In the attached you will find a curve that is the result of a circle being projected onto a sphere. When this curve is offset on the sphere, the result is rather buggy.
offset on srf.3dm (291.5 KB)
And continuing with that model, when the sphere is split with a fixed version of the offset curve and then offset solid, the result (in the attached) are two surfaces that are not connected but somehow are joined anyway.
offset on srf - test parting line.3dm (325.7 KB)
Hi Wim - Yeah… if you rotate the curve so that the seam point is on the sphere seam, it works as expected; if you explode the results you got, you can delete the extensions and join up the rest into the more expected result. The extending part is by design but I am not so sure it makes sense in this context, I’ll add a bug for this - it might be tune-upable.
http://mcneel.myjetbrains.com/youtrack/issue/RH-31235
I am not seeing the same thing as in your second example though - I get the expected solids. I’ll poke some more.
thanks,
-Pascal