Re: ElevationGrid

[Rikk Carey <rikk@wasabisoft.com>]

At 02:03 AM 2/13/98 -0500, Heiko Grussbach wrote:
>Hi,
>
>recently there is much interest in the ElevationGrid node, so IŽd like to
>ask a question to browser writers (or others who know). I assume, that
>rendering an EG is done by breaking it down into triangles. So for an EG of
>100*100 height values one would have almost 20000 triangles. Is that
>correct?

yes.


>...... And what happens if, in the extreme, all heights are equal, still
>20000 triangles or only 2 which would suffice for the geometry?

This is an implementation detail/variance.  The spec does not prevent them
from adaptively generating the grid and thus reduce the poly count.
However, I'll bet that all VRML browsers today simply generate an NxM grid.
It may be possible to create a PROTO that adaptively constructs itself
into a set of primitives (EG's or IFS's) based on some tolerances. Seams
will be a little tricky to handle properly.

I'm sure that the geo sciences users would benefit greatly from something
like this...


--
Rikk Carey   rikk@wasabisoft.com    http://www.wasabisoft.com/users/rikk
(650)969-5910            Wasabi Software, Inc: http://www.wasabisoft.com
          The Annotated VRML Ref. Manual: http://www.wasabisoft.com/Book

-------------------------------------------------------------------------
for list subscription instructions,
send email to www-vrml-request@vrml.org with text "info"