Is that a bug in the docs? I was 
considering simply replacing everything that says "thread locale" with 
"UI language".

> Regards,
> Filip

       Shachar

-- 
Shachar Shemesh
Lingnu Open Source Consulting ltd.
http://www.lingnu.com/