Oh wait, I see what you are talking about there - the centerline is instead shoving off more to one side. I'll take a look tomorrow, that probably should not be too difficult to fix.

- Michael

Yes Michael, I meant the centerline didn't go through the middle of object thickness. Just revolve & shell the original curve (arc) yourself. -Petr

Hi Petr, thanks for reporting this shelling problem and for posting the example file.

I have now fixed this up for the next beta release. I tracked it down to a bug in the geometry library involving spheres but just in the special case of a partial sphere with the surface normal pointing inwards towards the center of the sphere which is what you have in this case. The surface normal would get mistakenly flipped for this particular thing.

This is also related to some other changes in v2 that will examine surfaces when you open a file and see if they are primitives like a sphere, cone, or cylinder, and create the surface using a special analytic surface class if they are. That enables things like offsets and intersections to be special cased for these surfaces to generate more simple output geometry.

- Michael