Dave wrote:Update:
Remark 1: [tex]p_{k+1 } - p_{k } \ne 2 \lambda[/tex] over
E for some [tex]\lambda[/tex] such that [tex]1 \le \lambda < \frac{log^{2}(max(p_{k+1 }, p_{k}))}{2}[/tex].
Remark: We must redefine our current exceptional set,
E, to comply with remark one.
Remark: This problem is a big headache! Ouch!
Oops! Our current proof of Polignac's conjecture is wrong!! The proof of Polignac's conjecture should be about the spacing between consecutive odd primes.
Example: Suppose we want to exclude [tex]2 \lambda_{0 }[/tex] over
E.
Update:
We have [tex]p_{2 } - p_{1 } \ne 2 \lambda_{0 }[/tex] such that [tex]1 \le \lambda_{0 } < \frac{log^{2}(max((p_{1 }, p_{2}))}{2}[/tex].
_____________________________________________________________________________________________________________________________________________
Update:The chance that [tex]2 \lambda_{0 }[/tex] is the
wrong spacing between consecutive odd primes, [tex]p_{2 } > p_{1 }[/tex], is
roughly [tex]\frac{Floor( \frac{log^{2}(max(p_{1 }, p_{2}))}{2} - 1) }{Floor( \frac{log^{2}(max(p_{1}, p_{2}))}{2}) }[/tex].
Remark: "
Roughly" indicates too large.
However, over E, we generate the infinite product of similar values because of independence so that the chance [tex]2 \lambda_{0 }[/tex] is the
wrong spacing between consec
utive odd primes over E equates to zero:
Prob( [tex]p - q \ne 2 \lambda[/tex] over
E )
= [tex]\prod_{m=1}^{\infty }\frac{Floor( \frac{log^{2}(max(p_{m }, p_{m+1}))}{2} - 1) }{Floor( \frac{log^{2}(max(p_{m }, p_{m+1}))}{2}) } = 0[/tex].
_______________________________________________________________________________________________________________________________________________
In short,
Polignac's conjecture is still correct! But our previous reasoning was wrong! We hope we have
it right this time. We will review it later.
Remark: We apologized for the sloppy (flawed) math in previous posts.
Dave.
Go Blue! 