Dead project? - Page 5

mdettweiler · 2010-05-08, 16:30

Quote:

Originally Posted by em99010pepe

Thief...grow up.

Ugh...give it a rest, man, will you? Let's not turn another thread into a pointless flamewar.

em99010pepe · 2010-05-08, 16:44

Quote:

Originally Posted by mdettweiler

Ugh...give it a rest, man, will you? Let's not turn another thread into a pointless flamewar.

No, you guys are a bunch of thieves and that must be said. Leave the project alone until you see the opinion of the current users, rover and konrad127123.

mdettweiler · 2010-05-08, 17:18

Quote:

Originally Posted by em99010pepe

No, you guys are a bunch of thieves and that must be said. Leave the project alone until you see the opinion of the current users, rover and konrad127123.

At no point were we going to do anything without the input of the current users.

All that's occurred so far is discussion about possible ways to proceed.

henryzz · 2010-05-08, 19:13

part of the reason they are posting is because they want current users to post

ltd · 2010-05-08, 21:01

Quote:

Originally Posted by mdettweiler

Ah, you're right, I do see that the latest prime k has been removed from the status page now. When you say it's removed from the distribution queues, though, do you mean that it's actually removed from the LLRnet server, or just future batches of work to be automatically loaded into the server?

There is a cron job running every hour. It does several things.
First it removes all tests for a k in the actual llrnet queue if a prime is found
Second it removes all tests for which a valid factor was reported within the last hour. Again from the real llrnet queue.
Third if the size of the queue drops below a defined value new tests are imported to the llrnet queue. There are no batches of work just a continuous stream of tests feeded in as soon as they are needed.
Forth all statistic pages are updated.

There is no manual work needed for all this and as the job runs every hour it is much better looking at the client resources as any manual procedure as there is no way to forget to remove factors or to forget to look at found primes for a while.

By the way the only reason why the primes do not show up automatically is to give the administrator of a project a chance to validate if it is a real prime or just a false positive.

gd_barnes · 2010-05-13, 07:40

There seems to be a huge misunderstanding about my last statement. Yes, I see that the automated things have run and that the prime is showing up on the automated stats pages here.

But the admins have not updated this:

http://www.mersenneforum.org/showpos...44&postcount=1

nor this:

http://www.mersenneforum.org/showpos...20&postcount=2

Not only does the prime need to be shown but the k needs to be removed and the # of k's remaining needs to be reduced.

Also, this link which shows all of the primes for base 5 is a dead link:

http://geocities.com/base5_sierpinski_riesel/

It is in this thread:

http://www.mersenneforum.org/showthread.php?t=10448

Wouldn't it be wise to keep your threads up to date? All of these things would take < 10 mins. to update.

10metreh · 2010-05-13, 19:09

There aren't any admins here. They have been inactive for so long that they aren't mods anymore.

gd_barnes · 2010-05-13, 21:25

I am aware of that. What I was attempting to imply was: Wouldn't it be best if someone was a stand-in part time admin for the project so that various threads could be updated until such a time that a full time admin was found?

Probably if an admin from another math project of some type, whether it be factoring or prime search, were to take at least a small interest and offer to do nothing more than keep the various threads up to date, he might be able to contact Mike (Xyzzy) and receive admin access here. Mike will probably just want to see that the person has been around a while and that a couple of people on this project would be OK with it.

Perhaps LTD or 10metreh would be interested.

10metreh · 2010-05-14, 17:23

Quote:

Originally Posted by gd_barnes

Perhaps ... 10metreh would be interested.

Nope. One project is enough for me.

KEP · 2010-05-16, 13:28

I just did some testings, and on my "Intel(R) Pentium(R) D CPU 3.00GHz", it took for the range 102.04T-102.14T an average of 4352.6 s/factor running the sieve file availeable in the base 5 folder. I'm not sure how to compare this to beyonds testing times on his 1.8 and 2.0 GhZ Dual Cores, where testing takes ~58 and ~50 minutes (if I'm not mistakening). The times is when testing with the new LLRnet. So anyone who can tell if this project is actually at the point where new sieving is needed or if this project is good to go for further LLR testing?

If there is no moderator, I suggest that either Gary becomes moderator here and ensure the update of the given reservations and removal of primed k's. If this is not desired, I suggest that this project is put out for a vote, weather or not to move it over under the CRUS project or leave it here in a more or less dormant stage. Any thoughts?

KEP

Beyond · 2010-05-16, 17:25

I too did a small sieve test last night. File downlaoded from mersenneforums per link in the sieve thread. Removed the recently found k (by rover) from the file. I had noticed the last post by Rincewind stating he had completed up to 102210, so I ran 102210 to 102220. Found 5 factors at at rate of 4244 sec/factor. Odd thing when I uploaded the results all factors were already in the database so all duplicates or someone has been sieving without posting ranges completed?

Anyway it appears the sieve is good, for now.

I think CRUS already has a full plate as is ... best to leave SR5 where it is for now and let someone update the reservation file/posts.

2010-05-13, 07:40	#50
gd_barnes May 2007 Kansas; USA 3³·5·7·11 Posts	There seems to be a huge misunderstanding about my last statement. Yes, I see that the automated things have run and that the prime is showing up on the automated stats pages here. But the admins have not updated this: http://www.mersenneforum.org/showpos...44&postcount=1 nor this: http://www.mersenneforum.org/showpos...20&postcount=2 Not only does the prime need to be shown but the k needs to be removed and the # of k's remaining needs to be reduced. Also, this link which shows all of the primes for base 5 is a dead link: http://geocities.com/base5_sierpinski_riesel/ It is in this thread: http://www.mersenneforum.org/showthread.php?t=10448 Wouldn't it be wise to keep your threads up to date? All of these things would take < 10 mins. to update. Last fiddled with by gd_barnes on 2010-05-13 at 07:45

2010-05-16, 13:28	#54
KEP Quasi Admin Thing May 2005 1111000110₂ Posts	I just did some testings, and on my "Intel(R) Pentium(R) D CPU 3.00GHz", it took for the range 102.04T-102.14T an average of 4352.6 s/factor running the sieve file availeable in the base 5 folder. I'm not sure how to compare this to beyonds testing times on his 1.8 and 2.0 GhZ Dual Cores, where testing takes ~58 and ~50 minutes (if I'm not mistakening). The times is when testing with the new LLRnet. So anyone who can tell if this project is actually at the point where new sieving is needed or if this project is good to go for further LLR testing? If there is no moderator, I suggest that either Gary becomes moderator here and ensure the update of the given reservations and removal of primed k's. If this is not desired, I suggest that this project is put out for a vote, weather or not to move it over under the CRUS project or leave it here in a more or less dormant stage. Any thoughts? KEP Last fiddled with by KEP on 2010-05-16 at 13:30

2010-05-16, 17:25	#55
Beyond Dec 2002 444₁₀ Posts	I too did a small sieve test last night. File downlaoded from mersenneforums per link in the sieve thread. Removed the recently found k (by rover) from the file. I had noticed the last post by Rincewind stating he had completed up to 102210, so I ran 102210 to 102220. Found 5 factors at at rate of 4244 sec/factor. Odd thing when I uploaded the results all factors were already in the database so all duplicates or someone has been sieving without posting ranges completed? Anyway it appears the sieve is good, for now. I think CRUS already has a full plate as is ... best to leave SR5 where it is for now and let someone update the reservation file/posts. Last fiddled with by Beyond on 2010-05-16 at 17:26

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
PrimeGrid asks: Is PSP dead?	AG5BPilot	Prime Sierpinski Project	23	2015-11-19 12:59
Looks Dead Here	storm5510	Operation Billion Digits	5	2013-10-16 14:17
PRP and LLRNET = Dead	magnav0x	Prime Sierpinski Project	14	2005-12-14 22:02
CPU 1 dead in duallie?	Jeff Gilchrist	Hardware	10	2005-11-03 11:53
Dead P3	Prime95	Hardware	10	2003-09-10 22:41

2010-05-08, 19:13	#48
henryzz Just call me Henry "David" Sep 2007 Cambridge (GMT/BST) 1011011111000₂ Posts	part of the reason they are posting is because they want current users to post

2010-05-13, 19:09	#51
10metreh Nov 2008 2·3³·43 Posts	There aren't any admins here. They have been inactive for so long that they aren't mods anymore.

2010-05-13, 21:25	#52
gd_barnes May 2007 Kansas; USA 3³×5×7×11 Posts	I am aware of that. What I was attempting to imply was: Wouldn't it be best if someone was a stand-in part time admin for the project so that various threads could be updated until such a time that a full time admin was found? Probably if an admin from another math project of some type, whether it be factoring or prime search, were to take at least a small interest and offer to do nothing more than keep the various threads up to date, he might be able to contact Mike (Xyzzy) and receive admin access here. Mike will probably just want to see that the person has been around a while and that a couple of people on this project would be OK with it. Perhaps LTD or 10metreh would be interested.