mersenneforum.org  

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

Reply
 
Thread Tools
Old 2018-01-01, 09:10   #45
gd_barnes
 
gd_barnes's Avatar
 
May 2007
Kansas; USA

10,723 Posts
Default

Quote:
Originally Posted by KEP View Post
It includes no teeny tiny primes. Because of the way srbsieve works, it is very easy to automatize the process of finding out exactly how many k's each conjecture has that are Trivial k's, MOB and for the Sierpinski side GFN. It is not useless, at least thats not how it is viewed by the thousand new users that support CRUS through BOINC at srbase.

To be clear, I'm not (like when I created the srbsieve.ini files) searching for any primes other than on base 16 for now. But even though the stats we already have now is good and great, they say nothing at all about how far we have actually gotten

I'm btw gonna publish my version 0.1 of the more accurate statistics on srbase website, very soon. The stats I publish will contain the data available on january 1st on the various subpages on the CRUS website. So far more than 22T trivial factors has been counted and 374 GFN aswell as 2.5+T MOB's

Happy new year everyone

Take care

You took my posting out of context. Trillions of teeny primes...trillions of trivial k's...trillions of k's that are MOB. It's all of what I am talking about.

You could remove those statistics at SRBase and the BOINC users would be better off. What they should be encouraged to look at is the number of proven, 1k, 2k, 3k, etc. bases remaining. Encourage them to move bases into such colums.

Our statistics say everything that we need to about how far we have gotten. K's remaining on all bases is useless information and a waste of your time and CPU resources. CRUS will never take on bases like R280/S280. Why waste your time?

I indicated my displeasure in this thread with such statistics before. The starting of large-conjectured bases should not be encouraged.

We have all of the statistics that we need about the project here: http://www.noprimeleftbehind.net/crus/. We don't need any more.

I have an idea: Perhaps we can close off the project for future bases with conjectures > 1M. Then your statistics would be meaningless unless of course you want to take on managing those conjectures yourself.


Last fiddled with by gd_barnes on 2018-01-01 at 10:46
gd_barnes is offline   Reply With Quote
Old 2018-01-01, 09:48   #46
gd_barnes
 
gd_barnes's Avatar
 
May 2007
Kansas; USA

10,723 Posts
Default

This issue apparently has gotten out of hand on the SRBase site and that is my fault. The project admins here at CRUS needed to take more personal time to give the project a better direction. I now have more such time. I had not taken a look at the site in over a year and had not realized that these untested k's and bases have been over-emphasized on the main page there. I just now had a look at that page and was a little disappointed by what I saw in two places, especially after indicating a year ago in this thread how such information encourages the project to move in an incorrect direction. I would like to suggest that we shut that down because it is conterproductive to accomplishing what the project needs: proving bases and reducing k's remaining on reasonably-conjectured bases.

For the uninitiated, the SRBase BOINC home page is here:
http://srbase.my-firewall.org/sr5/

Two sections in question:

(1)
>>
Project status:
  • Riesel Base: 392/1031 proven (18.09.2017)
  • Sierpinski Base: 384/1032 proven (19.02.2017)
    278 unstarted bases (09.12.2017)

    706,136,815,151,905 Total k's remaining (01.06.2017)
    0.0055% done
<<

Reb, I'd like to ask for the removal of the total number of k's remaining as well as unstarted bases. You could show the number of bases with 1, 2, or 3 k's remaining. Alternatively, perhaps in that area you could provide a link to our pages that show the number of proven bases, and the number of bases with 1/2/3 k's remaining as shown here:
http://www.noprimeleftbehind.net/cru...crus-stats.htm Alternative #2: You could remove all of your "hard-coded" statistics there and just provide the link. That would give your searchers all of the info. that they really need for the project.

Even though the above link shows the number of unstarted bases, it can be seen in the larger context of how many bases are at different places in their search depth: a much better picture. (VBCurtis had also suggested showing something along these lines in post #2 in this thread.)

(2)
>>
Compared with last years the following process was made:

2014-2015 - 87M WUs processed, 1 base solved, 314 bases unstarted
2015-2016 - 88M WUs processed, 5 bases solved, 297 bases unstarted
2016-2017 - 99M WUs processed, 5 bases solved, 282 bases unstarted
in total - 274M WUs processed, 11 bases solved, 282 bases unstarted
<<

Reb, I'd like to ask for the removal of bases unstarted. CRUS will not be starting all bases such as R280/S280 anyway so it is misleading. Besides, all unstarted bases (except one) have conjectures greater than 1e5 now and so the remaining untested bases will be done very slowly.

I like all of the other info. in those two sections.

Reb, we greatly appreciate everything that BOINC and the community have done for the project. I'd like not to offend you in suggesting this but I feel it is necessary so that we encourage more searching where it is is needed.

And two final things:
First, this request has nothing to do with my or anyone else's effort involved in managing large conjectured bases. I now have much more personal time than I did a few months ago and that will not change in the near future.

Second, I do plan to request some admin help in the near future. Ian does help me from time-to-time but he has taken on a lesser role in the last few years.

In other words, this has to do with project direction, not my personal effort or anyone else's effort.


Gary

Last fiddled with by gd_barnes on 2018-01-01 at 10:59
gd_barnes is offline   Reply With Quote
Old 2018-01-01, 11:35   #47
rebirther
 
rebirther's Avatar
 
Sep 2011
Germany

3,121 Posts
Default

Thx for the suggestions Gary, I have removed the total k's and the percent status but will keep the unstarted bases, it is interesting for all to see how many bases are left to start and motivate users to run some of them.

If you have some important bases on the recommended side add them to the table.
rebirther is offline   Reply With Quote
Old 2018-01-01, 19:26   #48
pepi37
 
pepi37's Avatar
 
Dec 2011
After milion nines:)

101111011112 Posts
Default

If I may say something.
Many of us here are people that have one, two computers, so dont have many cores on disposal. I like CRUS project very much, and will stay, and will give my small contribution as long as I can. And there is one most important thing:you can choose any of many bases, and work on any K you wish ( that is part where you Gray come, and say: OK take only one K from that conjuncture and work on it, and for that part I am very grateful.) First part of making and solving any base is sieve. And that is most boring part since you spent time, and you cannot find anything. And I can say : finding prime is to all of us main goal. Now we have YOYO at home to help us.
Person behind that project is great. I ask to help me sieving my own repdigit sequences, and he didn't spend any micro seconds to say: of course that you can send me all files you have. Second, YOYO can in with current resources sieve to depth that I , like individual can only dream off. And since you can never over sieved any CRUS sequences, right question is: why dont "join" YOYO and CRUS.
From one side, YOYO will have work for many months, and for us: we will have properly sieved sequence , will have many of them and at least will removed many small exponent/candidates. We all know that will not solve all CRUS bases in our lifetime, but we can make more or less work, right? So make it more work :) Also we can concentrate on solving bases,we dont need to make sieve.
Reb , with his project will also have benefit:properly sieved files are welcome to him also. Why spend somebody else resources on any base because REB was taking under sieved sequence. If we have resources, then use it in most efficient way: dont spent time, because sieve depth is not good for even 500K candidate and he tested 1M candidate ( for example).

So : to all in this New year: lets make some work: lets remember that 2018 is and will be good CRUS year, with many eliminated K and bases!
pepi37 is offline   Reply With Quote
Old 2018-01-01, 19:27   #49
KEP
Quasi Admin Thing
 
KEP's Avatar
 
May 2005

3D016 Posts
Default

Reb, please remove my connection to the statistics, the current effort is cancelled.
KEP is offline   Reply With Quote
Old 2018-01-01, 21:14   #50
MisterBitcoin
 
MisterBitcoin's Avatar
 
"Nuri, the dragon :P"
Jul 2016
Good old Germany

2×52×17 Posts
Default

Quote:
Originally Posted by KEP View Post
Reb, please remove my connection to the statistics, the current effort is cancelled.


So, bad news on this site.
Oh I´m also trying to bring sr2sieve into yoyo@home to speed up the search.
But I have some problems with the file size and double-checking in order to prevent false/missed factors.

As a result of the latest discussion I´ll no longer start new bases or start new ranges until it´s wished from Gary. Sorry rogue, you´ll need to finish S3 by your own.
Running work will be finished.
MisterBitcoin is offline   Reply With Quote
Old 2018-01-04, 08:59   #51
gd_barnes
 
gd_barnes's Avatar
 
May 2007
Kansas; USA

10,723 Posts
Default

Quote:
Originally Posted by MisterBitcoin View Post

So, bad news on this site.
Oh I´m also trying to bring sr2sieve into yoyo@home to speed up the search.
But I have some problems with the file size and double-checking in order to prevent false/missed factors.

As a result of the latest discussion I´ll no longer start new bases or start new ranges until it´s wished from Gary. Sorry rogue, you´ll need to finish S3 by your own.
Running work will be finished.
MisterBitcoin,

We have greatly appreciated your help in coordinating the sieving with yoyo. The tremendously sieved files are great for us especially for BOINC.

Just to clarify: I never stated that I wanted to stop the project efforts for already started bases with conjectures > 1G. Continuing to work on S3 is great. It's over half done to n=25K at this point. We may as well get it completely searched to n=25K. Regardless I'm sure Mark (rogue) can finish it off sometime this year.

Will you be continuing on your base 7 and 15 reservations? What about new bases S196 and S732? I would especially like to see S732 finished to at least n=10K because that base and R411 (that I am working on) are the only bases left with conjecture < 100K that have not been completed to n=10K yet. It would be a great milestone for the project.

I know you enjoy working on new bases. I'm not against it but am ambivalent about them at this point. I'm certainly not stopping anyone. It's part of what makes the project interesting to some people. I just ask that we not start on any more new bases with conjectures > 1G. The project k's remaining statistic in this thread seemed to encourage as much.


Gary

Last fiddled with by gd_barnes on 2018-01-04 at 09:40
gd_barnes is offline   Reply With Quote
Old 2018-01-04, 09:12   #52
gd_barnes
 
gd_barnes's Avatar
 
May 2007
Kansas; USA

1072310 Posts
Default

Quote:
Originally Posted by rebirther View Post
Thx for the suggestions Gary, I have removed the total k's and the percent status but will keep the unstarted bases, it is interesting for all to see how many bases are left to start and motivate users to run some of them.

If you have some important bases on the recommended side add them to the table.
Thanks Reb.

I suppose that seeing the history of the not started bases is somewhat interesting for your searchers. I'll back off that request.

As for the recommended bases, BOINC can run any of those in that thread even if they are not in the BOINC table. There are quite a few of them right now. There are three 1-k bases recommended, one of which is reserved. I am working on sieve files for the other two. One will be done this weekend. BOINC could work on those.

MrBitcoin recently sent us a lot of sieve files for 1-k bases that were done by his effort with Yoyo all deeply sieved up to n=1M. You might check the sieving thread. There are a large number of 1-k bases that BOINC could work on. We could prove a lot of bases just from those sieve files.

Here is my reasoning behind not having k's remaining statistics: The project could just remove all bases with conjectures > 1T. There is only something like 7-8 of them on each side (Riesel and Sierp). What would this do to such statistics? It would remove 99.9%+ of all k's that are remaining on the project!

I simply feel that showing the percentage of k's remaining makes the project look bad at this point and is unnecessarily misleading.


Gary

Last fiddled with by gd_barnes on 2018-01-04 at 09:25
gd_barnes is offline   Reply With Quote
Old 2018-01-04, 09:27   #53
gd_barnes
 
gd_barnes's Avatar
 
May 2007
Kansas; USA

10,723 Posts
Default

Quote:
Originally Posted by KEP View Post
Reb, please remove my connection to the statistics, the current effort is cancelled.
I'm not sure what this means. Can I ask either KEP or Reb, were there other statistics (other than total k's remaining on started and unstarted bases) that KEP was working on for SRBase?

Last fiddled with by gd_barnes on 2018-01-04 at 09:34
gd_barnes is offline   Reply With Quote
Old 2018-01-04, 17:27   #54
rebirther
 
rebirther's Avatar
 
Sep 2011
Germany

60618 Posts
Default

Quote:
Originally Posted by gd_barnes View Post
I'm not sure what this means. Can I ask either KEP or Reb, were there other statistics (other than total k's remaining on started and unstarted bases) that KEP was working on for SRBase?
I think this thread http://srbase.my-firewall.org/sr5/fo...ead.php?id=203
rebirther is offline   Reply With Quote
Old 2018-01-04, 18:25   #55
MisterBitcoin
 
MisterBitcoin's Avatar
 
"Nuri, the dragon :P"
Jul 2016
Good old Germany

2×52×17 Posts
Default

Quote:
Originally Posted by gd_barnes View Post
MisterBitcoin,

We have greatly appreciated your help in coordinating the sieving with yoyo. The tremendously sieved files are great for us especially for BOINC.

Just to clarify: I never stated that I wanted to stop the project efforts for already started bases with conjectures > 1G. Continuing to work on S3 is great. It's over half done to n=25K at this point. We may as well get it completely searched to n=25K. Regardless I'm sure Mark (rogue) can finish it off sometime this year.

Will you be continuing on your base 7 and 15 reservations? What about new bases S196 and S732? I would especially like to see S732 finished to at least n=10K because that base and R411 (that I am working on) are the only bases left with conjecture < 100K that have not been completed to n=10K yet. It would be a great milestone for the project.

I know you enjoy working on new bases. I'm not against it but am ambivalent about them at this point. I'm certainly not stopping anyone. It's part of what makes the project interesting to some people. I just ask that we not start on any more new bases with conjectures > 1G. The project k's remaining statistic in this thread seemed to encourage as much.


Gary

S7, S15 will be done, hopefully this year.
S732 is nearly done, only a few LLR tests remaining.
S196 is on the fast lane! Passed 6K two days ago.

Don´t worry Mark, I´ll throw 8 cores into the S3 race.
MisterBitcoin is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
Progress bsquared YAFU 20 2014-05-22 16:52
Progress R.D. Silverman Factoring 0 2012-05-22 14:03
Special project #3b - Project 400 schickel Aliquot Sequences 307 2011-10-28 01:29
Special project #3a - Project 300 schickel Aliquot Sequences 29 2011-08-12 17:45
In Progress? R.D. Silverman Cunningham Tables 33 2010-05-07 14:02

All times are UTC. The time now is 13:51.


Wed May 25 13:51:27 UTC 2022 up 41 days, 11:52, 0 users, load averages: 1.76, 1.32, 1.24

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2022, 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.

≠ ± ∓ ÷ × · − √ ‰ ⊗ ⊕ ⊖ ⊘ ⊙ ≤ ≥ ≦ ≧ ≨ ≩ ≺ ≻ ≼ ≽ ⊏ ⊐ ⊑ ⊒ ² ³ °
∠ ∟ ° ≅ ~ ‖ ⟂ ⫛
≡ ≜ ≈ ∝ ∞ ≪ ≫ ⌊⌋ ⌈⌉ ∘ ∏ ∐ ∑ ∧ ∨ ∩ ∪ ⨀ ⊕ ⊗ 𝖕 𝖖 𝖗 ⊲ ⊳
∅ ∖ ∁ ↦ ↣ ∩ ∪ ⊆ ⊂ ⊄ ⊊ ⊇ ⊃ ⊅ ⊋ ⊖ ∈ ∉ ∋ ∌ ℕ ℤ ℚ ℝ ℂ ℵ ℶ ℷ ℸ 𝓟
¬ ∨ ∧ ⊕ → ← ⇒ ⇐ ⇔ ∀ ∃ ∄ ∴ ∵ ⊤ ⊥ ⊢ ⊨ ⫤ ⊣ … ⋯ ⋮ ⋰ ⋱
∫ ∬ ∭ ∮ ∯ ∰ ∇ ∆ δ ∂ ℱ ℒ ℓ
𝛢𝛼 𝛣𝛽 𝛤𝛾 𝛥𝛿 𝛦𝜀𝜖 𝛧𝜁 𝛨𝜂 𝛩𝜃𝜗 𝛪𝜄 𝛫𝜅 𝛬𝜆 𝛭𝜇 𝛮𝜈 𝛯𝜉 𝛰𝜊 𝛱𝜋 𝛲𝜌 𝛴𝜎𝜍 𝛵𝜏 𝛶𝜐 𝛷𝜙𝜑 𝛸𝜒 𝛹𝜓 𝛺𝜔