Hi Val, well it is just not quite well defined what the "center" of a wiggly arbitrary curve is.

Do you mean the center of the bounding box around the curve?

For a closed curve, the centroid of mass as if the curve was a solid plate of material?

The center of the curvature circle at a particular point of the curve?


> Or is this something that just cannot be coded?

There are several problems with coding it. Not so much in just doing the calculation, but rather in coding it to work quickly so that things don't slow down a lot. Some of the above calculations can be time consuming to calculate.. Even the center of the bounding box can be a bit tricky if you want to get a very accurate "tight" bounding box that hugs exactly against the shape of the curve. It's possible to do an approximate tight bounding box quicker but I worry about using approximations in snapping points.

On a larger model it would be hard to avoid things freezing up until all these center calculations were finished.

The basic problem is that the center can be quite a distance away from the actual outline of the curve shape itself, so it is hard to postpone doing the calculations - with some stuff like intersections I can postpone calculating the intersection until your mouse moves nearby the curves that are involved. But I can't really tell if your mouse is nearby the center without doing the full calculation to determine the center...

- Michael