[PATCH 1/2] comctl32/updown: Don't update buddy text if it's the same.

[Nikolay Sivov]

On 12/23/2009 23:50, Paul Vriens wrote:
> On 12/23/2009 07:55 PM, Nikolay Sivov wrote:
>> On 12/23/2009 21:49, Paul Vriens wrote:
>>> On 12/22/2009 10:43 PM, Nikolay Sivov wrote:
>>>> Don't update buddy text if it's the same.
>>>>
>>>> There's at least one application that depends on this behavior:
>>>> http://bugs.winehq.org/show_bug.cgi?id=18574
>>>>
>>>> This bug is closed cause crash no longer occurs in 1.1.35 due
>>>> a regression it contains, the real problem is fixed by this patch.
>>>
>>> Hi Nikolay,
>>>
>>> This one introduces a test failure on NT4 but I guess it's due to
>>> comctl32 4.72 and not the OS.
>>>
>>> NT4/4.72 sends a WM_SETTEXT instead of WM_GETTEXT (just tested, it's
>>> the only message).
>>>
>>> Could you have a look? The NT4 boxes at winetestbot.geldorp.nl have
>>> this issue.
>>>
>> Hi.
>>
>> I already saw it, we should just skip it on < 5.8. I don't have NT4 box,
>> could you please test application from
>> http://bugs.winehq.org/show_bug.cgi?id=18574 on it?
>>
>> I expect it to crash, if it doesn't that is not good.
>
> Yup on my NT4 box with just the mfc42.dll from the vc6redist it 
> crashes. That box has comctl32 4.70.
>
> On that same box with an updated comctl32 (running 50comupd.exe that 
> gives us comctl32 5.80) it doesn't crash.
>
Great, as expected. Not sure how to fix it properly in case of updown. 
Adding 'optional' entry is not acceptable here,
so we should somehow get that running on 4.7 and skip ok_sequence() 
call. Can you recommend something? I see only
UDM_SETPOS32 skipped now.