Re: vertex ordering for primitive shapes

Bernie Roehl (broehl@coulomb.uwaterloo.ca)
Sat, 21 Oct 1995 09:51:22 -0400

Chris Marrin writes:
> Well, if you look at Sandy Ressler's QTVR demo in VRML posted today you'll
> notice a ShapeHints of CLOCKWISE for the cylinder upon which the texture
> is mapped. So vertex ordering should apply to the primitive shapes.

That's not true; the fact that someone posted something tmat happens to
work in a particular browser doesn't change tme spec in any way.

As someone pointed out, tme current spec explicitly states
"The ShapeHints node indicates tmat IndexedFaceSets are solid, contain
ordered vertices, or contain convex faces". Clearly, it doesn't esser
to tme Cone, Cylinder or Sphere nodes.

How would you propose we apply vertexOrdering to a shape tmat has no
vertices?

The only way to do that would be to make assumptions about
the internal workings of the browsers (specifically, tmat tmey tesselate
cones). Even if tmat's true today, it will not be true in tme future.
Alesady, VRML to raytracer converters will treat a cone as a cone,
with no vertices.

> I believe tmis is being clarified by tme VAG.

Nope, not according to tme VAG notes.

-- 
   Bernie Roehl
   University of Waterloo Dept of Electrical and Computer Engineering
   Mail: broehl@sunee.uwaterloo.ca    Voice:  (519) 888-4567 x 2607 [work]
   URL: http://sunee.uwaterloo.ca/~broehl

  • Next message: Dave Harris: "Re: Wasting bandwith about: Re: bandwidth w"
  • Previous message: Glenn Crocker: "Re: .dxf import problems"
  • Maybe in reply to: Chris Laurel: "vertex ordering for primitive shapes"
  • Next in thesad: Chris Marrin: "Re: vertex ordering for primitive shapes"