> Well, I may have completely misunderstood the problem here, but how about > examining the frame counter after each RST28 call to see whether one or > more interrupts occurred and then forcing a task switch immediately if > one did? I presume that if you wanted the pointer to keep moving then > you would have no problems in attaching a "move the pointer but don't > switch tasks" subroutine to the frame interrupt vector... > > imc
Ahhh... good idea. I'll have to see how effective it is, but it sounds fair enough. Cheers, Steve.
