Re: [PATCH 2/2] user32/tests: Test that sending WM_CANCELMODE cancels tracking in window's menu bar.

[Dmitry Timoshkov]

Ken Thomases <ken(a)codeweavers.com> wrote:
> dlls/user32/tests/menu.c |   21 ++++++++++++++++++++-

Please add a message test instead, it would much more clearer show what is
going on.
> +    {MSG msg;   while (PeekMessage(&msg, 0, 0, 0, PM_REMOVE)) DispatchMessage(&msg);}

Please avoid creating such embedded blocks.

-- 
Dmitry.

[PATCH 2/2] user32/tests: Test that sending WM_CANCELMODE cancels tracking in window's menu bar.

[Ken Thomases]

On May 17, 2013, at 12:21 AM, Dmitry Timoshkov wrote:
> Ken Thomases <ken(a)codeweavers.com> wrote:
> > dlls/user32/tests/menu.c |   21 ++++++++++++++++++++-
>
> Please add a message test instead, it would much more clearer show what is
> going on.

I added to the existing tests of WM_CANCELMODE because that's what this new test is.  It's just like the existing tests except with a different way of initiating menu tracking.
> > +    {MSG msg;   while (PeekMessage(&msg, 0, 0, 0, PM_REMOVE)) DispatchMessage(&msg);}
>
> Please avoid creating such embedded blocks.

Again, that's just directly copied from the earlier code in the same test function.  I suppose I could break that up, if you prefer.

-Ken

[PATCH 2/2] user32/tests: Test that sending WM_CANCELMODE cancels tracking in window's menu bar.

[Dmitry Timoshkov]

Ken Thomases <ken(a)codeweavers.com> wrote:
> > Please add a message test instead, it would much more clearer show what is
> > going on.
>
> I added to the existing tests of WM_CANCELMODE because that's what this new
> test is.  It's just like the existing tests except with a different way of
> initiating menu tracking.

You will need to add a message test anyway if you plan to fix this properly,
your current test doesn't allow to see what its going on exactly and what is
needed to be fixed to make the test pass.

-- 
Dmitry.