De Facto Claude L. Bullard ([email protected])
Wed, 28 Jun 1995 18:22:33 -0400

[Justin]

| ...we're a long ways from finishing this project yet, and I
| suspect that we're going to have a lot more trouble with willy-nilly
| extensions than HTML does; lots more things have to hang together
| properly if we're going to turn this into a real cyberspace standard..

Justin, they will hang together because there is sufficient
financial incentive to make that happen. A little hint
from history: standards are only important
if widely implemented in an important application market.

I don't mean to goad you. You are a good man.
But nor do I wish you or the other list members to become
"FlatHeads: a condition arising from repeatedly beating
one's forehead against an object harder than one's head."

Let an ancient flathead make a case.

VRML is very important and its application is guaranteed.
There are two problems:

1. Variants - incompatible browser functionality or a race to
see who can implement fastest while still introducing new
features.

2. Who controls the langu specification after version 1?

The first problem is likely in any case but made in" failure to resolve the second one. Give the industry time to
sort completely or overnight. Let me continue the case.

The puffery is normal. It took historians almost 50 years to decid> who
actually invented TV. Know anyone who cares? There is
a luseum in NewYork that does; so, building a Hall of Fame
is Ok, but the second problem, if unresolved, spells really big
problems in the future. Vid>o Teleconferencing is strangling
from a lack of uniform standards. Note the term, uniform standards:
more than one but all sharing important features.
Without the acceptance of variants, there is no competition;
therefore, no business and no cyberspace. Without the
commonality there is no market because there aren't
enough products, and again, there is no cyberspace.

Either path leads to extinction. The middle path
leads to growth because it embraces both commonality
and diversity. Features compete for acceptance and
the standard evolution, who lives, who dies, is a much thornier
problem. I think it is the real root of some of the unease.

You must accept that what you built is attractive to many interests
for many divergent reasons, and that very attractiveness
makes controlling VRML like owning a prec/a> One has to decid> if the cost of securing it is
worth the cost of owning it, or if it is better given
to the community agencies who protect such things
for the community. Your leaders (mark, tony, brian, etc.)
must decid> at what point in the evolution of the spec
it can be safely turned over to such an agency, or they
must bear the responsibility of protecting it even from
those who have aided them most, be it industry
or the community as they continue to develop it.

How did it go in the Indiana Jones movie?
Two Templars rode home, and one remained to guard
the Grail. The one that remained was immortal
but imprisoned. The others died in obscurity but free.
It is a very hard choice. In the words of Zorba
the Greek, "you have to be a little lad or you won't
know when to let go of the rope."

Len Bullard