On Sun, 5 Aug 2001, Alexander R Angas wrote:
> > For the moment, I think sound is functional enough that I'm concentrating
> > on fixing the major win32 bugs (save games, fullscreen and scaling
> > corruption) first before I work on any new features (like a win32 native
> > soundserver).
>
> Yeah sure absolutely. I think I might add little trace markers to the save
> games where those purify errors are coming up so I can compare them between
> Win32 and Linux.
>
> Thanks for explaining the timing stuff though. :)
You're welcome. It might verify my observations further if you add a check
where the timer calculations are done and make sure currentime - lasttime
are always > 0. If not, print out an error. It would be interesting to see
if that error is printed when the game starts to go goofy.
If that does indeed happen, we could add a hack for win32 to sleep for
1 ms, then recalculate the currenttime - lasttime. (BTW, it's prbably not
literally currenttime - lasttime in the code; I have no way to look at it
right now.)
--
http://www.clock.org/~matt
