On 9/28/06, Neil Jerram <neil_at_ossau.uklinux.net> wrote:
> "Blake Coverett" <scheme_at_bcdev.com> writes:
>
> > Implementations of Scheme are required to implement the whole
> > tower of subtypes given in section 2.1.
> >
> > This puts a large burden on various specialized implementations.
>
> Why is that? I would have thought that the only effect of the new
> requirement would be that a specialized implementation could not say
> "This implementation conforms to R6RS." It would have to say instead
> something like "This implementations conforms to R6RS, except that it
> only implements exact fixnums, not the whole numerical tower." That
> doesn't seem like a large burden.
>
> > In fact, I would argue that the overall domain of applications for
> > which nothing but exact fixnums would suffice is larger that the
> > domain requiring any complex numbers at all.
>
> That may well be true. But surely the point of a standard is that
> someone can write a Scheme program that does use many of the described
> features, and then know that this program will run on any
> implementation that says it is R6RS compliant?
Your two statements are at odds with each other. How can developers
rely on R6RS compliance of implementations, if the specification is
complex enough that many implementations only partially comply?
--
Carl Eastlund
Received on Thu Sep 28 2006 - 10:44:40 UTC