From: c s <rb_dev_at_yahoo.com>
Date: Tue, 9 Dec 2003 02:25:27 -0800 (PST)

--- Hardeep Sidhu <hardeeps_at_pobox.com> wrote:

> An easy way to reproduce it is to do a quick ff/rew
> while the mp3 file
> is being loaded from disk. Your change guarantees
> that the mpeg thread
> gets one extra iteration before the wps updates but
> it doesn't guarantee
> that the ff/rew will occur during that iteration.
> The solution would be
> to not update the wps until the ff/rew completes in
> the mpeg thread but
> that's kinda messy and probably not worth it for
> such a minor issue.


Thanks. I was able to reproduce it as you described,
although you really have to watch carefully to catch
it as the jump is relativly small for that case.

I agree with you that that special case may not be
worth bothering trying to fix. The original glitch was
probably not really an issue that concerned many
people in the first place but the fix was so trivial
that I figured that there was no reason not to do it.

If only one mpeg_queue event gets processed during the
yield(), I suppose a quick and dirty fix would be to
do a series of yields before updating the wps, which
would catch more cases, but that still wouldn't
gurantee that we would always have processed the
MPEG_FF_REWIND queue event before we updated the wps.

=====
Craig

rb_dev_at_yahoo.com

__________________________________
Do you Yahoo!?
New Yahoo! Photos - easier uploading and sharing.
http://photos.yahoo.com/
Received on 2003-12-09