Ciao,
some time ago I asked if predicates like NUMBER? and
INTEGER? must return #t only when applied to Guile's built
in numbers, or they can be overloaded by extensions to
return #t when applied to other number-like objects upon
which it is possible to apply all the common number
operators, like +, - and SIN. And I think Neil J. answered
that: yes, only built in Guile numbers should be #t.
I still have the idea that it is convenient that INTEGER?,
REAL? and COMPLEX? can be applied to classes of number-like
objects; for example: I do not see why a user could not want
to test the realness of a custom vector with REAL?, so that
functions like:
(define (my-op a)
(if (real? a)
(some-thing a)
(some-other-thing (real-part b))))
can be used for both built in real numbers and extension
added number like objects, like vectors and matrices.
To distinguish built in reals from other "real" things
there is the IS-A? predicate, and to make "old" code
continue to work there are modules.
So how bad it is to break this?
--
Marco Maggi
"They say jump!, you say how high?"
Rage Against the Machine - "Bullet in the Head"
_______________________________________________
Guile-user mailing list
Guile-user@gnu.org
http://lists.gnu.org/mailman/listinfo/guile-user
