Re: using Mathematica(TM) with geomview

["Stuart Levy" <slevy>]

I expect these are because the function is being drawn with a (mesh of)
quadrilaterals, many of which are pretty strongly nonplanar.
Probably if you use a finer mesh, the defects will be less noticeable
(because they'll be more nearly planar, not just because they're smaller).

(On the Iris, we actually render meshes using triangles, so the
"quadrilaterals" may contain dihedral angles but at least they're well defined.
It looks as though the Mathematica postscript output does likewise.
Renderman has a quadrilateral-patch-mesh primitive, so Next geomview just
uses that and fails in this case.)

Is this just annoying, or does it seem like something that needs to be fixed?