mersenneforum.org  

Go Back   mersenneforum.org > Prime Search Projects > Conjectures 'R Us

Reply
 
Thread Tools
Old 2015-08-15, 19:28   #2322
rebirther
 
rebirther's Avatar
 
Sep 2011
Germany

2·32·151 Posts
Default

Reserving S708 to n=100k (25-100k) for BOINC

Last fiddled with by rebirther on 2015-08-15 at 19:41
rebirther is offline   Reply With Quote
Old 2015-08-15, 20:18   #2323
pepi37
 
pepi37's Avatar
 
Dec 2011
After milion nines:)

2·19·37 Posts
Default S808

Reserving S808 as new up to 25K
pepi37 is offline   Reply With Quote
Old 2015-08-16, 08:46   #2324
gd_barnes
 
gd_barnes's Avatar
 
May 2007
Kansas; USA

1030810 Posts
Default

Quote:
Originally Posted by pepi37 View Post
S708 if finished.
49K remains
ABCD file with all K remain up to 100K sent to Rebirther to process on SRBase-Boinc
Base released - results attached
Quote:
Originally Posted by rebirther View Post
Reserving S708 to n=100k (25-100k) for BOINC
Pepi and wombatman, PLEASE STOP all work using srbsieve!! Reb, please stop S708 immediately in BOINC. There are far more k's remaining than what Pepi sent you.

I found it highly unlikely that there would be only 49 k's remaining at n=25K for S708 because I had run this base to n=2500 in the recommended bases thread. There were 701 k's remaining at n=2500! That is far too large of a reduction so I began checking the primes file. There are many composites. In the first 1000 k's, here is a balancing of what I found:
845 primes shown in file
163 were actually composite
682 true primes

Attached is my PFGW run showing 163 composites. I ran my old reliable PFGW 3.3.6 against the primes file. I independently checked 30-40 of the composites at the factoring db. All had factors. The most surprising of all was the very first one:
2*708^1+1 = 13 * 109 (!!??)

Besides the first one, it appears that all of the remaining composites are from n=7 to 50.

Many of the k's with composites would prime at larger n-values n<25K but not all. With > 700 k's remaining at n=2500, there should likely be 200-300 k's remaining at n=25K.

Mark, I am concerned about the continual corrections to srbsieve that I am seeing in the "Testing new Ranges for Sierpinski/Riesel" thread. We cannot "test in production" on this project. We need independent parallel tests run on many different bases before bases are submitted for work here. I am not convinced that that is happening. (I did it myself on base 3 and it looked good.) My question is: Is this an srbsieve issue or are the users not using it properly or do they have possible bad versions of NewPGen, PFGW, LLR, or srsieve in their folder when running the program?

For people who have been at CRUS < 1 year, I would prefer it if they would work with people who have more experience with the project when running new bases. Starting new bases is the most challenging thing that we do here. S708 did not pass the smell test. 49 k's remaining is very unlikely for such a high base with such a high conjecture and should have been independently verified.
Attached Files
File Type: txt S708-1K-results.txt (39.0 KB, 67 views)

Last fiddled with by gd_barnes on 2015-08-16 at 09:42
gd_barnes is offline   Reply With Quote
Old 2015-08-16, 09:21   #2325
rebirther
 
rebirther's Avatar
 
Sep 2011
Germany

2×32×151 Posts
Default

ok, thats very bad. I have cancelled the base S708.
rebirther is offline   Reply With Quote
Old 2015-08-16, 09:25   #2326
gd_barnes
 
gd_barnes's Avatar
 
May 2007
Kansas; USA

22×3×859 Posts
Default

For independent verification of S708, attached are all primes and k's remaining at n=2500. Also included is the starting bases script that I used.
Attached Files
File Type: zip S708 to n=2500.zip (79.8 KB, 72 views)
gd_barnes is offline   Reply With Quote
Old 2015-08-16, 09:56   #2327
pepi37
 
pepi37's Avatar
 
Dec 2011
After milion nines:)

2×19×37 Posts
Default

Thanks for warning. All is stopped!
pepi37 is offline   Reply With Quote
Old 2015-08-16, 14:11   #2328
wombatman
I moo ablest echo power!
 
wombatman's Avatar
 
May 2013

1,741 Posts
Default

If I am currently running with pfgw (since srbsieve was having the checkpointing problem), can I continue on the ones I started? I'm also willing to verify the "primes" in pl_prime to confirm that they are actually primes. And if any turn out not to be, I can re-run that particular k up to the same n as the others.

If, however, that's not acceptable, please let me know and I'll stop.

Edit: Nevermind, I saw your request for the new-base script results as well, so I'll just do that to try and verify.

Last fiddled with by wombatman on 2015-08-16 at 14:17
wombatman is offline   Reply With Quote
Old 2015-08-16, 14:21   #2329
rogue
 
rogue's Avatar
 
"Mark"
Apr 2003
Between here and the

141458 Posts
Default

Quote:
Originally Posted by gd_barnes View Post
Mark, I am concerned about the continual corrections to srbsieve that I am seeing in the "Testing new Ranges for Sierpinski/Riesel" thread. We cannot "test in production" on this project. We need independent parallel tests run on many different bases before bases are submitted for work here. I am not convinced that that is happening. (I did it myself on base 3 and it looked good.) My question is: Is this an srbsieve issue or are the users not using it properly or do they have possible bad versions of NewPGen, PFGW, LLR, or srsieve in their folder when running the program?
The last version before I added the recovery logic is good. They are not sieving properly with newpgen. That is being discussed in the other thread.
rogue is offline   Reply With Quote
Old 2015-08-17, 03:10   #2330
paleseptember
 
paleseptember's Avatar
 
Jun 2008
Wollongong, .au

3×61 Posts
Default

Starting R576 from new using the NewBases v4.3 script. CK is 30651.
paleseptember is offline   Reply With Quote
Old 2015-08-17, 06:13   #2331
MyDogBuster
 
MyDogBuster's Avatar
 
May 2008
Wilmington, DE

54448 Posts
Default

Quote:
Starting R576 from new using the NewBases v4.3 script. CK is 30651.
R576 is already at n=25K. Do you mean S576?
MyDogBuster is offline   Reply With Quote
Old 2015-08-17, 06:15   #2332
paleseptember
 
paleseptember's Avatar
 
Jun 2008
Wollongong, .au

18310 Posts
Default

Quote:
Originally Posted by MyDogBuster View Post
R576 is already at n=25K. Do you mean S576?
Curses and phooey. Yes, S576. Thankfully it was churning away for <3 hours. I can't edit my original post, so let's say that I'm taking S576 from new with the NewBases script. AFAICS it's the lowest CK unstarted.

Last fiddled with by paleseptember on 2015-08-17 at 06:18
paleseptember is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
Bases 251-500 reservations/statuses/primes gd_barnes Conjectures 'R Us 2257 2021-02-27 09:18
Bases 101-250 reservations/statuses/primes gd_barnes Conjectures 'R Us 890 2021-02-24 18:35
Riesel base 3 reservations/statuses/primes KEP Conjectures 'R Us 1081 2021-02-23 16:13
Bases 33-100 reservations/statuses/primes Siemelink Conjectures 'R Us 1677 2021-02-21 21:42
Bases 4-32 reservations/statuses/primes gd_barnes Conjectures 'R Us 1431 2021-02-18 16:05

All times are UTC. The time now is 23:30.

Tue Mar 2 23:30:48 UTC 2021 up 89 days, 19:42, 0 users, load averages: 2.52, 2.52, 2.52

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.

This forum has received and complied with 0 (zero) government requests for information.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.
A copy of the license is included in the FAQ.