Re: PHIL: Performance Baseline for VRML-capable Machines

The Bagel-Meistro (cbentzel@media.mit.edu)
Wed, 15 Jun 94 10:54:13 -0400


Mark Pesce writes:
> I believe that a 486/25 represents a "low-end" baseline, for the following
> reasons:
First three reasons......

> D) Such a machine can sustain a rendering speed of approximately 25,000
> Gouraud-shaded or texture-mapped polygons, per second.

To make this language even more device independent, or at least
more low-end accessible while still providing the same power,
could there perhaps be a way to have the host end determine how
much detail they need, for example, allowing filled polygons, but
not texture-mapped ones? This could be a language independent
choice, such as Mosaic allowing one to not show any inlined images
to speed up the time to access documents. This would seem to
provide the best of both worlds: on Dec Alpha's running it on
leased T1 lines the graphics could be superb while the 386-33 user
running a 14.4 could experience something similar.
Bentzel