Grant Miner <firstname.lastname@example.org> writes:
> Ah, a divide by zero :)
Yes, well, the direct floating point implementation in scsh still has
a number of lurking bugs. That's also the reason why Scheme 48 doesn't
enable it by default but uses an indirect implementation via the
run-time system. This implementation is slower but correct.
Scsh users who would like to switch back to Scheme 48's implementation
should remove "floatnums" from the variable "opens" in line 866 of
Makefile.in and rebuild scsh. You can still enable the direct
implementation on demand by loading the package "floatnums".