On Thu, Sep 26, 2002 at 02:31:04PM -0700, Dan Hollis wrote:
> On Thu, 26 Sep 2002, Steve Freeland wrote:
> > On Thu, Sep 26, 2002 at 01:33:49PM -0700, Dan Hollis wrote:
> > Not necessarily. If the protocol is designed correctly, it can
> > guard against desync as effectively as a client-master system. Did you
> > think the desync you were getting with NeoGeo games was due to a flaw
> > in the protocol? I'm 99% certain that the netmame 0.5 protocol will
> > *not* allow desync (input desync at least, behaviour not depedant on
> > input is of course a whole other story) irrespective of network
> > conditions.
>
> actually we got lots of total xmame lockups with the netmame 0.5 protocol,
> but with the new one we wrote we don't get *any*.
>
Hmmm. Well, I don't know what would have caused that, but I'm
willing to bet it was an implementation bug rather than a flaw in the
protocol design. I still think it would be a mistake to go back to a
client/server protocol, but I guess we'll have to agree to disagree on
this.
Ah well, nothing like having something to prove to get the code
flowing!
--
Steve Freeland
[EMAIL PROTECTED]
_______________________________________________
Xmame mailing list
[EMAIL PROTECTED]
http://toybox.twisted.org.uk/mailman/listinfo/xmame
