mersenneforum.org > News holy tethered cow! new Mersenne prime? (M43-related)
 Register FAQ Search Today's Posts Mark Forums Read

2005-12-16, 14:11   #12
ppo

Aug 2004
italy

113 Posts

Quote:
 Originally Posted by pacionet What do you mean with "blooper" ? I don't speak and read English perfectly and Italian translators for "blooper" say "blooper" , they don't help much ..
you should use an online english dictionary, like
Code:
blooper

• noun informal, chiefly N. Amer. an embarrassing error.

2005-12-16, 14:17   #13
ppo

Aug 2004
italy

113 Posts

Quote:
 Originally Posted by fatphil verification should occur before the new year, almost certainly
not so sure, holidays are near

2005-12-16, 14:32   #14
pacionet

Oct 2005
Italy

3×113 Posts

Quote:
 Originally Posted by jasonp A colloquial term for a comical mistake. GIMPS has had machines report an assignment as prime but for which the verification run didn't work. Each time Prime95 gets more changes to avoid false alarms. I suspect the waiting is much worse on the machine's owner than the rest of us :) jasonp
Thanks , I understood.

 2005-12-16, 14:43 #15 akruppa     "Nancy" Aug 2002 Alexandria 246410 Posts Hm, guess it's me who should have used a dictionary. "Blooper" may be read like I was referring to the false positive we had once, but that wasn't my intention. I referred to the fact that most (all?) exponents of new Mersenne primes had been figured out by GIMPS members before the official announcement, usually because a little bit too much info on the exponent was leaked out. Maybe I should have said "slip of tongue". Alex
 2005-12-16, 15:54 #16 HiddenWarrior     Jun 2003 Russia, Novosibirsk D616 Posts Great! I think the verifications will be done to 31 december, than we will have super present hell yeah!!!
 2005-12-16, 16:12 #17 fatphil     May 2003 3×7×11 Posts What would the cost of Prime95 (or whatever it's called), dropping way-points, such that each quarter/tenth/whatever of the LL test could be performed in parallel, and so verification could be days rather than weeks? I can't believe this hasn't been suggested before (it's in C&P for example) - what was the conclusion?
2005-12-16, 16:41   #18
drew

Jun 2005

2·191 Posts

Quote:
 Originally Posted by fatphil What would the cost of Prime95 (or whatever it's called), dropping way-points, such that each quarter/tenth/whatever of the LL test could be performed in parallel, and so verification could be days rather than weeks? I can't believe this hasn't been suggested before (it's in C&P for example) - what was the conclusion?
I think that's a great idea. These residues can be stored locally, so it wouldn't require any extra server capacity.

The cost would be tens of megabytes stored on one's PC.

Drew

Last fiddled with by drew on 2005-12-16 at 16:42

2005-12-16, 17:16   #19
jasonp
Tribal Bullet

Oct 2004

3·1,163 Posts

Quote:
 Originally Posted by fatphil What would the cost of Prime95 (or whatever it's called), dropping way-points, such that each quarter/tenth/whatever of the LL test could be performed in parallel, and so verification could be days rather than weeks? I can't believe this hasn't been suggested before (it's in C&P for example) - what was the conclusion?
Doesn't Prime95 save the result 2000 squarings before the end? That would at least let someone else with the program verify in a few minutes whether things are obviously bogus

jasonp

 2005-12-16, 17:23 #20 pacionet     Oct 2005 Italy 3·113 Posts Sorry I have two last questions : 1) Verifications were usually done by Tony Reix e Jeff Gilchrist ,if I am not wrong, and they took usually some days (5-10 days). Is it correct ? Can these guys tell us the exponent or they can't (because it s forbidden ) until official announcement ? 2) If I found a new Mersenne prime ( ) could I post a thread on this forum with the probable Mersenne prime I' ve found before official announcement ? What happens if I reveal the exponent before official announcement ? They don't give me the money ? Thanks
 2005-12-16, 17:48 #21 ppo     Aug 2004 italy 113 Posts I don't see a real advantage in speeding up the verification. It is already fast enough, and, provided that everyone behaves according to the rules that are clearly spelled out, we can wait a few days. George has always provided some clues so to let us have some fun in trying to beat the official announcemet, for me it is part of the game. He is able to do a quick check, by repeating the last few iterations, just to make sure, but waiting for the double confirmation to complete is a necessary step to give credibility in the academic world to the discovery.
2005-12-16, 18:39   #22
T.Rex

Feb 2004
France

2·457 Posts

Quote:
 Originally Posted by pacionet 1) Verifications were usually done by Tony Reix and Jeff Gilchrist, if I am not wrong, and they took usually some days (5-10 days). Is it correct ? Can these guys tell us the exponent or they can't (because it s forbidden ) until official announcement ?
I cannot believe someone has ALREADY found a new Mersenne prime ! It is wonderful . Unbelievable !
I can see the page where this UNVERIFIED prime appears. Can't believe it !

And, again, it is during vacations ... So, I guess my colleagues do not fully use their big 16x or 32x Itanium2 machines and I could use one of them for checking this candidate. If I get authorizations. And if I find someone in the building ... I'm also in vacations but I could see if something can be done this Monday.

I do not know what is the exponent. George should.
If George needs help for quickly verifying this candidate new prime, I'm OK to help and to follow his rules.
The computation should take 6 days if no problem. Probably 7 or 8. I'm out of the town 23, 24 and 25th.

But maybe Jeff could use the new machine his University received in mid 2005. Jeff ?

Incroyable !

Tony

 Similar Threads Thread Thread Starter Forum Replies Last Post a1call Miscellaneous Math 23 2020-09-17 13:17 ixfd64 News 901 2008-11-10 19:29 Uncwilly News 132 2005-05-10 19:47 adpowers Lounge 40 2004-08-12 22:05 ixfd64 Puzzles 2 2003-09-23 12:53

All times are UTC. The time now is 01:58.

Sun Nov 1 01:58:02 UTC 2020 up 51 days, 23:09, 2 users, load averages: 1.94, 1.75, 1.63