On Fri, Sep 7, 2012 at 8:57 AM, Per Bothner <per@bothner.com> wrote:
> On 09/06/2012 03:58 PM, Alex Shinn wrote:ity -
>
>> Chez modules cannot be implemented by every
>> other system, nor can Scheme48 modules, but
>> the R7RS define-library form can.
>
>
> Not to argue for or against "Chez-style" modules,
> but is this really meaningful?  The point of R7RS is
> to specify and language, and so every other system
> that claims to modify R7RS would have to be *modified*
> to implemented the specification.
>
> I assume you mean that Chez modules are be so
> *fundamentally* different from other module systems
> that it would be need a huge and unreasonable re-write
> that it would not be reasonable for R7RS.

More or less.  What I actually mean is that
it should be trivial to define a static translation
between R7RS library declarations and any
existing module system.  It would be impossible
to do the same with Chez modules and any
system which wasn't basically already the
same as Chez.

In practice if we require Chez modules, the
existing incompatible systems will either not
support R7RS, or support it separately and
incompatibly from its own modules.  In that
implementation, R7RS will be disjoint and
second class, and not gain uptake with its
users.

--
Alex

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