On Sun, 14 Aug 2005, Martin Gasbichler wrote:

> Andre van Tonder <xxxxxx@now.het.brown.edu> writes:
>
>> I have been wondering about the following technical issue:
>>
>>    (let ((x 1))
>>      (let-syntax ((m (lambda (form)
>>                        (let ((x 2))
>>                          (syntax x)))))
>>        (m)))
>>
>> ...
>
> I don't think this has anything to do with namespaces. The variable
> generated by (syntax x) and the binding introduced by the LET live in
> different phases (or stages). Hence they must not interfere.
>
> The syntax-case model is like an interpreter whose source code
> bindings interact with the variables of the interpreted program.

I'm inclined to agree, even if doing this entails a break with the traditional
model.  It seems like the right thing but may be somewhat hairy to implement,
though.  For example, presumably the following:

   (let ((x 1))
     (let-syntax ((m (lambda (form)
                       (let ((x 2))
                         (let-syntax ((n (lambda (form)
                                           (let ((x 3))
                                             (syntax (syntax x))))))
                           (n))))))
       (m)))

should give 1.  For each identifier, the expander needs to keep track of its
binding at all different levels of the syntactic tower.

Cheers
Andre