I wrote:
> Here's my first batch of numerics bugfixes and other changes for
> improved mathematical correctness and R6RS compliance. As far as
> I can tell, they're ready to commit. Reviews solicited.
Apologies for the wasted bandwidth, but I now realize that R6RS does not
actually require that (eqv? +nan.0 +nan.0) be false, so I will let it
remain true for the sake of backward compatibility.
Therefore, for the time being, please disregard:
0008-equal-and-eqv-equivalent-for-numbers-and-not-eq.patch
and all patches after it. I will post revised patches soon.
Mark
