mersenneforum.org Working on the 28,000,000-30,000,000 range
 Register FAQ Search Today's Posts Mark Forums Read

 2013-12-25, 12:58 #34 tha     Dec 2002 5·173 Posts I am not so sure you would need to worry. First these four assignments. They are likely to become the lowest exponents in the weeks to come. 27897511 27906397 27988427 27988483 They have been worked on for a long time, more than a year. They are 90-95% done and move up by a few tenth of a percent per month. But they do report reliable progress. So almost all indicators make these assignments a no-no for taking since it really would be poaching. I tried to find another assignment below 30,000,000 that would meet the criteria spelled out in this thread but they all have already been poached or taken long time ago by others. CaptainEntropy is still working on some exponents in this range with no progress reported in the last 4 years. They would meet the criteria if they would become a block to a major milestone, but other GIMPS participants have completed them already. I agree that poaching should be prevented as much as possible, but I also think the criteria spelled out here make poaching very unlikely, definitely less likely than criteria actively used by others.
 2013-12-26, 03:59 #35 petrw1 1976 Toyota Corona years forever!     "Wayne" Nov 2006 Saskatchewan, Canada 10100100111012 Posts I'll take these next 8 ummm 10?. Code: 28465331 - Woody 28471337 - Woody 28474087 - Hand_In_The_Box 28476733 - Hand_In_The_Box 28485169 - Hand_In_The_Box 28489207 - Hand_In_The_Box 28491019 - CKDO still has this one 28491101 - CKDO still has this one 28497499 - Speck 28513531 - Speck "Sid & Andy" will be the owner Last fiddled with by petrw1 on 2013-12-26 at 04:41 Reason: Added Computer Names for my reference
 2013-12-26, 11:10 #36 tha     Dec 2002 5×173 Posts New overview of the taken assignments: Code: 28258819 * Tha 28759477 28262609 * Tha 28779329 28339513 * Tha 28784731 28351957 * Tha 28786301 28404611 * Tha 28809983 28425883 * Tha 28813847 28452059 * Tha 28818913 28818919 28465331 * Sid & Andy 28819093 28471337 * Sid & Andy 28819289 28474087 * Sid & Andy 28821313 28476733 * Sid & Andy 28824689 28485169 * Sid & Andy 28842977 28489207 * Sid & Andy 28861241 28491019 * Sid & Andy 28864109 28491101 * Sid & Andy 28869107 28497499 * Sid & Andy 28875697 28513531 * Sid & Andy 28875751 28514449 28876157 28514693 28876307 28519753 28876597 28531051 28877843 28531057 28879159 28533929 28881109 28535261 28887407 28539479 28888129 28544231 28888261 28546901 28890131 28550267 28890751 28550311 28891537 28550909 * Miszka 28892641 28563097 28894681 28566569 28897207 28573081 28898743 28574713 28899917 28579189 28900217 28579757 28900997 * Miszka 28588337 28939003 28589321 28944551 28592209 28968389 28598201 28970779 28602817 28975501 28611707 28976737 28612691 28977737 28613119 28980199 28617217 28980229 28622617 28981999 28627297 28986851 28655009 28990187 28658687 28995443 28663183 29029769 28663433 29031329 28668589 29031503 28679789 29031553 28693109 29126737 28730941 29127017 28751323 29127349 28757419 29204239 28759363 29206651
2013-12-26, 18:59   #37
chalsall
If I May

"Chris Halsall"
Sep 2002

11,087 Posts

Quote:
 Originally Posted by Brian-E Not really wanting to argue with anyone, and this will likely be my only post in this thread, but I'd like it on record that I feel quite threatened in my continuing tiny contribution to GIMPS by this sort of poaching. By the criteria indicated by a few in this thread, my slow but reliable DC work could easily be pinched from me before it reports.
I hear what you are saying.

But, at the same time, I think tha has brought into "the crucible" an important issue. And I believe he is trying to be fair in his approach, unlike several other active "poachers" who still submit results to GIMPS.

Every participant's cycles can important and useful, and no-one should feel that just because they don't have a lot of "fire-power" that they cannot contribute to GIMPS.

On the other hand, there are a few people who "don't play nice with others". Some reserve way more "preferred" candidates than they can process in a reasonable time. Other's reserve candidates in the middle of the "wave", and then "sit" on them for (literally) years until they become "milestone" blocking.

I would argue this is actually a "meta-bug" of GIMPS. As an example, why should "Captain Entropy" be able to hold up an assignment for TFing for more than three years?

Why should a Prime95 participant's machine be considered "trusted" and given the lowest available candidate at the moment of request just because it has "high confidence" -- shouldn't the historical through-put of the particular machine in relation to the candidate also enter the heuristics?

For the record:

1. I've spent the last three hours cleaning up the mess in the GPU72 database this causes by those who get such assignments from GPU72, since it warns users if the work they've been assigned has been completed, and suggests they stop the said work and unassign it from the GPU72 personal assignment report.

1.1. At least one user ignores this notice, and just keeps reserving new assignments. This causes work for me, personally.

1.2. This user is no longer able to reserve LL or DC work from GPU72.

2. GPU72 used to re-capture work which was assigned by it to one of our users but abandoned, and hold it for reassignment back to said user.

2.1. This will no longer be done. If a preferred LL or DC assignment is abandoned and recaptured, I personally will complete the assignment (and thus personally take the risk that my work will be useless if the user magically completes the work in the time it takes me to do it).

2.2. For those who get their LL or DC assignments through the GPU72 manual assignment page, please *ensure* you formally "claim it".

2.2.1. This ensures everyone knows who has ownership, and that you'll have 60 days to complete the assignment before it is recycled.

2.2.2. Rich et al... For your situation (not being able to have Prime95 talk to Primenet or the Proxy from work), simply copy your worktodo.txt file onto a memory stick or e-mail it yourself. Then when you're home place the assignments into a "slaved" Prime95 instance with your Primenet credentials and have it "call home".

P.S. We all *really* need to get out more....

 2013-12-27, 12:12 #38 Miszka     May 2013 Poland 10610 Posts LL test successfully completes double-check of M28550909
 2013-12-27, 17:32 #39 Miszka     May 2013 Poland 2×53 Posts I'll take these two exponents: 28514449 28514693
2013-12-27, 17:36   #40
chalsall
If I May

"Chris Halsall"
Sep 2002

255178 Posts

Quote:
 Originally Posted by Miszka I'll take these two exponents: 28514449 28514693
I would argue that, based on tha's proposed criteria above, the latter is reasonable, but the former isn't.

2013-12-27, 17:54   #41
Miszka

May 2013
Poland

2×53 Posts

Quote:
 Originally Posted by chalsall I would argue that, based on tha's proposed criteria above, the latter is reasonable, but the former isn't.
O.K. I'll take these two exponents (assigned to ANONYMOUS):
28514693
28531057

because
28514449
28519753
28531051
are assigned to "Carsten Kossendey"

Last fiddled with by Miszka on 2013-12-27 at 18:01

2013-12-27, 18:03   #42
chalsall
If I May

"Chris Halsall"
Sep 2002

11,087 Posts

Quote:
 Originally Posted by Miszka O.K. I'll take these two exponents (assigned to ANONYMOUS): 28514693 28531057
These two candidates satisfy the proposal's selection criteria.

 2013-12-27, 18:13 #43 tha     Dec 2002 5×173 Posts Hmm, I get a very uneasy feeling about this. The user that originally had all these assignments and did nothing with them for an extreme long time was asked to give up the assignments. The exponents were put back into the pool. Then we find out that he has taken them again, but now as ANONYMOUS. He also has not logged into this forum, at least not using his own account, since the exponents were released back into the pool. Again we see nothing happening to these exponents. Then we spell out some criteria in the forum. The next thing that happens, said user makes a minimal change to a bunch of these assignments (adding his old username again and posting a CPU name), even going as far as having one exponent reporting 0.1% of the work done. Weird, if that CPU can do 0.1% of the work on one exponent in so much time, then why have so many exponents reserved? Why does he not want to communicate, why does he suddenly shy away from contact in lieu of many previous postings? The list I posted was carefully composed and the actions taken since by said user are (very) suspect to me and look like they are designed to obstruct rather than contribute. I stick to the assignments I have taken, and I consider all the exponents on the list to be kosher. Last fiddled with by tha on 2013-12-27 at 18:40 Reason: grammar
 2013-12-27, 18:59 #44 tha     Dec 2002 5×173 Posts The machine "tycho" appears to be a dual core machine that claims to do a single exponent in 11 days. I stick to the work I've started and have progressed on. But since there are still enough assignments available that meet the original criteria I'll add a mark to those exponents that are claimed by "tycho". And we will keep a watch on them. Last fiddled with by tha on 2013-12-27 at 19:00

 Similar Threads Thread Thread Starter Forum Replies Last Post Dubslow Linux 3 2012-05-11 14:44 cheesehead Forum Feedback 1 2012-03-19 17:13 Dubslow Forum Feedback 2 2012-03-19 06:53 delta_t Marin's Mersenne-aries 15 2004-09-13 15:27 garo Lone Mersenne Hunters 2 2003-03-28 11:13

All times are UTC. The time now is 11:08.

Sun Feb 5 11:08:37 UTC 2023 up 171 days, 8:37, 1 user, load averages: 1.14, 0.95, 0.90