TriStrips, one last time.

Randy Stiles (
Wed, 7 Dec 94 14:48:33 PST

I'd like to address something that has been bugging me for months with
VRML. Some of the ultimate plans people have for VRML are a little
misguided in my opinion. The HTTP protocol is completely wrong for
developing multi-user, immersive VRML. If we ever wanted to go that way,
we would need to rethink the infrastructure that VRML sits upon. I would
also like to point out that there are already several corporate efforts
looking into providing this functionality.

It is useful for providing caches of scenes, etc.
Agreed, the actual motion updates of
participants/objects should be handled
differently. Earlier discussion mentioned DIS,
and that could be incorporated complementary to
the http portion later.

As for corporate efforts, corporations tend to
keep things to themselves because they think
exclusivity gives them an edge. It is usually
unintended that they let something like Unix slip
out and get used and improved by a larger
community. Some changes are underway, and a good
sign is that SGI is willing to let the Inventor
format be used as a starting point. But just
because corporate efforts are underway does not
guarantee it will get done. Corporations are
fickle at times.


# Randy Stiles Lockheed AI Center
# Orgn 9620 Bldg 254F
# Office: 415.354.5256 3251 Hanover Street
# Fax: 415.354.5235 Palo Alto, CA 94304-1191