[disgruntledphd2]

It converts really large integers to floats, which can lose precision. You need to handle really large numbers as characters, or use bit64.

I only really started hitting this in the private sector though, never a problem when I was in academia.

[infogulch]

Oh. That's kindof... gross. Better than wrapping silently I guess. Can you at least tell when a number has been 'promoted' to a float so you know you need a bigger type?