but, it has the sideffect that the rotatable part JUMPS to the mouseposition. it would be much nicer and more logic if the rotated object would keep a relative distance to the point from where the mouse drags it in circles. comparable to a simple drag-offset.
hmm. after trying some trigonometry i gave up and searched the web, but couldn't find any answer.
as the offset of the rotatable object may only be on the x axis, one may
calculate this difference from pivot point 0,0 on mouse down.
by adding this difference on (enter frame) where we calculate the degree
with some math formula we get what we want.