Specification bug? Alexey Radul 22 Sep 2009 12:28 UTC
Re: Specification bug? David Van Horn 22 Sep 2009 15:40 UTC

Re: Specification bug? David Van Horn 22 Sep 2009 15:40 UTC

Alexey Radul wrote:
> The specification of the procedure procedure-arity-includes? seems to
> imply the presence of arity information for all procedures.  In
> particular, the null implementation seems to imply that applying any
> procedure to any set of arguments will always result in an arity
> error.
>
> I suggest changing the specification of procedure-arity-includes?
> to say:
>
> Returns #t if the proc may be able to accept k arguments and #f
> otherwise. If this procedure returns #f, applying proc to k arguments
> will result in an arity error. If no arity information is available
> for proc, returns #t.
>
> (the last sentence above is an optional clarification only, that may
> not be worth including the SRFI document)
>
> and changing the null implementation to always return #t.

Yes, there's a bug -- I will change the specification and implementation
to the above for the next revision.

David