Originally posted by Greenbank

It's this last test that I noticed this line in results.txt (20% checked so far):-

p=32258662112491 k=24737 n=3007

32258662112491 is 1831 x 61231 x 287731

How was this missed in the initial sieve?
Actually it wasn't missed!

32258662112491 24737 3007 2437 61 e

The e says that this was an "excluded" result. ie it did not exist in the orginal sob.dat that was sieved to 1G

ProthSieve (and some versions of SoBSieve) produce extraneous factors. ProthSieve writes these to factexcl.dat This could also have been produced by a factor program (P-1, ECM or other) which produces factors, not necessarily prime factors.

We probably should not be keeping these excluded factors. They only clutter up the databases. Yes, I know, I was one of the people that urged we keep them when we started the sieve effort. They do have a purpose, but a limited one. Note, I am not saying that we don't need the factexcl.txt files. There are factors in them that are important to keep. The problem is that all sort of factors go into factexcl. The same is true of factrange. Not all of the factrange factors are valuable. Some of the factrange factors are also excluded factors.

We need to keep:
1) The lowest factor found for a k n pair
2) the first factor found for a k n pair

But to keep the database small, we need to limit these factors to those above some minimum value, 1M at least or 1G the way we have been doing for the most part.
And yes, I do have a Database of these factors. It is in Berkely DB form but I can produce it in a flat file. Most of these factors are also in this file produced daily by MikeH. This file also includes duplicate and excluded factors and non prime factors. Almost everything that has been submitted is in there. A "d" at the end, possibly followed by a number indicates a duplicate. An "e" at the end indicates an excluded factor, that in my opinion should be ignored.

What VJS said about missed factors is very true, but does not directly answer Greenbank's question. It took a few readings of Greenbanks post for me to realize that he was talking about a nonprime factor rather one that had just been found and should have been found long ago. We are still finding those as VJS said. There are still more to find. I have a few ranges in mind and hope to have someone sieve them soon. Any volunteers?