On Thu, Jun 7, 2012 at 8:45 PM, John Cowan <cowan@x>
Emmanuel Medernach scripsit:
I suppose that's all right.
> IMHO this is restrictive, having payload to be an exact non-negative
> integer (or a sequence of bits) precludes using it to store symbols
> if one (future) implementation wish to do so.
I have taken your remarks into account in the wiki page.
> Yes, one implementation may decide to always return the same oldWhat is the benefit of this?
> NaN. My point is that (nan) and +nan.0 should have the same behaviour:
I see it mostly as a consistency requirement. And to be able to distinguish between NaNs coming from different places.
About signaling NaNs, do we have to specify that the payload should be part of the exception signaled ?