Waldek Hebisch wrote: > So while old algorithm > remains correct, a superior one may be invented and replace > the old.
Sure. And I would agree with that. I was thinking more along the lines of the build system, the interpreter, client-server communication, etc. I would expect the mathematics to grow and change, although I DO hope that category theory as a framework will endure and scale. > There is an extra twist to this: I suspect that polynomials > in question are relatively prime. With high probability > a simple modular method will discover this. But there is > a question how to hook modular method so that it gets > used. I belive that the polynomials have type 'UP(Expression Integer)' > and you need some extra work to determine if modular method > is applicable. > > In short: new algorithms are discovered and to use new algorithm > you may be forced to change parts of the system which at first > glance have nothing to do with new code. That sounds reasonable to me - I would regard that as an investigation into new mathematics and as such a legitimate reason to update Axiom's mathematical logic. What I'm hoping to avoid is needing to do anything major with the non-mathematical components of the system - they should do what is needed and not break. How to achieve this is probably where the debate comes in, but I'll not start that up again. Cheers, CY P.S. - how does one sign up for the new email list? _______________________________________________ Axiom-developer mailing list Axiom-developer@nongnu.org http://lists.nongnu.org/mailman/listinfo/axiom-developer