I would rather put the burden on the authors than on the viewers.
Frame rate is of the utmost importance when viewing VRML scenes. Anything
that detracts from the frame rate buck must be balanced by an equivalent
amount of presentation bang, or it's not worth it.
The trouble is that we can't know how much slower the "ultra-dynamic"
model is until we try it. So it's hard to judge whether or not it's
worth it. I have a feeling that this entire debate is based on differing
views on how much slower things would be if you couldn't optimize for
performance. Gavin, an acknowledged 3D programming expert, is strongly
recommending that we err on the side of optimization. I'm more inclined
towards his point of view on this issue because of his background, just
as I'm more inclined to your point of view, Mitra, when you talk about
networking issues.
James
-- James Waldrop / Technical Directorsulam@construct.net / Construct Internet Designsulam@well.com /http://www.construct.net