user32: After handling an internal message give a chance to real message to arrive.

[Dmitry Timoshkov]

Alexandre Julliard <julliard at winehq.org> wrote:

> > I understand that the patch that removes WM_WINE_SETWINDOWPOS message
> > probably needs more tests, it's already in the pending state, I'm asking
> > about the tests that are waiting in the queue. First thing I'm trying to
> > do is understand where exactly the source of the problem is, and the tests
> > I sent exersize the suspected things one by one.
> 
> Sure, that's a good thing to do, but it doesn't mean that all these
> intermediate tests need to end up in the tree, or that they need to be
> committed before you can make progress. Just take your time to
> understand that thing properly, then write tests demonstrating the
> actual problem, and submit them along with the solution.

That's a usual thing with hard to diagnose problems like race conditions
which is the case here: find something suspicious -> write a test -> try
a fix -> find that the fix still not enough -> go to start. In the end
there might be a fix somewhere in a different place, but in the way there's
a good chance that other problems could be found and fixed, that wouldn't
happen otherwise. If you believe that the tests by itself prove nothing or
don't deserve to be accepted on their own that's highly discouraging. Is
that how Wine development works these days? Probably I should add a local
hack and be done with it then.

-- 
Dmitry.