Newer milestone thread

[cheesehead]

Originally Posted by cheesehead

I think that's a well-written message. In particular, the second paragraph quoted above avoids putting (what I'd consider to be) too much pressure on the recipient to "hurry up". I'm satisfied with its balance.

I withdraw that statement. After due consideration of the ethical issues, I'm no longer satisfied with its balance, or indeed its use at all.

Added: Just to be clear, my personal preference is to send no such message. I think the above suggestion strikes a reasonable balance between my preference and those of others.

I withdraw the second sentence. I no longer consider the suggestion of any "prodding" message to be ethically acceptable.

[chalsall]

Originally Posted by cheesehead

You've misconstrued the "100 vs. 100.000001" argument. That argument has nothing to do with whether the exponent is tested twice.

It does have to do with violating GIMPS's pledge of exclusive assignments (in license.txt, I remind you) -- which does apply in this particular case.

Sigh...

Cheesehead... You appear to have a lot of time on your hands.

Might you consider studying some law theory before trying to make claims such as a there exists a binding agreement by all GIMPS participants of a "pledge of exclusive assignments" based on the language in license.txt?

As has been said before, the language in license.txt is "ass covering" for George et al.

The language doesn't actually restrict anyone from doing anything they want, so long as George et al get credit for any new MPs found using the software.

Get over it.

[MooMoo2]

Forget poaching, prodding, or bugging. The best (and least controversial) option would just be a reminder. The letter doesn't have to be so long; a simple one like this will be fine:

Dear User,

Thank you for contributing your spare computer time to the GIMPS project. Every particpant's work is appreciated, but we'd like to let you know that you're a special part of the project's history. This is because one of the Mersenne numbers you are testing, namely M31494937, has become the last open first-time testing assignment below the Mersenne primes M32582657 and M37156667. Therefore, once M31494937 has had a completed first LL test, there will be no remaining candidates needing a first-time test below what are possibly the 44th and 45th Mersenne primes, respectively.

You can complete the test at your own pace; we don't mind if you take a month, a year, or even a decade to finish it. But the day you complete the test will be a day that will forever be remembered here:
http://www.mersenne.org/report_milestones/

Once again, thank you for your time and for your contributions to GIMPS.

Best Regards,
The Participants of GIMPS

[Uncwilly]

Originally Posted by chalsall

Would it be the end of the world if someone with a really fast machine "poached" this exponent? When the contributor in question returned their result they would still receive the same GHzHours credit (assuming the residual matched) as a DC.

That is what I hinted might happen. I was not suggesting, encouraging, or condoning it.

Originally Posted by Uncwilly

I suspect that there will be a real early double check of that number completed soon enough.

[NBtarheel_33]

Originally Posted by MooMoo2

Forget poaching, prodding, or bugging. The best (and least controversial) option would just be a reminder. The letter doesn't have to be so long; a simple one like this will be fine:

Dear User,

Thank you for contributing your spare computer time to the GIMPS project. Every particpant's work is appreciated, but we'd like to let you know that you're a special part of the project's history. This is because one of the Mersenne numbers you are testing, namely M31494937, has become the last open first-time testing assignment below the Mersenne primes M32582657 and M37156667. Therefore, once M31494937 has had a completed first LL test, there will be no remaining candidates needing a first-time test below what are possibly the 44th and 45th Mersenne primes, respectively.

You can complete the test at your own pace; we don't mind if you take a month, a year, or even a decade to finish it. But the day you complete the test will be a day that will forever be remembered here:
http://www.mersenne.org/report_milestones/

Once again, thank you for your time and for your contributions to GIMPS.

Best Regards,
The Participants of GIMPS

Even better than mine. In other words, we're saying that nothing but good can come out of accelerating the test.

[davieddy]

Originally Posted by NBtarheel_33

No one is actually sitting around (are they?) on pins and needles, eyes bloodshot from staring at the screen, as they wait for the milestone LL to come in, are they? (If so, they may want to take a GIMPS vacation and consult professional assistance...)

That point is beautiffuly phrased, and in my case rather too close
to the mark for comfort.

I often have two instances of the Primenet summary on the go,
updating the oldest one hourly, and noting the progress. "With interest"
would be phrasing it a bit too strongly

I will probably miss the most conspicuous feature of the table
more than most when it goes: that 1 in the 31M row.

There are other things I do on the internet, as Brian will kindly
refrain from telling you about. (Publically at least).

David

[cheesehead]

Originally Posted by chalsall

Sigh...

Cheesehead... You appear to have a lot of time on your hands.

You'd like it if I ceased exposing the flaws in the right-to-interfere side, wouldn't you?

I'll stop refuting the right-to-interfere arguments when the right-to-interfere proponents stop posting them.

Might you consider studying some law theory before trying to make claims such as a there exists a binding agreement by all GIMPS participants of a "pledge of exclusive assignments" based on the language in license.txt?

Might you consider ceasing to use rhetorical trickery such as this straw-man introduction of "law theory" and "binding agreement" [implying legally-binding agreement]?

By escalating what I actually wrote and shifting it from one basis (ethics) to another (law), you find it easier to pretend that what I wrote is refutable than you would if you were to stick to the non-exaggerated, non-shifted (ethics) version of what I posted. Use of straw-man exaggeration is consistent with lack of confidence in your ability to win the argument without rhetorical trickery.

As has been said before, the language in license.txt is "ass covering" for George et al.

By pretending that I was approaching it in a legal manner, you avoid confronting that my approach is ethical, rather than legalistic. All your "legal" claims are just attempts to sidestep and ignore what my postings really said.

How about meeting my ethical arguments honestly and fairly, without dodging, without pretending that mine was a legal argument? Can you face that square-on in this discussion?

[Mr. P-1]

Originally Posted by cheesehead

In my v25.8 distribution of prime95, the key sentence for this discussion is:

"Every effort has been made to ensure that you will be assigned an exponent that no one else has tested...

This is no longer true. Current practice is to rerelease exponents with one suspect LL for "first-time" testing again. In some cases the suspect LL turns out to be good.

...or is testing."

But there is no claim that nobody will test that exponent in the future.

IOW, GIMPS officially pledges that it intends assignments to be exclusive.

Sure. Assignments are exclusive. We are not talking about exponents being assigned to more than one participant, but about exponents being tested by individuals to whom they are not assigned.

That has to include not condoning, and indeed making effort to discourage, poaching after the assignment is made, or else the pledge is meaningless.

No. The pledge is grant exclusive assignments. There is no pledge, nor could there be, that nobody will test an exponent which they have not been assigned.

You can argue that there is, or should be, a community standard that participants do not poach. I take no issue at that. All I'm arguing is there is no formal rule to that effect.

[ATH]

31494937 LL LL, 92.90% 760 137 2011-05-01 2010-12-26 2010-11-28 2008-11-15 alaendle P3-900

This assignment is already 760 days old and with current progress it won't be done for another 6-12 months. According to George own suggestion for recycling exponents, this should have already been recycled, but I guess those rules are not implemented yet:

http://www.mersenneforum.org/showpos...&postcount=443

My new ideas for recycling assignments:

For LL tests on exponents < 80M and not manual testing:

If assignment is one year old and < 50% complete it is recycled.
If assignment is one-and-one-half years old it is recycled.

For non-LLs:

If assignment is 180 days old and < 50% complete it is recycled
If assignment is 270 days old it is recycled

[davar55]

I still agree with Ch....H..ed.
Liked that letter, though.

[cheesehead]

Originally Posted by Mr. P-1

This is no longer true. Current practice is ...

You're trying to dodge the ethical intent by separating out details. George's statement isn't perfect. But the intention of the PrimeNet assignment system is as I stated.

Originally Posted by Mr. P-1

This is no longer true. Current practice is to rerelease exponents with one suspect LL for "first-time" testing again. In some cases the suspect LL turns out to be good.

So, channel this into a suggestion for more precise wording.

But there is no claim that nobody will test that exponent in the future.

Straw man. The claim is that GIMPS makes an effort (within its powers, obviously) to assure that assignments are exclusive, not that "nobody will test that exponent in the future". I've never claimed otherwise, so this pretended protest at a misinterpretation is just a straw man.

Why are you introducing distortions like that? Don't you think you can win the argument by sticking to honest, nondistorting statements meeting my arguments head-on, instead of using rhetorical trickery to dodge?

Sure. Assignments are exclusive. We are not talking about exponents being assigned to more than one participant

Of course not ... and you know very well that's not what I'm addressing -- so why do you bring it up, other than to distract your audience?

What I'm addressing is the situation in which someone with an assignment is proceeding properly, but other folks think they have some right to interfere by either poaching or prodding that assignee to "hurry up".

, but about exponents being tested by individuals to whom they are not assigned.

Yes, that is what I'm addressing -- exponents being tested by individuals to whom they are not assigned and know that very well, because they are intentionally targeting the assignee for interference.

The pledge is grant exclusive assignments.

Yes, and in order to make "exclusive assignment" more than an empty phrase, the pledge implies a reasonable effort to dissuade other people from testing the same exponent while that assignment is in effect. Without that follow-on, the assignment process wouldn't mean anything.

There is no pledge, nor could there be, that nobody will test an exponent which they have not been assigned.

Straw man.

Why do you keep denying something that needs no denial? By interspersing such straw men among the statements you address to me, you give the impression that they address, and oppose, something I actually said. That attempted impression is false and misleading. You could avoid imputing such falsities to me if you wanted to do.

The way you keep using rhetorical trickery to distort what I said and mislead your audience indicates that you lack confidence that you could win the discussion with straightforward, honest, non-distorting, non-misleading, non-tricking statements.

All I'm arguing is there is no formal rule to that effect.

Then why are you throwing in all the rhetorical devices, if that's really "all" you're arguing?

Clearly, that's not "all" you're arguing. Please either admit it or cease it.