Not sure I should have this DC....

[petrw1]

Exponent Status Data  
18728873 No factors below 2^67 
 P-1 B1=225000, B2=4500000 
 Unverified LL 46D44BAAF2DE42__ by "Ezram van Rijnberk" 
 Unverified LL 9FF61215444796__ by "Stefan Klinke" 
 Unverified LL FB879860A8D339__ by "petrw1" on 2008-12-21 
 Assigned Double-checking to "petrw1" on 2008-12-21 
 History FB879860A8D339__ by "petrw1" on 2008-12-21

Three LL's/DC's. Three different residues. That alone is a concern.

But, one DC was mine and it is back assigned to me to DC again. Should I release it?

[lycorn]

What is normally considered "good pratice" is to have, for a given exponent, the first time test and the DC perfomed by different participants.
So I would say "yes, release it".

[cheesehead]

The use, since several versions ago, of a pseudorandom bit-shift at the beginning (and reversal at the end) provides noticeable protection against faulty hardware's producing the same erroneous residue from the same cause during two successive L-Ls on the same system. So there is not quite the same suspicion of same-system results that there had been before. However, there are still sources of error that would not be affected by that bit-shift.

IIRC v4 reports showed that shift count along with the returned residue.

[petrw1]

Originally Posted by cheesehead

The use, since several versions ago, of a pseudorandom bit-shift at the beginning (and reversal at the end) provides noticeable protection against faulty hardware's producing the same erroneous residue from the same cause during two successive L-Ls on the same system. So there is not quite the same suspicion of same-system results that there had been before. However, there are still sources of error that would not be affected by that bit-shift.

IIRC v4 reports showed that shift count along with the returned residue.

So are you implying I should or should not drop it OR it doesn't really matter anymore?

[cheesehead]

Originally Posted by petrw1

So are you implying I should or should not drop it OR it doesn't really matter anymore?

Be above suspicion, like Caesar's wife. (http://www.brainyquote.com/quotes/qu...aes161092.html)

Drop it like a hot potato.

[petrw1]

Hereby dropped....

[S485122]

I searched the double-checked results for exponents above 40M. There where quite a lot double checked by the same user and sometimes the same computer. Some of the results are obviously multiple submissions of the same result... See the attached file.

Jacob

[cheesehead]

Originally Posted by S485122

quite a lot double checked by the same user and sometimes the same computer

Caesar would've ordered them to be DCed independently ... or else divorced them.

[lycorn]

And those exponents are lucky that Caesar was not like Henry VIII...