[Axiom-developer] Re: Problem kTuple wish 2.

[Stephen Wilson]

Hello Ralf,

Ralf Hemmecke <address@hidden> writes:
[...]
> Try to define this apply function.
> 
> Important, I want to have as much type safety as possible, i.e. If I
> would have written
> 
> c := m(1, "x", false);
> 
> the compiler should already shout that the type of the first argument
> doesn't fit.

Initial thinking suggests that this would be difficult, if not
impossible, to implement genericly using static types.  I know of no
way to specify a non-homogeneous `cross' of arbitrary dimension, for
use as a meaningfull signature for `m' which the compiler could check.

Its been said before that static typing, when pushed to the limit,
becomes equivalent to dynamic typing -- that is to say, they are not
polar opposites.

Certainly this kind of generality is not on the radar for me any time
soon.

> 
> Good luck.
> 
> Ralf


Thanks,
Steve