"Paul F. Dietz" <dietz@...> writes:
> The reader goes off into lala land when given a potential
> number token with the format of a float with an excessively
> large exponent:
> (read-from-string "1.0s1000000000") ==> loops
I don't think it does, actually. It's trying to construct a very
large bignum, that it will eventually attempt coerce to a short float.
Arguably, this behaviour sucks and needs to be fixed, but I don't
think it's wildly wrong.