Negative exponents for MPolys

[SoongNoonien]

In the context of Nemocas/AbstractAlgebra.jl#2385 I found a problem with negative exponents in Nemo:

julia> R,(x,y)=ZZ[:x,:y];

julia> R([ZZ(1)], [[ZZ(-1), ZZ(0)]])
x

julia> R([ZZ(1)], [[-1, 0]])
x^18446744073709551615

In the case of negative ZZRingElems the sign is ignored and in the case of negative Ints the actual bits of the Ints are interpreted as a UInt. I think both should just raise an error. What do you think?

[lgoettgens]

i agree

[fingolfin]

Yes, should raise an error