News - Page 10

gd_barnes · 2010-03-19, 12:43

I'd like to suggest that we take a complete break from new bases for at least a week now. We've garnered a tremendous amount of new information over the last 1-2 weeks.

I counted at least 30 new bases over about a 5 day span and many have complex algebraic factor situations that are taking an inordinate amount of time to properly analyze and add to the pages. After many hours last night, I still have at least 10 of them to add for the base 501-1024 thread and a few from the base 33-100 thread. I'd really like to get on to some other things like responding to my Email and PMs.

Let's work on finishing the new bases that we have already reserved and then extending already searched bases to deeper depths.

Thank you,
Gary

gd_barnes · 2010-03-20, 10:26

We're currently in the process of a server/web pages move to a new machine. The noprimeleftbehind pages, which contain the CRUS pages will be down for a little while.

The good news is that I've fully updated the pages for all work done through this posting. Once the pages come back up, they should be up to date.

Edit: They came back up but they're not up to date. It's still the "old" pages from a day ago. We're experiencing some difficulty with the port forwarding related to the move.

Gary

AMDave · 2010-03-20, 13:00

Change completed
URL tests for CRUS and NPLB appear to pass.
AMDave

gd_barnes · 2010-03-20, 22:47

I confirmed that all web pages were correct and updated as of about 8 AM CDT (1 PM GMT) today.

If anyone sees any reservations or ranges completed not properly reflected as of the above time, please let me know. I'm now working on just the few more that have come in since then.

Gary

gd_barnes · 2010-03-23, 21:04

OK, guys. Feel free to unload on me again to a certain extent. That is, go ahead and test some new bases if you'd like. I'd only ask that we stick to no more than 1-2 new bases per person per day. That will help me not get too far behind when stuff comes up in my personal/business life for several days.

The break from the web page updates helped me quite a bit in getting caught up here and with some other admin stuff at NPLB. Thanks for going easy on me for a few days.

It also gave me time to "mentally" generalize the "new" kind of algebraic factors across all bases so that won't take me so long to show on the pages now. Although I want to get it down "in writing" in the algebraic factors thread, the analysis to generalize it across all bases is quite involved so it may be several days or possibly weeks before I get to that. In the mean time if they aren't clear to you, feel free to ask Serge, Ian, or myself if you have a new base with k's remaining that you think could be removed.

In other news, late tonight, I'm going to start a recommended bases and efforts thread in an attempt to give a little more direction to the project so I know a little more of what to expect about what people are going to reserve or work on. There will be a mixture of new bases, 1k remaining bases, and bases that are somewhat behind their "neighbors (i.e. bases nearby with similar conjectured k)" as to how far they have been searched. That will help avoid some of the unexpected new bases load that seems to come around from time to time.

Thank you to everyone for your immense interest in the project. I'd much rather be too busy with it than not busy at all. It's certainly turned out to be far more interesting math-wise than I would have ever originally expected when the project started.

Gary

gd_barnes · 2010-04-20, 07:41

Hey all,

I'll be on a business trip from Apr. 20th (later today) to Apr. 29th. If everyone can please limit their new base reservations to one per day during this time, I'd appreciate it. I'll be especially busy this coming weekend.

To clarify: If you've already reserved the bases and they've been shown on the pages, post as many results/primes/k's remaining on multiple bases as you would like per day. Bases not already on the pages are what take a while.

Thanks,
Gary

gd_barnes · 2010-04-21, 09:18

Hi all CRUS folks,

How'd you like a change from the conjectures for at least a couple of days? We'll be running an LLRnet rally at "No Prime Left Behind" (NPLB) from April 30th to May 2nd. See details at http://www.mersenneforum.org/showthread.php?t=13334.

For those who have never run an LLRnet rally, what we do is pile all of our machines on one server and see how many pairs we can process and primes we can find for 48 hours. We have both team and individual scores for pairs processed. Come join a team or be your own team. Either way, come join in the fun!

The tests will be at n=~765K-775K base 2 and should take ~10-11 mins. each on a modern machine.

Gary

gd_barnes · 2010-04-24, 09:28

After some contemplation today about what has transpired in the last few days/weeks/months on small-conjectured new bases and a PM that I received about someone wanting to do a large # of them on the Sierp side in the near future, I've decided to drop all bases > 512 from the CRUS project.

Don't worry, existing work will not be wasted and I'm not "cutting everyone off" at this moment in time. What I will do within the next 2-3 weeks starting in about 10 days is I will split up the web pages into bases <= 512 and > 512. If others wish to manage and update the pages for bases > 512, they are welcome to copy the pages to their own website and do so. I will be glad to assist in this process.

I also welcome someone creating a new project for bases > 512. For that matter, he can also extend it to as high of a base as he or everyone wants. Before doing that, obviously bases 250 to 512 will now have to be in one thread here with bases 513 to 1024 in another thread (vs. 250-500 & 501-1024). This will only amount to a little bit of tweaking and post moves. Then if someone wants to take bases > 512 into a different project or forum, we can get that thread moved to the new forum. (I guess I should have split them up in powers of 2 to begin with.)

I'm sorry but it's just become way too much for me to manage and it's taking too much time away from my personal and business life. When Robert posted all of the conjectures up to base 1024, a few people did some higher bases here and there and I was perfectly willing to accomodate those people and decided to extend the project definition to include bases up to 1024. But now that the new bases script has made it so easy to start new bases, I'm willing to admit that I've bitten off way more than I can chew from an administrative perspective. Also, admittedly, I've grown to dread the web page updates whereas previously, it was my favorite thing to do, mainly because removing the k's and adding bigger primes was fun, especially for the bases where a final large prime is found to prove them.

I had assumed that we would slowly work our way higher on the bases and I attempted to do that with the recommended bases/efforts thread, but it has gone largely ignored. For instance, at this time all bases <= 250 with a CK of < 10K are either at or nearly complete to n=25K. It is my thinking to keep slowly working our way higher on the bases in that manner...not just search only the small ones.

I will do what I can to keep up with what people post, including for bases > 512 until such a time that I get things split apart. I am still not willing to accept anything more than one new base per day while I'm out of town so please respect that. At this moment, I won't be able to make any updates until at least Monday.

I do appreciate everyone's enthusiasm for the project. I also appreciate people's offer to help with the HTML but ultimately, it's still too much to manage, because it's more than just updating the web pages. It's the checking, thread updates, asking for specifics when they aren't there, etc. Regardless, I'd much rather have it this way than the other way, with little to no activity.

Thank you,
Gary

kar_bon · 2010-04-24, 10:19

I think it's not a good idea to split bases >512 into a new project or new thread here!

All information should be available at one location to minimize irritations wher to search for a special info or posting people in the wrong place!

Suggestion: Freeze all reservations for bases >512 for now and push bases with only searched to n=25k higher and so minimize the remaining pairs. Especially bases with only one k remaining should be done more than n=50k or 100k.

It should also be manageable to write a script to automate the html-code for a base given with files of primes and info like remaining pairs, level tested. But as you posted, the most work is to verify all those results.

MyDogBuster · 2010-04-24, 13:14

Quote:

Suggestion: Freeze all reservations for bases >512 for now and push bases with only searched to n=25k higher and so minimize the remaining pairs. Especially bases with only one k remaining should be done more than n=50k or 100k.

I have to agree with Karsten. Just freeze any new bases work on anything > 512. Advancing bases already started shouldn't be a problem, I hope.
You can then adjust your upper limit as you see fit.

If someone wants to start a project for bases > 1024, then go ahead.

I have a few bases > 512 that I'm currently working on but will not start anymore. I still have quite a few cores doing bases < 512 and those will stay within that range. I do enjoy starting new bases, so I'll be working on the larger CK bases < 512 in the future.

Hope this helps. Ian

mdettweiler · 2010-04-24, 14:55

I'd be inclined to agree as well: freeze all reservations for new bases > 512, but keep the existing ones on the pages. Then, hopefully down the road when some sort of automation has been developed to make the process easier, we can reconsider opening them up again. In the meantime, though, there's no reason not to continue work on the already-started bases in that range, many of which have only one k remaining and which I'm guessing are not too hard to do updates for.

2010-03-19, 12:43	#100
gd_barnes May 2007 Kansas; USA 24243₈ Posts	I'd like to suggest that we take a complete break from new bases for at least a week now. We've garnered a tremendous amount of new information over the last 1-2 weeks. I counted at least 30 new bases over about a 5 day span and many have complex algebraic factor situations that are taking an inordinate amount of time to properly analyze and add to the pages. After many hours last night, I still have at least 10 of them to add for the base 501-1024 thread and a few from the base 33-100 thread. I'd really like to get on to some other things like responding to my Email and PMs. Let's work on finishing the new bases that we have already reserved and then extending already searched bases to deeper depths. Thank you, Gary Last fiddled with by gd_barnes on 2010-03-19 at 12:47

2010-03-20, 10:26	#101
gd_barnes May 2007 Kansas; USA 101·103 Posts	We're currently in the process of a server/web pages move to a new machine. The noprimeleftbehind pages, which contain the CRUS pages will be down for a little while. The good news is that I've fully updated the pages for all work done through this posting. Once the pages come back up, they should be up to date. Edit: They came back up but they're not up to date. It's still the "old" pages from a day ago. We're experiencing some difficulty with the port forwarding related to the move. Gary Last fiddled with by gd_barnes on 2010-03-20 at 12:06

2010-03-23, 21:04	#104
gd_barnes May 2007 Kansas; USA 101·103 Posts	OK, guys. Feel free to unload on me again to a certain extent. That is, go ahead and test some new bases if you'd like. I'd only ask that we stick to no more than 1-2 new bases per person per day. That will help me not get too far behind when stuff comes up in my personal/business life for several days. The break from the web page updates helped me quite a bit in getting caught up here and with some other admin stuff at NPLB. Thanks for going easy on me for a few days. It also gave me time to "mentally" generalize the "new" kind of algebraic factors across all bases so that won't take me so long to show on the pages now. Although I want to get it down "in writing" in the algebraic factors thread, the analysis to generalize it across all bases is quite involved so it may be several days or possibly weeks before I get to that. In the mean time if they aren't clear to you, feel free to ask Serge, Ian, or myself if you have a new base with k's remaining that you think could be removed. In other news, late tonight, I'm going to start a recommended bases and efforts thread in an attempt to give a little more direction to the project so I know a little more of what to expect about what people are going to reserve or work on. There will be a mixture of new bases, 1k remaining bases, and bases that are somewhat behind their "neighbors (i.e. bases nearby with similar conjectured k)" as to how far they have been searched. That will help avoid some of the unexpected new bases load that seems to come around from time to time. Thank you to everyone for your immense interest in the project. I'd much rather be too busy with it than not busy at all. It's certainly turned out to be far more interesting math-wise than I would have ever originally expected when the project started. Gary Last fiddled with by gd_barnes on 2010-03-23 at 21:07

2010-04-21, 09:18	#106
gd_barnes May 2007 Kansas; USA 101·103 Posts	NPLB rally Hi all CRUS folks, How'd you like a change from the conjectures for at least a couple of days? We'll be running an LLRnet rally at "No Prime Left Behind" (NPLB) from April 30th to May 2nd. See details at http://www.mersenneforum.org/showthread.php?t=13334. For those who have never run an LLRnet rally, what we do is pile all of our machines on one server and see how many pairs we can process and primes we can find for 48 hours. We have both team and individual scores for pairs processed. Come join a team or be your own team. Either way, come join in the fun! The tests will be at n=~765K-775K base 2 and should take ~10-11 mins. each on a modern machine. Gary

2010-04-24, 09:28	#107
gd_barnes May 2007 Kansas; USA 101×103 Posts	Dropping bases > 512 After some contemplation today about what has transpired in the last few days/weeks/months on small-conjectured new bases and a PM that I received about someone wanting to do a large # of them on the Sierp side in the near future, I've decided to drop all bases > 512 from the CRUS project. Don't worry, existing work will not be wasted and I'm not "cutting everyone off" at this moment in time. What I will do within the next 2-3 weeks starting in about 10 days is I will split up the web pages into bases <= 512 and > 512. If others wish to manage and update the pages for bases > 512, they are welcome to copy the pages to their own website and do so. I will be glad to assist in this process. I also welcome someone creating a new project for bases > 512. For that matter, he can also extend it to as high of a base as he or everyone wants. Before doing that, obviously bases 250 to 512 will now have to be in one thread here with bases 513 to 1024 in another thread (vs. 250-500 & 501-1024). This will only amount to a little bit of tweaking and post moves. Then if someone wants to take bases > 512 into a different project or forum, we can get that thread moved to the new forum. (I guess I should have split them up in powers of 2 to begin with.) I'm sorry but it's just become way too much for me to manage and it's taking too much time away from my personal and business life. When Robert posted all of the conjectures up to base 1024, a few people did some higher bases here and there and I was perfectly willing to accomodate those people and decided to extend the project definition to include bases up to 1024. But now that the new bases script has made it so easy to start new bases, I'm willing to admit that I've bitten off way more than I can chew from an administrative perspective. Also, admittedly, I've grown to dread the web page updates whereas previously, it was my favorite thing to do, mainly because removing the k's and adding bigger primes was fun, especially for the bases where a final large prime is found to prove them. I had assumed that we would slowly work our way higher on the bases and I attempted to do that with the recommended bases/efforts thread, but it has gone largely ignored. For instance, at this time all bases <= 250 with a CK of < 10K are either at or nearly complete to n=25K. It is my thinking to keep slowly working our way higher on the bases in that manner...not just search only the small ones. I will do what I can to keep up with what people post, including for bases > 512 until such a time that I get things split apart. I am still not willing to accept anything more than one new base per day while I'm out of town so please respect that. At this moment, I won't be able to make any updates until at least Monday. I do appreciate everyone's enthusiasm for the project. I also appreciate people's offer to help with the HTML but ultimately, it's still too much to manage, because it's more than just updating the web pages. It's the checking, thread updates, asking for specifics when they aren't there, etc. Regardless, I'd much rather have it this way than the other way, with little to no activity. Thank you, Gary Last fiddled with by gd_barnes on 2010-04-24 at 09:39

2010-03-20, 13:00	#102
AMDave Jan 2006 deep in a while-loop 1010010010₂ Posts	Change completed URL tests for CRUS and NPLB appear to pass. AMDave

2010-03-20, 22:47	#103
gd_barnes May 2007 Kansas; USA 101×103 Posts	I confirmed that all web pages were correct and updated as of about 8 AM CDT (1 PM GMT) today. If anyone sees any reservations or ranges completed not properly reflected as of the above time, please let me know. I'm now working on just the few more that have come in since then. Gary

2010-04-20, 07:41	#105
gd_barnes May 2007 Kansas; USA 101·103 Posts	Hey all, I'll be on a business trip from Apr. 20th (later today) to Apr. 29th. If everyone can please limit their new base reservations to one per day during this time, I'd appreciate it. I'll be especially busy this coming weekend. To clarify: If you've already reserved the bases and they've been shown on the pages, post as many results/primes/k's remaining on multiple bases as you would like per day. Bases not already on the pages are what take a while. Thanks, Gary

2010-04-24, 10:19	#108
kar_bon Mar 2006 Germany 5533₈ Posts	I think it's not a good idea to split bases >512 into a new project or new thread here! All information should be available at one location to minimize irritations wher to search for a special info or posting people in the wrong place! Suggestion: Freeze all reservations for bases >512 for now and push bases with only searched to n=25k higher and so minimize the remaining pairs. Especially bases with only one k remaining should be done more than n=50k or 100k. It should also be manageable to write a script to automate the html-code for a base given with files of primes and info like remaining pairs, level tested. But as you posted, the most work is to verify all those results.

2010-04-24, 14:55	#110
mdettweiler A Sunny Moo Aug 2007 USA (GMT-5) 3×2,083 Posts	I'd be inclined to agree as well: freeze all reservations for new bases > 512, but keep the existing ones on the pages. Then, hopefully down the road when some sort of automation has been developed to make the process easier, we can reconsider opening them up again. In the meantime, though, there's no reason not to continue work on the already-started bases in that range, many of which have only one k remaining and which I'm guessing are not too hard to do updates for.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
News	gd_barnes	No Prime Left Behind	251	2021-02-15 03:00
P!=NP in the news	willmore	Computer Science & Computational Number Theory	48	2010-09-19 08:30
Other news	Cruelty	Riesel Prime Search	41	2010-03-08 18:46
The news giveth, the news taketh away...	NBtarheel_33	Hardware	17	2009-05-04 15:52
News	KEP	Riesel Base 3 Attack	4	2008-12-17 11:54