Rob's respons on NURBS test

["Cruikshank, David W" <david.w.cruikshank@boeing.com>]

All,

Here is Rob's response to the NURBS "circle" test.

"Yes, I think the circle representation of NURBS01.cgm is correct.  The
NURBS literature that I've read states that a NURBS can exactly
represent a circle (something a NUBS cannot do), and I believe that is
what Dick produced.
I've been diverted (momentarily I hope) onto another project, so I did
not go through the degree elevation exercise myself.  However, Dick
certainly seems to be a NURBS expert, and the NURBS that he produced
closely resembles circular NURBS that you can find in textbooks on
NURBS.  So, I think it is unlikely that he made a mistake.  This is all
somewhat counter-intuitive to me, as I expected a more noticeable
difference in the shape of the curve with a 20% difference in weights.
My initial attempt at evaluating the shape of the curve with an open
source NURBS library did produce a very noticeable difference in the
shape of the curve.  At some point I'll go back and try to figure out
why.

This all leads to an observation/question.  Why is it that in all of the
time we've been soliciting files for the NURBS test suite, the only
NURBS example (with weights not equal to 1.0) was created by hand?  Are
these things actually produced by any CAD systems in common use?"

Dave
Technical Fellow - Graphics/Digital Data Interchange
Boeing Commercial Airplane
206.544.3560, fax 206.662.3734  <-- NEW NUMBERS
david.w.cruikshank@boeing.com