Less than 10,000 left....

[henryzz]

I think that one condition of being given preferred exponents should be that the prime95 client automatically moves it to the front of your work queue.

[retina]

Originally Posted by Kevin

If a computer is only making progress equivalent to 400GHZ, then yes, those computers shouldn't have been trusted with the low exponents.

Indeed I agree with this. Those overclocker folks are just taking it too far these days.

[Kevin]

Originally Posted by lfm

Well I would be very surprized if someone runs over the 1 year limit and for the first time in the history of the project is officially kicked off a unit while still reporting regular progress (it would seem to be a manual operation) does not subsequently have the exponent assigned (manually) to someone known, fast and reliable. What am I missing?

One thing is that the system technically cares about "regular updates" as opposed to "regular progress". You could keep shuffling an assignment to the bottom of your queue or be running it on a laptop you only use 1 hour every few days and still make "regular updates", but not be making "regular progress". The only thing I'm really talking about is if preferred exponents that are supposed to go to fast/trusted systems end up taking this long, which would indicate some type of loophole with/abuse of the system. In the old system, some people would horde hundreds of low assignments and have expected completion dates of more then 200 days. Some people not as knowledgeable about the assignment system might notice that their main desktop is getting smaller assignments, and use them to stock up the queue on an aging PII. Even if some of the current sub-33M exponents end up taking a year to finish, I don't think George would kick them off, but it's also not the kind of thing you necessarily want happening on a consistent basis.

Originally Posted by lfm

It is one of the reasons I like this project. It is inclusive of people with lesser computers. It doesn't demand you upgrade every couple years to keep contributing (but of course you can upgrade if you want). If you discourage anyone from helping even just to say "Hey why are you so slow?" it is liable to have repercussions for the project as a whole quite beyond the obvious. I for one would not want to see it.

As I said before, I'm not suggesting we somehow contact random users and ask them why they're so slow. As cheesehead suggested, maybe some kind of anonymous reporting could be built into the program. But with the current system, preferred exponents are supposed to go to people with a certain P4 GHZ level. If this isn't always happening, it would be nice to know why this is happening. The most obvious way to address this is (if it can be done in the proper manner) just ask. It was done previously in this thread between regular forum members without incident and I'm sure it could be done again without incident if the circumstances called for it.


This line of thread is really getting off track. At this point, all we really should be doing is watching the numbers to see if any irregularities arise. I really don't think there will be, I expect the assignment system will do a satisfactory job. But if it does turn out that the system doesn't work as well as some people would like, then we start debating whether or not the issue is severe enough to warrant action and how to properly proceed. But for right now, there's really no reason to keep debating these points.

[Kevin]

Originally Posted by henryzz

I think that one condition of being given preferred exponents should be that the prime95 client automatically moves it to the front of your work queue.

Maybe as a configurable option, but I don't think that would be a good idea in general. Lots of possible issues, not especially much gain for something that doesn't seem like much of a problem. Even if an assignment starts at the bottom of a 90 day work queue, then the person stops running their computer on weekends, and also takes a 1 month vacation, it should still take less than 6 months to return the result. That's not especially bad for a pretty extreme example, given that the general policy is that people can't be cut off until more than 12 months. I think the only instance this helps is if there's deliberate action on the users part to make things take exceptionally long, and the amount of restrictions you'd need to place to keep that from happening would hamper lots of legitimate queue shuffling.

[henryzz]

Originally Posted by Kevin

Maybe as a configurable option, but I don't think that would be a good idea in general. Lots of possible issues, not especially much gain for something that doesn't seem like much of a problem. Even if an assignment starts at the bottom of a 90 day work queue, then the person stops running their computer on weekends, and also takes a 1 month vacation, it should still take less than 6 months to return the result. That's not especially bad for a pretty extreme example, given that the general policy is that people can't be cut off until more than 12 months. I think the only instance this helps is if there's deliberate action on the users part to make things take exceptionally long, and the amount of restrictions you'd need to place to keep that from happening would hamper lots of legitimate queue shuffling.

some people have more than a 90 day work queue though
if the eta is over 90 days should someone be assigned a preferred exponent? IMHO no
it would also be nice to see reporting progress on v5

[cheesehead]

Originally Posted by storm5510

A friend of mine would call this a p**sing contest.

... because your friend would have as little knowledge of the background for my stance, going back years to the era when it was considered acceptable to "poach" assignments of other people, as you do. Educate yourself.

[Kevin]

Originally Posted by cheesehead

... because your friend would have as little knowledge of the background for my stance, going back years to the era when it was considered acceptable to "poach" assignments of other people, as you do. Educate yourself.

Just because some people poached assignments back in the day doesn't mean it was ever considered acceptable. You're not the only one that's been around for a while. Plus, the addition of the UID really makes your concern about poaching more or less irrelevant.

[cheesehead]

Originally Posted by Kevin

I was going off the fact that this thread exists,

Exactly.

You seem to have no appreciation of the historical basis for my opinions. You seem to presume that nothing happened before you were here, and that past objections by others did not exist simply because they do not appear in this thread about the subject of interfering with the assignments of others.

I have never seen anybody other than you support the same position you hold,

... which is probably simply a consequence of your inexperience in this forum. Have you bothered to read past threads where similar discussions/arguments took place?

especially as vociferously as you do.

My "vociferousness" is intended to deter those who might want to propose such things in the future (and who happen to read this thread before doing so), as well as you.

I'd be fine with something like that.

Then why don't you stop arguing that anything else is necessary?

But I think George would rather spend his time optimizing code and working on other things,

Well, of course you frame it as a work-saver for George (even though it's really substituting your desires for some of his goals).

so it's in the best interest of everybody if we do the leg-work first to see if such implements are even necessary.

Well, of course you frame your desire to "prod" slowpokes as a public service. So did all those others who proposed doing so in the past.

For the nth time, I was referring to contacting known users on the forum.

... and others not currently known, but who you want to be made known.

If all you wanted was to contact currently known users on the forum, you would have withdrawn your proposal to change anything because you can already do that now.

Quit referring back to the first post when I was vague after I have clarified what I was talking about multiple times since then.

You've refined your arguments, but have never repudiated your basic positions that (a) it's acceptable for you to claim as "your" own a portion of the project that hasn't been assigned to you, and (b) it's all right to try to hurry-up those whom you deem to be "too slow" without having ever shown that they do any harm to GIMPS. You've danced around my objections without ever directly and unambiguously declared that you no longer propose, or consider acceptable, doing those things.

- -

In post #220, I asked you

Originally Posted by cheesehead

Then why do you need more information on their progress? So someone can poach their assignments?

That wasn't an accusation; it was a question about your motive for wanting more information about users who weren't proceeding as fast as you liked.

You could have given a straightforward answer such as, "No, it's because such-and-such." But instead you responded with a counter-question in post #225:

Originally Posted by Kevin

Second of all, when did I say anything about < snip > or poaching exponents?

I responded to your question in post #227 with a direct answer:

Originally Posted by cheesehead

I never claimed you had mentioned poaching.

Now you say:

That seems like a pretty direct accusation, unless you want to hide behind the "rhetorical question" argument.

Maybe you need to re-read "I never claimed you had mentioned poaching." I never accused you of anything. I asked you something (which you've never answered).

If anything's rhetorical, it's your attempt to frame my question as an accusation instead of simply giving a simple straightforward answer to it.

- -

You can just look up who has what assignments on Primenet. I never suggested seeking more information than is already publicly available.

... except in post #208, where you posted, "it might be nice to have slightly more frequent updates (maybe every 1-3 weeks instead of every month). It would give a better idea of how quickly things are trailing off. Also, once there's fewer exponents left, we might want to start keeping track of them to possibly identify stragglers (and maybe prod them along", for instance.

That's more than once now that you've claimed that you "never" suggested something or "only" suggested something, when your words to the contrary are right up there where everyone can read them. If you no longer agree with some of what you previously wrote, why don't you say so straight out, explicitly and unambiguously, instead of pretending you didn't write those things?

Most people were able to see

... as it appears to you, because you haven't bother researching the past history of this subject ...

that there was an issue with lagging assignments in the previous system without any objective standard in place.

... so you use your own, subjective standard to decide what is "lagging" and what is not.

The real "issue" is that some people can't control their impatience and their desires to control what other participants are doing. GIMPS, the project, is suffering no damage!

You could make it anything more than 10 standard deviations past the average.

But why? GIMPS, the project, is suffering no damage!

You could set a time limit of 6 months instead of 12 months for the 100 lowest exponents remaining.

But why? GIMPS, the project, is suffering no damage!

I don't really know, we're not at that point in the discussion yet.

It seems to me that if GIMPS were actually suffering real damage from those "stragglers", you would know. But it's not, and so you don't.

At this point in time, all I said we might want to pay more careful attention, and made some loose suggestions for things to discuss in the future.

Why do you keep backing away from taking responsibility for your own previously-posted words, trying to minimize what you actually said, but without ever declaring that you have abandoned those positions, such as the "prod" business?

So what you're really doing is arguing against the system of having minimum P4 equivalences/reliability rankings/preferred exponents.

No, I'm not. Quit dodging. Everyone else here can see perfectly well that I'm criticizing your proposals and the attitudes behind them, not the assignment system, and you're trying over and over again to avoid taking responsibility for your own posted words.

You do realize that what you're arguing against is precisely the system that has been put in place,

No, I'm not, and you know it. Quit dodging, and have the courage to face what you yourself have written.

_I_ have proposed no change. _I_ am satisfied with the current checks and balances of the assignment system for handling assignments that are taking a long time. (Hint: It's ultimately the administrators, who, unlike you, have shown great patience, to decide what to do about "slow" progress. They are quite capable of handling any "stragglers" without your hurrying them up.)

It's _you_ who want to "prod" people beyond what is already done by the administrators.

The only thing I've done is propose that we check to make sure the system is working as intended,

It is working as intended! All the checking that's needed for that is already in place!

What you really propose is to substitute some of YOUR intentions for those of the project administrators because you're more impatient than they are.

and that assignments meant to go to people above a certain threshhold aren't being worked on by systems significantly below that threshhold.

Remind me again:

Exactly, precisely, what damage is done to GIMPS (not to your sense of impatience, but to GIMPS) by those systems?

Yes, it would be, which would be explain why nobody has suggested that except you.

Bull.

YOU wrote, "Also, imagine people who might make it their goal to clear out the lower end. It's a little disheartening if you spend a lot of time working on clearing everything beneath a certain range, and it takes an extra year to actually reach the goal because a few people take their time. There's never really a sense of accomplishment or moment of triumph for your wok when you get within arm's length of the goal, but things don't officially end for another few months."

See? You're talking about people who think it's all right not just to "make it their goal to clear out the lower end", but further to complain when "a few people" interfere with their desired goal.

You've never abandoned your position that it's all right for impatient folks to direct "slow" folks about what to do (or not do).

But even if we can't change the assignments that have already been made, we can change how assignments are made in the future,

See? Just a few sentences after claiming that it's _me_ who is "arguing against ... the system that has been put in place," it is _you_ who wants to change it.

Kevin, your own words keep tripping you up by contradicting your false claims!

and perhaps even change the standard of what is considered "acceptable progress" for certain sets of exponents.

See? You suggest changing the standard of what is considered acceptable progress (because you want to engineer some way to "prod" those whom you consider too-slow)!

[cheesehead]

Originally Posted by Kevin

Just because some people poached assignments back in the day doesn't mean it was ever considered acceptable.

... except, that is, for the folks who not only _did_ consider it acceptable, but even argued on the mailing list and early forum that it was desirable.

If you'd read those past postings, you'd know that.

You're not the only one that's been around for a while.

... but you, in particular, show no knowledge of the past history of this issue, so you're not one of those who has been around very long, are you?

Plus, the addition of the UID really makes your concern about poaching more or less irrelevant.

... except that the _attitudes_ underlying poaching are the same as you express to justify "prodding" folks you consider slow!!

Technical changes don't obsolete those attitudes, so I continue exposing and opposing them.

[retina]

Sorry to barge in on your debate here, but I am also quite a late comer to this GIMPS thing (since crica 2000) and I never really considered "poaching" to be a bad thing. I always just thought that the work has to be done anyway so the order of completion doesn't matter. But I would be interested to hear about the previous problems that any poaching has caused.

Although perhaps it is best to mention that I've never bothered to test "someone else's" exponent except for one time when I saw there was an exponent that had taken over a year to DC (something around 8M IIRC) and still not finished. I tested it locally. It took about a week to finish and came up as not prime. I never reported that result to GIMPS. And I realised later that I had actually just wasted a weeks worth of CPU time so I never got the desire to do it again.

[cheesehead]

Originally Posted by retina

But I would be interested to hear about the previous problems that any poaching has caused.

You've given your own answer about one of the problems poaching causes:

I realised later that I had actually just wasted a weeks worth of CPU time so I never got the desire to do it again.

Poaching wastes CPU time because it unnecessarily duplicates effort _outside the assignment system_.

An obvious follow-up is: What's the difference between what you did and an assigned double-check?

I won't take the time for a full answer now, but here's just a couple of things:

1) You've robbed the properly-assigned user his/her legitimate chance of being a Mersenne prime discoverer!

Think about what happens, from the assignee's point of view, if you poach someone else's assignment, and happen to discover, before they do, that the exponent you've poached is that of a Mersenne prime.

2) One of the reasons for having an assignment system is to make the search for Mersenne primes efficient. By violating the rules of the assignment system, you reduce that efficiency.