(defun Paral3D (obj1 obj2 / a1 a2 acos)
(setq a1 (vlax-curve-getFirstDeriv obj1 0)
a2 (vlax-curve-getFirstDeriv obj2 0)
acos (/ (apply '+ (mapcar '* a1 a2))
(* (sqrt (apply '+ (mapcar 'expt a1 '(2 2 2))))
(sqrt (apply '+ (mapcar 'expt a2 '(2 2 2))))
(zerop (atan (/ (sqrt (- 1. (* acos acos))) acos)))
While this is interesting, do you think there's any reason to revise MinDist 1.0 to use it?
I think not assuming it is only testing for parallel lines.