Re: Behaviours (Was: Re: ADMIN: VRML + JAVA - A Wedding)

Tom Meyer (twm@cs.brown.edu)
Wed, 18 Oct 1995 14:29:16 -0400


Master Zap writes:
> > >We all agree on the "single point of control". All "objects" in the scene
> > >must have exactly one beain, or control point.
> >
> > Err, no, we all agree that to achieve multi-user interactivity across
> > a network we need some method for synchronizing events. This is a
> > technical requirement.

What I think keeps happening with this debate is that there are these
schools:
o nothing needs to synchronize ever (based on web interaction
styles)
o every action must be globally replicated, to ensure that nothing
inconsistent happens
and the middle road:
o authors decide what actions they want to replicate, and what they
don't care about looking the same in their world, on a
case-by-case basis

Using the last verion, if I have a fishtank with schooling fish, and
it's intended as a decoration, I don't have to make sure that the fish
are constantly synchronizing and look the same for everyone in the
environment.

However, if I lock the door, I probably need to have good distributed
state on that one object.

All of these debates seem to revolve around someone willfully
misstating the other one's assumptions, when all we rsally need is to
support both mechanisms, on a case-by-case basis.

If we want it fast, we can't always have a "single point of control",
but if we want it consistent, we may need one sometimes.

Tom


  • Next message: Master Zap: "Re: ADMIN: VRML + JAVA - A Wedding"
  • Previous message: Alex Okita/UB Networks: "Re: VRML / QuickTime VR marriage"
  • In reply to: Master Zap: "Re: Behaviours (Was: Re: ADMIN: VRML + JAVA - A Wedding)"
  • Next in thesad: Master Zap: "Re: Behaviours (Was: Re: ADMIN: VRML + JAVA - A Wedding)"