Re: PHIL: What will VRML be used for?

Anthony Parisi (
Tue, 14 Jun 1994 12:55:29 -0700 (PDT)


I feel that you're exactly right on all three points.

> I think that everyone involved in VRML should have a very clear idea of what we
> are trying to achieve. From my understanding, there are two uses of VRML:
> 1) To provide a new, exciting interface to static WWW pages. The user will step
> into a room, where they can read a book to retrieve textual information, look
> at pictures on the wall (e.g. for GIFs), or watch a TV set (e.g. for MPEGs).
> Hyperlinks will be in the form of doors/stairs/wormholes whatever.

This seems intrinsically useful; a much more friendly environment for
exploring than what Mosaic provides.

> 2) To provide a standard language with which users can interact with each other
> and objects, where the users are in the same virtual space, but not
> nesscessarily the same physical space.

Absolutely. I think many of us understand the potential for adding this
sort of interactivity to the Web.

> To me, 1 is the most immediately achievable goal, but 2 is certainly the 'next
> generation' and should be considered in the design of VRML. Any comments? What
> should we be aiming for?

Read Mark Pesce's GOALS document. I'll pitch in my two cents: I think
the initial goal should be to arrive at a workable solution for #1 soon--
within the next few months. This is eminently doable and, frankly, it's
not rocket science. There are several standards and popular technologies
out there for doing 3d graphics which we can use a starting point. Mark
has posted an RFI for these.

Goal #2 is going to take time. I have already seen a few discussion
threads that touch on important pieces of the problem-- how are "objects"
uniquely identified on the Net? What's the system for specifying object
taxonomy, i.e. classes? How do you specify behavior? How does a network
architecture affect the design of inter-object communication protocols?

Many of the above questions *must* be addressed by standards bodies. Some
of them already are-- I'm sure by now you've seen postings on the Object
Management Group's CORBA (Common Object Request Broker Architecture).
My intuition tels me that the way to go here is to bring OMG and the Web
community together to form joint standards regarding object management and
the Web. I'd be interested in seeing comments on this from the list as a

Some of the above questions threaten to start religious wars. The truly
scary one is: what language do I use to specify behavior? I personally
think that if we try to pick one, we are doomed from the start. We
should take an open approach. In any case, there's no way that question
will be answered in a few months.

My suggestion for a strategy is: solve #1 now, and in so doing don't do
anything that prevents the design of a practical, open, standards-based
solution for #2.

tony parisi "In cyberspace, no one can hear you type"

tony parisi		"In cyberspace, no one can hear you type"