[FOM] When is it appropriate to treat isomorphism as identity?

[karim zahidi]

Thanks to Andrej en Vaughan for their responses.

One remark though.

That the number of roots is not computable for an arbitrary real
polynomial is non-computable is ofcourse true (I never disputed that).
However the non-computability is of a trivial kind (since you're working
over a non-computable field). It is trivial certainly compared with 
Martin Davis' result that there is no algorithm to decide whether an
arbitrary integer valued multivariate polynomial has infinitely many
integer solutions.

I was ofcourse mistaken in assuming that if you restrict to computable
coefficients, the problem becomes computable.

 Regards

Karim