(2022-02-24, 22:22)David Manley Wrote: There's some challenging angles in there (LDraw model obtained from Technicopdia) :
My code is (in effect) simply drawing circles in 2D space. Solving it for this model looks like it would need some kind of (having Google-d it earlier today) Jacobian Inverse Kinematics solver. Way beyond the capabilities of the part alignment script.
Yikes.

But as long as there's only a single solution, then it should be solvable just with triangles…shouldn't it? I'm sort of tempted, but I think I'll take a little break from Technic for the moment. :-)