I'd think interrupts would be ideal...

by (Login Mikrondel)
R

"interrupts have crummy timing anyway (a horizontal scan takes tens of microseconds, you might wait that long just to get the interrupt handler, especially if more priviledged I/O is going on)"

What are you talking about? Interrupts have great timing. BANG and your handler is running. Even if more privileged I/O is going on, *a polling routine isn't going to be running anyway*. An interrupt handler has the advantage of being able to wait even after the trigger has gone away.


...buuuut as far as I know there is no way for the horizontal retrace to trigger an interrupt.

Posted on Jul 22, 2008, 4:06 AM
from IP address 220.245.178.132

Respond to this message   

Goto Forum Home


Response TitleAuthor and Date
Reference (URL) on Jul 22
 And? on Jul 22
  * In particular was the latency part, tens of microseconds. on Jul 22
   And you expect interpreted BASIC to compete with that? on Jul 22
    Re: And you expect interpreted BASIC to compete with that? on Jul 23
     No, I don't on Jul 23