Sagemath and pari/gp disagree about the degree of the zero polynomial over arbitrary rings. Since pari/gp is proper subset of sagemath, this gives us inconsistency in sagemath.

Session:

sage: K.<x>=QQ[];K(0).degree()
-1
sage: gp.poldegree(K(0))
-oo