John Cowan <cowan@mercury.ccil.org> writes:

> How about this compromise:  simply remove the clause defining `eqv?` on
> non-IEEE flonums?  It is arguably not a proper domain for standardization
> anyway, since there are no such implementations today.  That would allow
> future implementations to return `#t` or `#f` at their discretion.

This would be *vastly* better than the current situation.  If it's the
best we can hope for, then _please_ do this.  This would make it very
likely that implementations would correctly extrapolate the definition
of 'eqv?' to other representations.

> I have changed "conforming to" to "implemented in the style of",
> which I think eliminates that problem.

This is also helpful.  Thanks.

    Regards,
      Mark

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