Thank you for the fast response and correction. You are right.
Sorry, I have seen only one problem with the sign function and had no look at
the whole mathematic.
I think a complete solution of (-1)^(1/3) would be
-1, 1/2 + %i*sqrt(3)/2, 1/2 - %i*sqrt(3)/2
So we don't have a bug.
Next time I will think more carefully about the mathematic.
Dieter Kaiser