On Mon, 24 Oct 2005, Taylor Campbell wrote:

>If Aubrey Jaffer's suggestion of NaNs encapsulating information about
>their origin be adopted, I think it would make sense to extend the NAN
>procedure to accept more arguments, of any nature, to specify any
>information whatsoever about the condition, like T's UNDEFINED-VALUE
>mechanism.  Of course, one possible value would be (LAMBDA ARGS (/ 0.0
>0.0)).

Sure. But it only makes sense on implementations that have more than one
NaN (and aren't using it for anything else like characters or fixnums),
so it's probably a good idea to standardize that the (nan) routine can
take additional arguments, but not specify that it has to do anything in
particular with them.

				Bear