On Wed, Aug 08, 2001 at 11:05:50AM -0400, Erik Mooney wrote:
> >Well, there are a lot of tricks you can do to make it less sensitive to
> >those issues, although none of them perfect, obviously. For example,
>
> And it has to be absolutely perfect for peer-peer parallel emulation to work. If
> one paddle read is off by one scanline, the player misses the ball on one
> machine and not the other; wham, the two emulators are in divergent states.
> Then:
Obviously you have to send all the state involved in a paddle read, not
just that it occured.
> You just described client-server emulation in fact, with client-side prediction.
> One machine is updating the other after-the-fact when necessary. You can't
> have peer-peer with both machines updating the other, because each machine
> would be updating the other with divergent results.
State synchronization is not equivalent to what you described, which was
a system where only one machine actually emulates the game. Nor is this
necessarily client-server, because both sides are performing this prediction.
Neither side considers itself the authority on anything except the specific
input registers its user may change.
In any case, I think this discussion has gone far enough out of the
realm of relevance.
Cheers.
--
-/
|/| Julian Squires <tek@xxxxxxx>
/-
-
Archives (includes files) at http://www.biglist.com/lists/stella/archives/
Unsub & more at http://www.biglist.com/lists/stella/
