Re: [Update REQ 5390]: Hello (spherical convex hulls)

[rgeorge at hsv.tybrin.com]

Stuart,

Thank you for your suggestions. I have a few questions though.

Suggestion #1
If all of the points were cospherical (which they are in general),  
mapping the vertices of the resulting 3D solid back to the sphere  
could produce the original set of points. Am I missing something  
here?

Suggestion #2
I don't actually have a good reference point. I could use the first  
data point as my reference and proceed, but I cannot guarantee that  
all points will be within 90 degrees of that or any reference  
point. I will be processing thousands, tens of thousands, hundreds  
of thousands and perhaps millions of points. Any method which  
required transforming each point should be as cheap as possible.

Any response or suggestions are greatly appreciated.

Russell.