Poaching - Page 5

Xyzzy · 2003-01-22, 07:58

No censoring is being done in this forum... Possibly it could happen in one of the other forums but I imagine the mod would indicate if a change was made...

Vulgar language is not encouraged though... Remember, there are kids that read this forum...

Worst case scenario I can turn on the bad-word filter thingie, but I'd really like to avoid that...

Xyzzy · 2003-01-22, 08:00

Quote:

Originally Posted by cheesehead

What is "LLR"?

The best I can figure it is this...

http://www.utm.edu/research/primes/bios/titans/LLR.html

eepiccolo · 2003-01-22, 14:18

Wow, I didn't realize some people were so sensitive

. C-c-c-can't we all jus' get along

:D

Kevin · 2003-01-22, 18:25

One thing we could do to help out the older, slower machines would be to break up factoring into smaller pieces. We could give machines with <300 CMHZ (corrected MHZ, accounting for hours in a day it's on) the bit ranges 57-64, then have any P4's on factoring finish whats left. This would give slower machines more managable work units. Also, if a slow machine checks "Get first-time/double check", we could pop up a menu similar to that for 33M that says "Assuming your computer is running xx hours of the day, it will take this long to finish. We would like to be done in no more than yy amount of time. If you don't feel you will be able to complete your exponent in a timely fashion, please continue with factoring work." This would work even better if we made sure that these slower computers got leading-edge exponents instead of the stragglers. Then we could have less aggressive requests for how long they have to work on it, so in general less people will complain about the deadlines.

Xyzzy: I don't know about the other whipper-snappers around here, but I've heard (and used) my fair share of colorful language. Don't feel like you have to censor on my behalf or anyone else's until you're asked to.

nomadicus · 2003-01-22, 20:25

Quote:

Originally Posted by binarydigits

Nomadicus: the exponents being poached are in the 2 million digit range and thus not eligible for the $100000 prize, so it is not a factor at all.

Only the 2M range? If that is the case, seems to me a poacher is doing someone a favor if we can give the original person the credit, ignore the poacher -- except for their work ;) , thus the original person gets the added benefit of moving on to the next M much sooner then otherwise.

Prime95 · 2003-01-22, 20:35

Quote:

Originally Posted by nomadicus

Only the 2M range? If that is the case, seems to me a poacher is doing someone a favor if we can give the original person the credit, ignore the poacher -- except for their work ;) , thus the original person gets the added benefit of moving on to the next M much sooner then otherwise.

This assumes the only motivation for the poachee is CPU credit. Much more likely he is 1) peeved because he was robbed of his chance to find a new prime, 2) peeved because he has literally wasted months of computing time

nomadicus · 2003-01-22, 20:52

Quote:

Originally Posted by Prime95

This assumes the only motivation for the poachee is CPU credit. Much more likely he is 1) peeved because he was robbed of his chance to find a new prime, 2) peeved because he has literally wasted months of computing time

The poachee would not have wasted months of computing time. I am thinking that the poachee would get all the credit (CPU as well as fame), and that the poacher would be cut off at the knees. The poacher having ill intentions, would give the poachee a boost. If we could achieve that, then poaching would be a non-issue and a deterent. Within GIMPS, is this not possible or am I in outer space?

cheesehead · 2003-01-22, 20:59

Well ... back to my questions:

_IF_ PrimeNet has automatic time limits on assignments, ordinarily requiring no manual intervention to expire assignments or re-assign them, then why would any GIMPS participant, other than a system administrator or a would-be poacher, need to know someone else's:

(a) current iteration,

(b) days-to-go,

(c) days-to-expire, or

(d) last date-updated?

If there's no non-poaching non-administrating user's need-to-know for those items, then just stop including them in public reports. Include them only in administrative reports and private password-requiring individual reports.

That would deny target-selecting information to would-be poachers, right?

I'm not claiming this would stop poaching, just that it denies poachers the information they currently use (I presume) to select straggling assignments to poach. A poacher could still target the assignments of userids that had previously been slow, but then runs a greater risk of poaching a LL that's actually on-schedule, perhaps even one that's running on that user's freshly-upgraded faster system the poacher doesn't know about. Then the would-be poacher might need to consider that his/her efforts would (a) fail to "help" achieve milestones, and (b) more obviously interfere with GIMPS's orderly progress. (We could help the denser poachers figure this out by publishing this reasoning. :) )

- - -

Indeed, couldn't my suggestion be done NOW?

Make the current assignments report password-protected, then substitute a new public assignments report that omits the above four items.

Do the system administrators currently need the eyes and attention of others to detect stragglers _about whom action needs to be taken_? If not, then why provide this information to poachers? If so, just give some other trusted individuals the password for the full assignments report.

Does anyone see any problem with this scheme? (... other than that some individuals' curiosities would go unsatisfied?)

cheesehead · 2003-01-22, 21:11

(I moved the comments I originally made here into my preceding post.)

Hmmm ... I'll ask this on the mailing list, too.

QuintLeo · 2003-01-22, 22:26

I don't see any reason for anyone (except perhaps George - it *is* his project) to need access to the a-d items Cheesehead listed for other folks. I don't know if having that info unavailable would eliminate poaching, but I suspect it would reduce it quite a bit.

cheesehead · 2003-01-23, 00:24

I'd suggest eliminating days run and date assigned also, but:

A) We need some way to monitor for runaway assignments which probably ought to be available to anyone, not just the trusted few, and date-assigned seems fairly useful, if not absolutely necessary, for that.

B) If date-assigned is public, then the easily-derivable days-run may as well be used for report-column filler since the elimination of all those other fields is going to make the public report downright narrow.

So, am I right about the retention of those two fields? Or can someone demonstrate their nonnecessity for the public assignments report?

2003-01-22, 20:59	#52
cheesehead "Richard B. Woods" Aug 2002 Wisconsin USA 2²×3×641 Posts	Reporting requirements if there are automatic time limits Well ... back to my questions: _IF_ PrimeNet has automatic time limits on assignments, ordinarily requiring no manual intervention to expire assignments or re-assign them, then why would any GIMPS participant, other than a system administrator or a would-be poacher, need to know someone else's: (a) current iteration, (b) days-to-go, (c) days-to-expire, or (d) last date-updated? If there's no non-poaching non-administrating user's need-to-know for those items, then just stop including them in public reports. Include them only in administrative reports and private password-requiring individual reports. That would deny target-selecting information to would-be poachers, right? I'm not claiming this would stop poaching, just that it denies poachers the information they currently use (I presume) to select straggling assignments to poach. A poacher could still target the assignments of userids that had previously been slow, but then runs a greater risk of poaching a LL that's actually on-schedule, perhaps even one that's running on that user's freshly-upgraded faster system the poacher doesn't know about. Then the would-be poacher might need to consider that his/her efforts would (a) fail to "help" achieve milestones, and (b) more obviously interfere with GIMPS's orderly progress. (We could help the denser poachers figure this out by publishing this reasoning. :) ) - - - Indeed, couldn't my suggestion be done NOW? Make the current assignments report password-protected, then substitute a new public assignments report that omits the above four items. Do the system administrators currently need the eyes and attention of others to detect stragglers _about whom action needs to be taken_? If not, then why provide this information to poachers? If so, just give some other trusted individuals the password for the full assignments report. Does anyone see any problem with this scheme? (... other than that some individuals' curiosities would go unsatisfied?)

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Poaching	blip	Data	8	2016-01-30 01:59
Poaching	davieddy	Lounge	6	2010-10-16 12:31
Poaching and v5	PrimeCruncher	PrimeNet	6	2004-04-05 19:17
Officially poaching very old exponents	Prime95	Data	17	2003-11-13 02:13
New fashion poaching (???)	lycorn	Lounge	6	2003-01-31 08:33

2003-01-22, 07:58	#45
Xyzzy "Mike" Aug 2002 8224₁₀ Posts	No censoring is being done in this forum... Possibly it could happen in one of the other forums but I imagine the mod would indicate if a change was made... Vulgar language is not encouraged though... Remember, there are kids that read this forum... Worst case scenario I can turn on the bad-word filter thingie, but I'd really like to avoid that...

2003-01-22, 14:18	#47
eepiccolo Dec 2002 Frederick County, MD 2×5×37 Posts	Wow, I didn't realize some people were so sensitive . C-c-c-can't we all jus' get along :D

2003-01-22, 18:25	#48
Kevin Aug 2002 Ann Arbor, MI 433 Posts	One thing we could do to help out the older, slower machines would be to break up factoring into smaller pieces. We could give machines with <300 CMHZ (corrected MHZ, accounting for hours in a day it's on) the bit ranges 57-64, then have any P4's on factoring finish whats left. This would give slower machines more managable work units. Also, if a slow machine checks "Get first-time/double check", we could pop up a menu similar to that for 33M that says "Assuming your computer is running xx hours of the day, it will take this long to finish. We would like to be done in no more than yy amount of time. If you don't feel you will be able to complete your exponent in a timely fashion, please continue with factoring work." This would work even better if we made sure that these slower computers got leading-edge exponents instead of the stragglers. Then we could have less aggressive requests for how long they have to work on it, so in general less people will complain about the deadlines. Xyzzy: I don't know about the other whipper-snappers around here, but I've heard (and used) my fair share of colorful language. Don't feel like you have to censor on my behalf or anyone else's until you're asked to.

2003-01-22, 21:11	#53
cheesehead "Richard B. Woods" Aug 2002 Wisconsin USA 2²×3×641 Posts	(I moved the comments I originally made here into my preceding post.) Hmmm ... I'll ask this on the mailing list, too.

2003-01-22, 22:26	#54
QuintLeo Oct 2002 Lost in the hills of Iowa 2⁶·7 Posts	I don't see any reason for anyone (except perhaps George - it is his project) to need access to the a-d items Cheesehead listed for other folks. I don't know if having that info unavailable would eliminate poaching, but I suspect it would reduce it quite a bit.

2003-01-23, 00:24	#55
cheesehead "Richard B. Woods" Aug 2002 Wisconsin USA 1111000001100₂ Posts	I'd suggest eliminating days run and date assigned also, but: A) We need some way to monitor for runaway assignments which probably ought to be available to anyone, not just the trusted few, and date-assigned seems fairly useful, if not absolutely necessary, for that. B) If date-assigned is public, then the easily-derivable days-run may as well be used for report-column filler since the elimination of all those other fields is going to make the public report downright narrow. So, am I right about the retention of those two fields? Or can someone demonstrate their nonnecessity for the public assignments report?