morphisms between multivariate polynomial rings

[Martin R]

I was a bit surprised about the following:

sage: k = ZZ["q"].hom(QQ["q"])

sage: isinstance(k, Morphism)
True
sage: m = ZZ["q,t"].hom(QQ["q,t"])
sage: isinstance(m, Morphism)

False

However, I vaguely recall a discussion about morphisms between multivariate polynomial rings being problematic.  Does my memory trick me?

Martin

[Travis Scrimshaw]

Hi Martin,

   That is coming from this:

sage: type(m)
<class 'sage.structure.coerce_maps.DefaultConvertMap_unique'>

Although I don't think we have a specific "all morphisms should inherit from this ABC (or generic base class)" specification. It might be easy to get something to work for it.

Are you thinking of the difficulty of determining a consistent way of setting coercions between polynomial rings? That's the only thing I know that is problematic.

Best,

Travis