On Wed, Apr 12, 2006 at 05:50:40PM +0100, Malcolm Wallace wrote:
The argument John was making is that this is a useful distinguishing point to tell whether your concurrent implementation is cooperative or preemptive. My argument is that, even if you can distinguish them in this way, it is not a useful distinction to make. Your program is simply wrong. If you have a sequential program whose value is _|_, your program is bad. If you execute it in parallel with other programs, that does not make it any less bad. One scheduler reveals the wrongness by hanging, another hides the wrongness by letting other things happen. So what? It would be perverse to say that the preemptive scheduler is semantically "better" in this situation.
Oh, I didn't mean it was necessarily a useful quality to the end programmer, I was actually just trying to make the point you were making that such programs are incorrect and getting the non-termination case over with. So we can get to the fairness discussion without adding caveats like "if no thread is in an infinite loop". But I didn't want to just say "assuming your program is correct" without giving some indication of what that actually means for a program to be correct. In any case, it is something we can point to and say "this! this is a difference!" whether it is a useful one or not. now for the contrived counter-example :) start two threads, one trying to prove goldbachs conjecture, the other trying to find a refutation. in a preemptive system this will terminate*, in a cooperative system it may not. John * insert goedel incompleteness caveat. -- John Meacham - ⑆repetae.net⑆john⑈