preliminary comments Peter McGoron (31 Aug 2026 13:12 UTC)
Re: preliminary comments Hernán Ibarra Mejia (01 Sep 2026 09:37 UTC)
Re: preliminary comments Peter McGoron (01 Sep 2026 11:10 UTC)
Re: preliminary comments Hernán Ibarra Mejia (02 Sep 2026 14:47 UTC)
Re: preliminary comments Peter McGoron (02 Sep 2026 15:18 UTC)
Re: preliminary comments Peter McGoron (01 Sep 2026 18:28 UTC)
Re: preliminary comments Hernán Ibarra Mejia (02 Sep 2026 15:14 UTC)
Re: preliminary comments Marc Nieper-Wißkirchen (03 Sep 2026 08:48 UTC)

Re: preliminary comments Peter McGoron 02 Sep 2026 15:14 UTC

I tried a bunch of R7RS-Small implementations and they seem to agree
with you, so it doesn't seem to be a problem and the sample
implementation likely doesn't need to export them.

The R7RS-Small describes auxiliary syntax as a "binding," and a binding
usually has to be imported. Some implementations (erroneously?) match
auxiliary syntax not by their binding. Example:

     (define else #f)
     (cond (else 10))

On CHICKEN 6, this returns `10`, and on Chibi, it returns nothing. This
is because CHICKEN 6 matches by symbolic name and Chibi matches by
binding (I think).

One benefit of handling auxiliary syntax with the library system is that
one can rename them to avoid conflicts. However, this is pretty rare:
people are unlikely to shadow `else` and `return/m` with their own thing.

-- Peter McGoron