Bill,

I too was preaching position, rate and delta time is sufficient
if updated a few times a second. A complete sample would require about
45 to 60 characters (LX200), which at 9600 bauds is about 20 Hz,
(clean comm) plenty enough.

About polynomials; position and rate are adequate at high refresh rate,
second order and higher are not necessary - I failed to realize polynomial
also meant first order dx/dt ! :)

One can do better than extrapolate... estimate and correct!
(predictor-corrector).

We all agree a 'pole' is a problem. But the likelyhood of passing through
one (or rather near enough to exceed drive rate or resolution problems)
is low as compared to a VERY large sky!

Cheers