On Sat 21 May 2011 08:42, Emmanuel Medernach <emmanuel.medernach@gmail.com> writes:

> On Fri, May 20, 2011 at 9:44 PM, John Cowan <cowan@mercury.ccil.org> wrote:
>
>     a compiler can reject (map (foo)) as erroneous, but under a
>     "return on the stack" implementation[0], (foo) might return both the
>     function and the list to map. 

Note that I disagree with John's terminology here, as explained in
another mail.

> Why not specifying that arguments in procedure calls should each
> evaluate to exactly one value ?

The current vague situation is best IMO.  An implementation may truncate
the result to one value.  AFAIK S7 splices the values into the procedure
call, as John is describing.

Andy
--
http://wingolog.org/

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