[plt-scheme] Please help test version 359.100

At Tue, 14 Nov 2006 16:33:49 +0100 (CET), hufflen jean-michel wrote:
> >From plt-scheme-bounces at list.cs.brown.edu  Tue Nov 14 13:55:51 2006
> >(...)
> 
> >> I don't think this is relevant. The true question is whether
> >>
> >>   (+ (for-each add1 '()) (for-each sub1 '()))
> >>
> >> steps to
> >>
> >>   (+ some-value some-value)
> >>
> >> or to
> >>
> >>   (+ some-value some-different-value)
> 
>    Since the value returned by "for-each" is left unspecified

But it isn't left unspecified. It is specified to be the specified
thing that the r5 report just calls "unspecified". (how confusing is
that?!)


> some
> implementations may choose to return zero value.  So does MIT Scheme:
> 
> 1 ]=> (for-each (lambda (x) x) '(1 2 3))
> 
> ;No value
> 
>    I suppose that this effect results from the "(values)" expression.  If we
> admit this, the second point is how to compare expressions resulting from the
> "values" function.  In MIT Scheme, the equality functions already return
> false:
> 
> (eq? (values) (values))    ====> #f
> (equal? (values) (values)) ====> #f
> 
>    So it seems to me that unspecified values cannot be compared or composed
> w.r.t. a portable way.

You can only admit this, as Jacob points out, if you have coercians for
multiple values to allow them to work in single value contexts.

Robby