Re: [PATCH 2/2] gdi32/tests: Fixed broken tests on non English system Part 2.

2012-10-07 Thread Dmitry Timoshkov 
Qian Hong  wrote:

> Any feedback for this patch? Thanks a lot for any comments!

I'd try to avoid extra EnumFontFamiliesEx call, and probably do the checks
you need in the previous one.

-- 
Dmitry.

Re: [PATCH 2/2] gdi32/tests: Fixed broken tests on non English system Part 2.

2012-10-07 Thread Qian Hong 
Hello Dmitry,

Any feedback for this patch? Thanks a lot for any comments!

On Thu, Oct 4, 2012 at 3:02 PM, Qian Hong  wrote:
> Dmitry, could you review this patch? Thanks a lot ;-)
>
> This series of patch fix some broken tests on Japanese Windows and Chinese 
> Windows:


-- 
Regards,
Qian Hong

-
Sent from Ubuntu
http://www.ubuntu.com/

Re: ole32: Fix dwClsContext parameter of a CoCreateInstance call in DefaultHandler_Run.

2012-10-07 Thread Nikolay Sivov 

On 10/6/2012 19:29, Roman Dadkov wrote:
This patch change dwClsContext parameter of a CoCreateInstance call in 
function DefaultHandler_Run.
Because even if there is some clsid in the registry, the function will 
not be able to run the newly created object.



-  hr = CoCreateInstance(&This->clsid, NULL, CLSCTX_LOCAL_SERVER,
+  hr = CoCreateInstance(&This->clsid, NULL,
+CLSCTX_INPROC_SERVER|CLSCTX_INPROC_HANDLER|
+CLSCTX_LOCAL_SERVER|CLSCTX_REMOTE_SERVER,
  &IID_IOleObject, (void **)&This->pOleDelegate);

This could be replaced with CLSCTX_ALL.

Is it possible to add a test for that?

Re: iphlpapi/tests: Skip tests if functions are not supported

2012-10-07 Thread Dmitry Timoshkov 
André Hentschel  wrote:

> For Mac, BSD, Solaris

If it's Wine specific, then probably there shouild be a todo_wine somewhere
to make it clear that under Windows there is no skips.

-- 
Dmitry.

Updated Wine Wiki Repo

2012-10-07 Thread Kyle Auble 
Just wanted to let everyone know I've created a
git repo for the Wine Wiki from the CVS one at
SourceForge. You can find it at:
https://bitbucket.org/kauble/wine-wiki-migration

Since it's a fresh repo, I thought editing the
history should be ok (just moved the right info
into Author & Committer fields and made the commit
message formatting more consistent). I do still
have the unedited branches from cvsimport and can
put those up too if there's any demand for them.

I've been talking with Dimi about what we would
need to upgrade MoinMoin, and once he can find
some time, he said he would setup a regular
archive of the Wine Wiki's data. I also like the
idea of using tags to mark which wiki version is
in use, but I'm not sure how many times the wiki
has been upgraded. If anyone knows what versions
of MoinMoin we've used and when the upgrades took
place, I can add the extra tags.

I don't know if there's much interest in just
moving this to the Wine git repo yet. If not, I
can probably review patches, but would be happy to
add someone else as an admin if they have a
bitbucket account.

- Kyle

Re: user32: Fix the edit class extra byte count on non-x86 platforms (try 2)

2012-10-07 Thread André Hentschel 
Am 04.10.2012 10:40, schrieb Alexandre Julliard:
> André Hentschel  writes:
> 
>> @@ -5173,10 +5173,10 @@ const struct builtin_class_descr EDIT_builtin_class =
>>  editW,/* name */
>>  CS_DBLCLKS | CS_PARENTDC,   /* style */
>>  WINPROC_EDIT, /* proc */
>> -#ifdef __i386__
>> -sizeof(EDITSTATE *) + sizeof(WORD), /* extra */
>> -#else
>> +#ifdef _WIN64
>>  sizeof(EDITSTATE *),  /* extra */
>> +#else
>> +sizeof(EDITSTATE *) + sizeof(WORD), /* extra */
>>  #endif
> 
> Why do you need that?
> 

Sry, seen this too late.
I need it for user32/tests/edit.c: test_extra_values()
It expects the size of the edit class like this:

#ifdef _WIN64
{"Edit",8,8},
#else
{"Edit",6,8},
#endif

So this fails on e.g. ARM because code and test use different conditions.

-- 

Best Regards, André Hentschel