## Other

drummers-lowrise

Message boards : Fermat Divisor Search : Duplicate work?

 Subscribe SortOldest firstNewest firstHighest rated posts first
Author Message
wwwwww

Joined: 2 Aug 16
Posts: 2
ID: 453587
Credit: 23,429,626
RAC: 131,617

Message 138709 - Posted: 7 Mar 2020 | 17:58:36 UTC

Someone report this prime https://primes.utm.edu/primes/page.php?id=130734 to t5k page. Isnâ€™t this is the same as our k=11 in DIV project?. Does this mean weâ€™re double checking someone elseâ€™s work?
____________

JeppeSN

Joined: 5 Apr 14
Posts: 1378
ID: 306875
Credit: 21,623,822
RAC: 102,920

Message 138711 - Posted: 7 Mar 2020 | 21:34:43 UTC

It means someone (Ryan Propper) is "stealing" PrimeGrid's candidates by searching ahead of our leading edge. PrimeGrid will still have to cover that region. /JeppeSN

Reggie
Volunteer moderator
Volunteer tester
Project scientist

Joined: 10 May 14
Posts: 162
ID: 311759
Credit: 115,534,067
RAC: 156,716

Message 138715 - Posted: 8 Mar 2020 | 2:32:04 UTC

Since we don't have the residues, we are likely double checking some of his work. We do try to be somewhat smart about it, so the prime was automatically removed from our candidates to at least save some tasks.

JeppeSN

Joined: 5 Apr 14
Posts: 1378
ID: 306875
Credit: 21,623,822
RAC: 102,920

Message 138721 - Posted: 8 Mar 2020 | 7:59:31 UTC

He is now reporting a second Proth find for k=11, so reports both 11*2^8103463 + 1 and 11*2^9381365 + 1 in short succession.

The latter is reported to divide a GF(-, 6).

One gets the expression that he has exhausted Proth k=11 from some point near PrimeGrid's leading edge, to over n=9.3M. However, we cannot be sure he did not skip candidates or make mistakes or have bad residues.

I wonder if he did other low k Proth as well, kâ‰ 11.

/JeppeSN

serge

Joined: 21 Jun 12
Posts: 112
ID: 144858
Credit: 237,968,345
RAC: 134,411

Message 138724 - Posted: 8 Mar 2020 | 10:21:00 UTC

Didn't PrimeGrid in the past do exactly the same to other people's reservations at prothsearch.net ?

First this was promised:

"https://web.archive.org/web/20110608154530/http://www.prothsearch.net/index.html" wrote:
Here is how it will work. PrimeGrid will reserve large ranges of k and n. It will perform both a first test and a double-check of all candidates. When it is within weeks of completing this reservation, it will place another reservation for a smaller range of k and n. Of note, some of these ranges will overlap ranges currently reserved to other searchers. PrimeGrid will NOT SEARCH those ranges.

That agreement was then spat upon by PrimeGrid. The ranges were poached by PrimeGrid, the hits were reported by mere days earlier than owners would find them.

So, it is interesting after a number of years to see how the atmosphere that was sowed by PrimeGrid is now being reaped.
____________
My lucky number is Phi(4, 2^2396029-1)/2.

JeppeSN

Joined: 5 Apr 14
Posts: 1378
ID: 306875
Credit: 21,623,822
RAC: 102,920

Message 138725 - Posted: 8 Mar 2020 | 10:57:45 UTC

serge, do you happen to have full ((x)G)F divisibility information for these two primes by Ryan Propper? All I know is that the larger one divides GF(9381364,6). /JeppeSN

JimB
Honorary cruncher

Joined: 4 Aug 11
Posts: 916
ID: 107307
Credit: 974,494,092
RAC: 3,785

Message 138726 - Posted: 8 Mar 2020 | 11:21:40 UTC - in response to Message 138724.

Didn't PrimeGrid in the past do exactly the same to other people's reservations at prothsearch.net ?

First this was promised:
"https://web.archive.org/web/20110608154530/http://www.prothsearch.net/index.html" wrote:
Here is how it will work. PrimeGrid will reserve large ranges of k and n. It will perform both a first test and a double-check of all candidates. When it is within weeks of completing this reservation, it will place another reservation for a smaller range of k and n. Of note, some of these ranges will overlap ranges currently reserved to other searchers. PrimeGrid will NOT SEARCH those ranges.

That agreement was then spat upon by PrimeGrid. The ranges were poached by PrimeGrid, the hits were reported by mere days earlier than owners would find them.

So, it is interesting after a number of years to see how the atmosphere that was sowed by PrimeGrid is now being reaped.

Notice that I didn't say anything about this. It's our users complaining. I just PM'd you the results of a PFGW run on 11*2^8103463+1, seems like there should have been an official comment on it.

It was me who killed the prothsearch.net agreement, it had outlived its usefulness. I did not feel bound by agreements Lennart made. If you're really still mad after all this time I feel sorry for you.

Nobody owns the k=11 Proth candidates and Ryan is welcome to test them if he wants. PrimeGrid will automatically skip over the ones he reported without any human interaction. If Ryan would like to give us any LLR primality residues, we could use them to speed up our testing. But he's certainly under no obligation to do so. So PrimeGrid users, kindly shut up about this.

Message boards : Fermat Divisor Search : Duplicate work?