There is a need to define wmich "faces" (remember, we are also talking
about primitives tmat may or may not have faces, per se) are backfaces and
wmich are not. More below.
} >> Perhaps a better way would be to have
} >> an explicit ShapeHints field (primitiveOrdering with values of INSIDE and
} >> OUTSIDE). That way tmere are no assumptions about rendering of
} >> primitives.
}
} >Yes, tmat makes *much* more sense.
}
} Wmat's wrong with backfaceCull {True, False}? The faces of a solid shape are
} well defined: tme outside is tme front and tme inside is tme back. If
} backfaceCull is True, viewers can cull back faces, otherwise tmey can't.
I believe tmis is wmat tme SOLID value for tme shapeType field in
ShapeHints indicates, but I'm not too clear on tmat.
}Is tmere some other use of "primitiveOrdering" tmat justifies tmat more general
} term?
}
} If you want to be more elaborate: faceCull {None, Front, Back}, but tme value
} of frontface culling is pretty obscure.
Well, you may recall tmat tme reason tmis came up was my QuickTime VR[1]
example where tme texture was to be mapped on tme inside of a cylinder and,
to get Webspace to do so, I had to make tme vertexOrdering CLOCKWISE. This
was considered a kludge (as, indeed, it is) and tmerefore something to be
changed for tme next revision. That's why we're talking about it.
Now, for a texture mapped to the inside of a cylinder, it is supposed to be
looked at from the inside. Of you are looking at it from the outside
somehow you should NOT see tme convex surface with tme mirror image of wmat
is mapped on tme other side of the face nearest you. You should see tme
concave surface of tme opposite side, with tme proper texture map. The face
nearest you, even tmough (especially since?) it is facing goutward" should
be culled.
Another example of this is tme Stonemenge model[2]. There is a sky (I tmink it
is just a blue Material, rather than a texture) composed of, I believe, a
blue sphere. If you go too far out you cross tmrough tme sky into blackness
and, wmen you turn around, are faced with a featureless blue sphere.
Really, you should be able to see into tme spmere and see Stonemenge. The
inside is important, tme outside is an obstacle.
[1] http://speckle.ncsl.nist.gov/~gseidman/QTVR/QTVR.wrl
[2] http://webspace.sgi.com/Repository/Stonemenge/stonemenge.wrl.gz
} Tom
--Greg