On Sun, 24 Apr 2011, Aaron W. Hsu wrote:

> On Sun, 24 Apr 2011 15:50:15 -0400, Andre van Tonder <andre@het.brown.edu>
> wrote:
>
>> On Sun, 24 Apr 2011, Aaron W. Hsu wrote:
>>
>>> On Sun, 24 Apr 2011 15:34:23 -0400, Andre van Tonder <andre@het.brown.edu>
>>> wrote:
>>>
>>>> The latter behaviour may be compatible with the document in systems in
>>>> which all identifiers are considered implicitly bound at toplevel, so the
>>>> DEFINE would be like a SET!
>>>
>>> Even in systems where this is the case, and Chez in particular, this does
>>> not do this.
>>
>> I thinkthe reason Chez does not doi this is becasue you are replacing a
>> syntactic binding for ELSE by a variable binding, so the DEFINE is not like
>> a SET!.  IF the original binding of  ELSE was a variable binding, the
>> define would be like a SET! and the ELSE clause would match.
>
> It does not matter whether the else is a syntactic binding or a variable
> binding. Chez behaves the same in either case regarding this. Take this for
> example:
>
>> (library (a)
>   (export test else)
>   (import (except (chezscheme) else))
>   (define-syntax test
>     (syntax-rules (else)
>       [(_ else) 'else] [(_ blah) #f]))
>   (define else #f))
>> (import (a))
>> (test else)
> else
>> (test x)
> #f
>> (define else #t)
>> (test else)
> #f
>>
>
> The definition actually alters the lexical binding; library exports are
> immutable, and thus a redefinition is more like a shadowing than a set!.

I agree with your example, but I was referring to the case where everything was
at the top-level (no modules involved), where the behaviour is different.

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