[GiNaC-devel] why does power(add) switch the sign of the add?

[Alexei Sheplyakov]

Hi, Jan,

> Why not leave away this bit of code, improve performance by a tiny
> little bit, and get somewhat more predictable behaviour of GiNaC?

This code is responsible for rewriting rules like (x-y)^3/(y-x)^2 -> x
- y,  (2*x-2*y)^2*(y-x) -> -4*(x-y)^3, etc.

Also GiNaC does not care of term ordering (in Planck we trust).

Hope this helps,
        Alexei