Re: f0dder's Fabulous Wait States.

hutch-- wrote:
> With the same level of assumption about unproven truisms, feel free to
> assume whatever you like but do fix that procedure you posted as an
> example, I can nitpick lousy code in my sleep and until you zero fill
> that structure and put the correct length into it, your code is
> dangerous and has no garrantee that it will run on a later version.
> Then put the structure witin the procedure as a local so it is both
> re-entrant and free of cache problems.

As has already been mentioned a multitude of times, the version in
Donkey's library doesn't have any of these flaws. Considering that
you whine when people don't use the very latest up-to-the-minute service
pack of masm32, you should shut your trap :)


> Its a case of picking the tree out of your own eye before you try and
> invent a splinter in mine. If you know so little about testing, at
> least don't burden me with your mistakes but do feel free to create
> the non-polling version of a message loop and a WndProc with
> precanned high level wait state API function calls.

Guess what? GetMessage is a blocking API. B-l-o-c-k-i-n-g, spelled out
for you. This means not polling for a message, but being woken up once
a message is ready.


.