1. RE: nan
Robert Craig wrote:
>
>
> That way the Translator won't output "inf".
> In C, there doesn't seem to be any standard for
> symbols representing inf and nan. I'll see if I can
> do something in the Translator to get around this.
>
You might consider making NAN & INF built-in constants in Euphoria that
we can reliably detect. (i.e. "if x = NAN then...") I remember we went
round & round with that one a while back with everyone getting different
results on different systems and interpreter vs. translator. The only
way I found that worked in all cases was to convert the value to bytes
and then check that, which is a bit expensive if you need to check a
bunch of possibly-NAN function return values...
