6 May
2016
6 May
'16
5 p.m.
Aha, I see it now. For an infinite set of primes P, a sequence obeying the fibonacci recurrence ought to exist, which is never zero modulo P. Therefore, never divisible by P or P^2. Therefore, my "random" model will not apply to those prime moduli and you'll never need to divide out powers of those P to keep it squarefree. When this realization is added to the probability model, it should easily suffice to make it nonrigorously clear that the thing will blow up to infinity, if appropriate starting integers are used. -- Warren D. Smith http://RangeVoting.org <-- add your endorsement (by clicking "endorse" as 1st step)