From the nimble fingers of Robert Covington (artlythere@...)
(4/5/2001 2:20 AM) came...

> 
>> The program below seems to meet your specification, with some changes to
>> make the sum of weights = 1.0. I don't think it will be much good in a
>> Gouraud shading context, though. You will end up with colour
>> discontinuities between adjacent triangles.
>> 
>> Robert P.
> 
> <big snip>
> 
> Nice demo. Thank you.
> 
> You are correct, as  a test I did proved.
> I think this is due to my misstatement of the problem and incorrect theory
> behind it.

It seems this ground has been trod before, but the nub of it remains: it's a
3D problem, and all that can be determined from a 2D concept is one (of an
infinite number) projection of that 3D thing onto 2D space.

The three colours must originate (at black) from one vertex of a cube - of
unit dimension, if you like -, and pure colours will travel along the three
edges common to that point. Any colour (vector) will be specified by the
vector sum of the three prime values.

Or not...
-- 
Ted Spencer; ted@...
--
Size isn't everything: the whale is endangered while the ant continues to do
just fine.