In general, I believe, sending a compressed, colormapped bitmap
(GIF) rendered by the server will actually be faster than
transmitting all the rendering primitives (X) to the client. The
more complex the scene, the more GIF wins.
On the other hand, while downloading full geometry (VRML) to the
client has similar cost to transmitting rendering primitives (X),
that cost is only for the first frame, after which VRML wins big.
So even aside from portability issues, I don't see where X would fit
in.
P.S. For those interested, Jan's remark above refers to Cyberview
<URL:http://www.geom.umn.edu/apps/cyberview3d/about.html>.
--------------------------------------------------------------------
Paul Burchard <[email protected]>
``I'm still learning how to count backwards from infinity...''
--------------------------------------------------------------------