This is a relatively small and hopefully simple script to calculate when you
should stop running ECM curves on a number and start dedicating your resources
to NFS instead. Work in progress, though primary functionality works already.

The basic premise is that when the ECM-work-done/odds-of-success-for-that-work
ratio becomes larger than the estimated NFS work, that's when you switch. In
turn, this module takes care of (most of) the details in actually calculating
the odds of success.