Bob Margulies skrev:

> A few hours ago, Prime95 version 23.8.1 reported M32091881 is not prime
> and appears to have gone directly to LL testing of M27471827 without
> doing any factoring. Is that OK? Usually factoring takes at least a day.

Regarding the factoring discussion, quoting from whatsnew.txt:

> New features in Version 24.12 of prime95.exe
> --------------------------------------------
>
> [---]
>
> 3)  The factoring breakeven points have been recalculated using my 2 GHz
P4.
>     This version of prime95 will do less trial factoring.

In version 24.12 of Prime95 the default factoring depth of Prime95 was
changed. Trial factoring is normally done as a separate work type and then
handed out to someone else for the Lucas-Lehmer test. (See menu
Test/PrimeNet for types of work to request.) P-1 factoring is done just
before the Lucas-Lehmer test and not assigned separately.

If the trial factoring is done by someone with version 24.12 or newer, then
when the Lucas-Lehmer test is assigned to you your client (v 23.8) might
decide to do the trial factoring one bit deeper before starting on the P-1
factoring and L-L test.

On the other hand, if trial factoring was done by someone with a client
older than v 24.12, then your Prime95 client would start the P-1 factoring
and L-L test directly.

The new client also chooses lower P-1 bounds if the number has been trial
factored deeper (as compared to what is normal for the new client). I don't
know what the bounds are compared to the old client.

A line in worktodo.ini like

Test=35263819,62,0

means a first-time Lucas-Lehmer test of M35263819, where trial factoring has
been done to 62 bits depth and P-1 factoring has not been done. (Last number
is 1 when P-1 is done.)
--
Patrik Johansson
http://www.bahnhof.se/~pjoh/index.html


_______________________________________________
Prime mailing list
[email protected]
http://hogranch.com/mailman/listinfo/prime

Reply via email to