On Sun, Jan 22, 2012 at 3:09 PM, John Cowan <cowan@mercury.ccil.org> wrote:
> Ray Dillinger scripsit:
>
>> I feel that having them eq? would be "the right thing" but that
>> right thing-ness in this case isn't a sufficient reason by itself
>> to change the standard.  Now, if right thing-nness is accompanied
>> by established practice in more than, say, 80% of implementations,
>> or widespread support for the proposition among implementors, then
>> I'd be for changing the standard to require it.
>
> Okay, I'm back home and ran tests.  The vast majority of my
> implementations return #f to both (eq? (vector) (vector)) and
> (eq? (string) (string)).  I also tested bytevectors in those
> implementations that made it easy to do so.  Here are the cases that
> returned #t:
>
> Chez returned #t for both strings and vectors.
>
> Icarus (but not Vicare) returned #t for vectors but not strings or
> bytevectors.
>
> Ypsilon returned #t for all three cases.
>
> NexJ returned #t for vectors but not strings.

You missed Chibi, which returns #true for vectors and
#false for strings and bytevectors.

--
Alex

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