“…Next, one computes the action of φ T on on this basis, which gives an explicit matrix for the action of φ T as an R-linear transformation of End k (φ). It is then claimed in [15] (proof of Proposition 5.14) that this calculation yields a basis of End k (φ) as an A-module, although the details of this deduction are not explained. We have not pursued this line of calculations, so we are unable to say how complicated it is in practice, and whether it suffices for deducing the finer number-theoretic properties of End k (φ), such as its discriminant over A or the Frobenius index.…”