My only problem with that is that with only two points used, I can think
of at least THREE possible circles (i.e., with centers at different
vectors); two points could lie somewhere along either one of two of the
component semicircles, or the two points could be at either end of a
diameter of a circle with a different diameter.  Am I clear?
What I'm thinking is that given THREE coplanar points, only ONE possible
(coplanar) circle could encompass all three points.   Maybe I'm wrong
about there being only one such circle?

--Mark

Ben Chambers wrote:
> 
> Actually, you can do it with two points. Try this:
> 
> #declare P1 = <x1, y1, z1>;
> #declare P2 = <x2, y2, z2>;
> 
> sphere {
>     (P1+P2)/2, sqrt(sqr(x2-x1)+sqr(y2-y1)+sqr(z2-z1))
>     texture {stuff}
> }

Post a reply to this message