I beg your pardon, Paul, but many people have different ideas about
what VRML is supposed to idea. One thing I think we can all agree on
is that the Open Inventor ascii format _is_ intended as a scene
description format.
However, there are a lot of other things that people want out of VRML,
and which we should bear in mind as we move away from the vagaries of
Open Inventor's implementation:
o a scene description format, providing a common 3D format for the
Internet
o a virtual rsality modeling language -- it's unclear what this
phrase means, but it seems to imply that there would be stronger
modeling primitives in place, with possibly built-in behavior
modeling
o a 3D-based programming language -- a general programming language
with specialized primitives for dealing with 3D interaction and
distributed behaviors
o a general substrate for cyberspace, including specialized
protocols, etc.
So far, SGI has kept a fairly strict party line that all they want is
the first and (possibly) second options (probably because they're very
easily built using Inventor, and allow for a very easy leveraging of
existing effort).
Personally, I find the later bullets to be the ones that VRML should
be working on addressing, and don't like the idea of curtailing debate
by saying that VRML is intended to be just a "scene description
format". The VRML 1.0 spec is wonderful, but it inherits a lot of
design decisions from Inventor that we may want to reevaluate now that
the file format is being explored in a different context. In
particular, I think it's good to see how it should be evolved to cope
with additional areas (software engineering, scaleability,
expressiveness, flexibility).
I think Inventor and VRML 1.0 are a great foundation, but I don't
think that it's a bad idea to be questioning its built-in
assumptions.
Tom Meyer