Re: [TRY 4] user32/tests: Add tests for STM_SETIMAGE and STM_SETICONmessage processing

[Dmitry Timoshkov]

"Ilya Shpigor"  wrote:
> - flush_sequence();
> - SendMessageW(hwnd, STM_SETIMAGE, (WPARAM)IMAGE_BITMAP, (LPARAM)hbm);
> - todo_wine ok_sequence(WmSetImageErrStaticSeq, "STM_SETIMAGE on a static", FALSE);
> -
> - flush_sequence();
> - SendMessageW(hwnd, STM_SETIMAGE, (WPARAM)IMAGE_ICON, (LPARAM)hicon);
> - todo_wine ok_sequence(WmSetImageErrStaticSeq, "STM_SETIMAGE on a static", FALSE);
> -
> - flush_sequence();
> - SendMessageW(hwnd, STM_SETICON, (WPARAM)hicon, 0);
> - todo_wine ok_sequence(WmSetIconErrStaticSeq, "STM_SETICON on a static", FALSE);
> -
> - flush_sequence();
> - SendMessageW(hwnd, STM_SETIMAGE, (WPARAM)IMAGE_ENHMETAFILE, (LPARAM)hemf);
> - todo_wine ok_sequence(WmSetImageErrStaticSeq, "STM_SETIMAGE on a static", FALSE);

You should use SendMessageA instead of SendMessageW to make the test run
under Win9x as well. Testing the return value of SendMessage would be also
helpful.

[TRY 4] user32/tests: Add tests for STM_SETIMAGE and STM_SETICONmessage processing

[Ilya Shpigor]

On Tuesday 29 September 2009 10:57:45 you wrote:
> You should use SendMessageA instead of SendMessageW to make the test run
> under Win9x as well. Testing the return value of SendMessage would be also
> helpful.

Hi,

The TRY 5 of tests had beent sent. Is it correct or my test so bad that nobody 
doesn't like to answer? :)

[TRY 4] user32/tests: Add tests for STM_SETIMAGE and STM_SETICONmessage processing

[Dmitry Timoshkov]

"Ilya Shpigor"  wrote:
> The TRY 5 of tests had beent sent. Is it correct or my test so bad that nobody 
> doesn't like to answer? :)

Try 5 looks good to me (therefore the lack of the comment), not sure
why it wasn't committed.

[TRY 4] user32/tests: Add tests for STM_SETIMAGE and STM_SETICONmessage processing

[Alexandre Julliard]

"Dmitry Timoshkov"  writes:
> "Ilya Shpigor"  wrote:
> > The TRY 5 of tests had beent sent. Is it correct or my test so bad
> > that nobody doesn't like to answer? :)
>
> Try 5 looks good to me (therefore the lack of the comment), not sure
> why it wasn't committed.

It's still using W functions, it's leaking an hdc, and it cannot
possibly pass on Wine. Not to mention that it would be more convincing
if it showed some successful sequences as well.