here's where i am but you're just going to tell me i'm wrong without
letting me understand it. you don't like my interest in finding what's
possible
# the collision point is the earliest intersection of
the swept-sphere paths the two spheres travel on
# selfdot(pos1+vel1*t) = rsquared
# selfdot(pos2+vel2*t) = rsquared
# selfdot(touchpt - (pos1 + vel1*t)) = rsquared
# selfdot(touchpt - (pos2 + vel2*t)) = rsquared
# simpler is to compare their distances to the sum of
their radii (inhibition forced us to look up on the internet >( )
# selfdot(pos1 + vel1*t, pos2 + vel2*t) = 4rsquared
# even simpler is to consider one the reference frame
for the other
# selfdot(pos12+vel12*t) = 4rsquared
# sum(pos12)**2 + sum(pos12*vel12)*t + sum(vel12)**2*t
# it's a quadratic equation where A = sum(vel12)**2, B
= sum(pos12*vel12), and C = sum(pos12)
# the quadratic equation is (-B +- sqrt(B^2-4AC))/(2A)
# so it's immediately rational from A, meanwhile you'd
need B^2-4AC to be a square number or a rational of square numbers
