branch cut fixup for atanh(bfloat(x))



>>>>> "Robert" == Robert Dodier <robert.dodier at gmail.com> writes:

    Robert> COMPLEX-ATANH applies the identity arg(x - y i) = - arg(x + y i)
    Robert> but that fails if y = 0 and x < 0. So I plugged in a special case for y = 0.

I think there needs to be more discussion about this.  Some Lisps
support signed zeroes (cmucl, sbcl), others do not.  Maxima's
bigfloats don't support signed-zeroes.  

What are we going to do about that?  I think that is a bigger issue
because when we've answered that, then we will know better about what
we want for the branch cuts.

Ray