[Fwd: Re: dlls/winmm/tests/capture.c - yes, we can open a device in 2MHz...]

[Jakob Eriksson]

Robert Reif wrote:

> How about starting a blacklist framework.  You could have
> a table with the function that fails and some caps info that
> identifies the driver.  I would suspect that there is more
> than one buggy driver out there.  You could then just add
> it's info to the table when it is found.
>
> You could then do:
>
> if (!inBlacklist(WAVE_IN_OPEN, &capsA)) {
>    // do the test
>    ...
> } else
>    trace("blacklisted ..... \n");
>
> There are modems that do strange things and the Remote
> Desktop Protocol driver and other third party sound
> rerouting pseudo drivers also do strange things.  These
> will need to be added someday also.
>
> This could become a real maintenance nightmare if not
> done properly from the start.



Yes, I too thought about creating a blacklist framework.  But I want
to wait a bit before starting on it, until I see:

a) how common buggy drivers are - the framework must not be
   heavier than the issues it solves.

b) with some more issues at hand, it's easier to see the
   patterns and what a blacklist framework should look like.
   I don't like coding without a clear vision of what I want.


regards,
Jakob