William D Clinger <will at ccs.neu.edu> writes:
>>From that, it sounds as though the editors are happy
> to require every NaN to be eqv? to every other NaN,
> but several people have already expressed their
> opposition to such a requirement, and I would like
> to make it clear that I too oppose it.
I may have misunderstood Alan's sentence: (eqv? +nan.0 +nan.0) is
unspecified, and at least I am happy with it. However, this makes the
recent addition of a restriction on `put-datum' premature.
--
Cheers =8-} Mike
Friede, V?lkerverst?ndigung und ?berhaupt blabla
Received on Sat Jun 30 2007 - 08:14:16 UTC