[Axiom-developer] Re: fixing SPAD

[Martin Rubey]

Gabriel Dos Reis <address@hidden> writes:

>    I believe you raise a fundamental problem, which is that Spad
> should allow combination of predicate through usual logical operators.

"combinations of predicates through usual logical operators." *are* allowed.
Only, in the context of conditional exports, predicates are limited to "is" and
"has".

Aldor, in contrast, makes no (or maybe, not much) difference to other parts of
the code here.

Thus, my question is, would it be easy, or at least, doable for you, to make
SPAD behave like Aldor here, i.e. drop the restriction on the predicates
allowed?

> For a given specific case, one might look for a workaround that simulates
> type checking at runtime.

I doubt this.  Of course, using Waldek's idea we could say that 'Complex F' is
only allowed if either F is not a field or x^2+1 is irreducible in F.  But I'm
not sure if that's what we want.  And I doubt that you can achieve what I first
proposed in current SPAD, namely, make 'Complex F' export 'Field' only if x^2+1
is irreducible in F.

I guess, your answer (whether it would be "easy"), is "no"...

Martin