Harald Geyer wrote:
>
>
> I think this would unify all suggestions we had in this thread so far
> and it should be consistently extendable for inf/minf/und etc.
>
Even if we agree on this, all the possibilities for operations on
these things must be defined, and they are not obvious. :(
> Also it might help to implement some kind of nan-arithmetic.
>
There are papers on this topic trying to make a consistent selection of
definitions.
NaNs are, I think, real. There could be NaN+%i*NaN also :)
RJF