9 minutes ago, Ray Dillinger wrote:
> What I could and did implement without running into problems was
> lazy and eager functions as opposed to delayed and forced versus
> eager evaluations. But the semantics of mixing calls to lazy and
> eager functions can be confusing, especially when the functions are
> evaluated in the course of expanding macros.

... Which leads to another bad point in R5RS that R7RS retains: saying
that the language might treat a promise as equivalent to its forced
value.  That kind of behavior is not only difficult to do and
confusing as a result, it also leads to a very different language
which "suffers" from some of the things that lazy languages do.  (For
example a "value" can now actually be a computation).  Leaving it as a
possible option in a standard is IMO reckless.

--
          ((lambda (x) (x x)) (lambda (x) (x x)))          Eli Barzilay:
                    http://barzilay.org/                   Maze is Life!

_______________________________________________
Scheme-reports mailing list
Scheme-reports@scheme-reports.org
http://lists.scheme-reports.org/cgi-bin/mailman/listinfo/scheme-reports