Quote:
Originally Posted by pepi37
On Kamada page the is huge number of low nash base 10 sequences ( plus or minus side) Even with very small sieve depth I can remove many candidates, but I would like to do sieve to at least 1e14. And then I have problem since sr2sieve cannot do such depth on sequence like this 5539879179*10^n+1.
So does anyone here have some simpler solution except sieving one by one with srsieve and sr1sieve?
Update: for now srsieve2 is also in game (and since it is MT) it is fast solution :)

This k is too large (>2^32), thus sr2sieve cannot handle, like that sr2sieve cannot handle the sequence (k*b^n+c)/gcd(k+c,b1) (k>=1, b>=2, c != 0, gcd(k,c) = 1, gcd(b,c) = 1) such that gcd(k+c,b1) is even (if gcd(k+c,b1) is odd, then we still can use sr2sieve to sieve the sequence k*b^n+c for the prime not dividing gcd(k+c,b1) and remove the n such that there is some prime p dividing gcd(k+c,b1) which also divides (k*b^n+c)/gcd(k+c,b1))