Validator and Quorum
Message boards : Number crunching : Validator and Quorum
| Author | Message | |
|---|---|---|
|
Seems to be a little problem with validation.
|
||
ID:
253 | Rating: 0
| rate:
/
|
||
|
This begs another question.
|
||
|
ID:
254 | Rating: 0
| rate:
/
|
||
|
This begs another question. There is a quorum of 2 at the moment, if two 2 results back the lowest value get credits for all. The 3rd result get the same amount of credits as 2 others. |
||
|
ID:
256 | Rating: 0
| rate:
/
|
||
|
This begs another question. I think awarding the lowest value to all will be generating a big credit uproar. If you're going to use a quorum, make it an average. |
||
|
ID:
289 | Rating: 0
| rate:
/
|
||
|
Is the minimum quorum is now 3? Do I understand that correctly?
|
||
|
ID:
533 | Rating: 0
| rate:
/
|
||
|
At the moment we are trying to understand how the system works in order to try to come up with the best set up possible. In this process testers play a big role and of course this seems like something logical. I will discuss this with the team.
|
||
|
ID:
536 | Rating: 0
| rate:
/
|
||
Andre Kerstens
Forum moderator Project tester Volunteer tester 
Joined: Sep 11 06 Posts: 749 ID: 1 Credit: 15,199 RAC: 0 |
||
|
Yes, this was a problem we had with app version 5.1 where errored results were validated successfully, while correct results were invalidated. That problem was fixed in 5.2, but there might still be some 5.1 results around (although we have asked people to abort all of these).
Seems to be a little problem with validation. ____________ D@H the greatest project in the world... a while from now! |
||
|
ID:
538 | Rating: 0
| rate:
/
|
||
|
What happened here:
[*]12746 130 18 Sep 2006 6:21:51 UTC 18 Sep 2006 7:23:24 UTC Over Success Done 207.84 1.05 pending [*]12747 217 18 Sep 2006 22:59:36 UTC 18 Sep 2006 23:26:23 UTC Over Client error Compute error 189.42 0.37 --- [*]12748 233 18 Sep 2006 23:40:12 UTC 19 Sep 2006 3:32:59 UTC Over Success Done 5,631.64 9.70 pending [*]20647 103 22 Sep 2006 13:31:40 UTC 23 Sep 2006 9:51:43 UTC Over Success Done 7,256.46 13.77 pending [*]21827 207 23 Sep 2006 10:04:42 UTC 23 Sep 2006 10:06:19 UTC Over Client error Compute error 0.00 0.00 --- [*]21828 137 23 Sep 2006 10:20:01 UTC 23 Sep 2006 14:44:43 UTC Over Success Done 5,082.33 9.75 pending [*]21830 190 24 Sep 2006 3:16:35 UTC 24 Sep 2006 16:43:10 UTC Over Success Done 5,750.20 9.90 pending [*]21940 --- --- --- Unsent Unknown New --- --- ---
|
||
|
ID:
618 | Rating: 0
| rate:
/
|
||
|
I'm not sure what's going on here. I'll see if we can figure out why these are not reaching consensus. Thanks.
What's wrong with the success ones to reach no consensus? |
||
|
ID:
619 | Rating: 0
| rate:
/
|
||
|
Andre Kerstens
Forum moderator Project tester Volunteer tester
Joined: Sep 11 06 Posts: 749 ID: 1 Credit: 15,199 RAC: 0 |
||
|
Currently on vacation, but got a wireless connection in a hotel somewhere :-)
What happened here: ____________ D@H the greatest project in the world... a while from now! |
||
|
ID:
631 | Rating: 0
| rate:
/
|
||
|
Everybody who is still crunching 5.1 results, please abort these as soon as possible!! 5.01 is the only application available to Macs. ____________ Dublin, CA Team SETI.USA |
||
|
ID:
638 | Rating: 0
| rate:
/
|
||
|
Andre Kerstens
Forum moderator Project tester Volunteer tester
Joined: Sep 11 06 Posts: 749 ID: 1 Credit: 15,199 RAC: 0 |
||
|
Sorry, this was only meant for Linux users, not for Macs. Macs seem to do fine with 5.1 for the moment (but we'll release 5.2 anyway as soon as we get our compiler problem solved). You see 501 in the boinc client, but this is really 5.1. I don't know why the Boinc client puts a 0 in the version number.
Everybody who is still crunching 5.1 results, please abort these as soon as possible!! ____________ D@H the greatest project in the world... a while from now! |
||
|
ID:
668 | Rating: 0
| rate:
/
|
||
|
Sorry, this was only meant for Linux users, not for Macs. Macs seem to do fine with 5.1 for the moment (but we'll release 5.2 anyway as soon as we get our compiler problem solved). Intel Macs do fine yes, but many of the valid results get 0 credits (just so everyone is aware). ____________ Dublin, CA Team SETI.USA |
||
|
ID:
671 | Rating: 0
| rate:
/
|
||
|
Andre Kerstens
Forum moderator Project tester Volunteer tester
Joined: Sep 11 06 Posts: 749 ID: 1 Credit: 15,199 RAC: 0 |
||
|
I've looked at one of your results that got 0 credits:
http://docking.utep.edu/result.php?resultid=22831
Sorry, this was only meant for Linux users, not for Macs. Macs seem to do fine with 5.1 for the moment (but we'll release 5.2 anyway as soon as we get our compiler problem solved). ____________ D@H the greatest project in the world... a while from now! |
||
|
ID:
680 | Rating: 0
| rate:
/
|
||
|
I've looked at one of your results that got 0 credits: http://docking.utep.edu/result.php?resultid=22831 Yes, there are lots of those for the Macs. |
||
|
ID:
683 | Rating: 0
| rate:
/
|
||
|
Currently on vacation, but got a wireless connection in a hotel somewhere :-) Will the sending of the last result for this WU help? I've just seen it:
21940 306 28 Sep 2006 20:31:09 UTC 3 Oct 2006 20:31:09 UTC In Progress Unknown New --- --- ---
|
||
|
ID:
684 | Rating: 0
| rate:
/
|
||
|
I've looked at one of your results that got 0 credits: http://docking.utep.edu/result.php?resultid=22831 How can the WU be labeled as "Sucess" and "Invalid" at the same time? Also, why is it invalid? There are no error messages of any kind that I can see. ____________ Dublin, CA Team SETI.USA |
||
|
ID:
686 | Rating: 0
| rate:
/
|
||
|
How can the WU be labeled as "Sucess" and "Invalid" at the same time? Also, why is it invalid? There are no error messages of any kind that I can see. BOINC client said it was success (ie. application returned error code 0, no CPU limit exceeded, no problems downloading/uploading, etc) but the validator determined the result was invalid by looking at output files. I think the validator doesn't even run on a result that doesn't have Outcome: Success |
||
|
ID:
687 | Rating: 0
| rate:
/
|
||
|
BOINC client said it was success (ie. application returned error code 0, no CPU limit exceeded, no problems downloading/uploading, etc) but the validator determined the result was invalid by looking at output files. I think the validator doesn't even run on a result that doesn't have Outcome: Success I see. How do we find out why the validator determined it to be invalid? ____________ Dublin, CA Team SETI.USA |
||
|
ID:
688 | Rating: 0
| rate:
/
|
||
|
I see. How do we find out why the validator determined it to be invalid? Ask the project team ;) Afaik the validator only compares the different results, they will probably all be somehow different through rounding errors and such. If the results are within some limits the same, they are declared "valid". If one or two differ too much, they are "invalid". The single computer can't know that. If a puter constantly deliveres invalid results, although it thinks they are valid, something will be broken inside (too much overclocking, broken circuits, flawed memory...). |
||
|
ID:
689 | Rating: 0
| rate:
/
|
||
|
Ask the project team ;) You're telling me that I get screwed out of credits if someone else's WU gets messed up to the point where it doesn't match mine? That's nuts. This is a stock machine, with the official client. So there is either a problem with the application, or the other WU(s) this is being compared against. I'm guessing the former, so we should investigate this further (that's why we're here, right?). ____________ Dublin, CA Team SETI.USA |
||
|
ID:
690 | Rating: 0
| rate:
/
|
||
|
Same thing is happening to Intel PIII as well. It would be something to do with the application.
|
||
|
ID:
691 | Rating: 0
| rate:
/
|
||
|
You're telling me that I get screwed out of credits if someone else's WU gets messed up to the point where it doesn't match mine? That's nuts. That's always a possibility, of course, but a unreasonable one methinks. The validator doesn't know what's right or wrong, otherwise it wouldn't be necessary to send the WUs in the first place. What it does is treat all results that proclaim validity alike, and if some messed up puters will agree on the same wrong result, it has to take it as velid, how should it know otherwise? In the Wiki is an simple example of the validation process, look here . But as most puters are probably working correct, you usually only have to wait a bit longer until consensus is reached. |
||
|
ID:
693 | Rating: 0
| rate:
/
|
||
|
The results calculated on different platforms are just too different to validate at this point. Whether that means the science is accurate or not I couldn't say. I thought we were doing only homogeneous validation? In other words, my intel macs WUs are only being validated with other intel mac WUs. ____________ Dublin, CA Team SETI.USA |
||
|
ID:
694 | Rating: 0
| rate:
/
|
||
|
For these 0 credit WUs, consensus *was* reached, but mine gets labeled Invalid. The indefinately "pending" WUs is a different problem I am having. And this is not a rare thing. For computer 210, there are *10* of these on the results that are still showing. Like I said, this is completely stock mac, with stock client. Something is wrong with the application. ____________ Dublin, CA Team SETI.USA |
||
|
ID:
695 | Rating: 0
| rate:
/
|
||
|
It's not running homogenous validation. This has been discussed in another thread that I can't find at the moment.
|
||
|
ID:
698 | Rating: 0
| rate:
/
|
||
|
It's not running homogenous validation. This has been discussed in another thread that I can't find at the moment. According to the developer, we are using homogenous validation. See #3 here: http://docking.utep.edu/forum_thread.php?id=53&nowrap=true#590 FWIW, I agree. This really needs to be at the top of their To Do list. Credits are needed to keep people here. Yes, people expect there to be issues with alpha projects. But if you don't fix it quickly, people will leave. ____________ Dublin, CA Team SETI.USA |
||
|
ID:
700 | Rating: 0
| rate:
/
|
||
|
Afaik the validator only compares the different results, they will probably all be somehow different through rounding errors and such. If the results are within some limits the same, they are declared "valid". This is correct. But obviously it also checks that files are valid before comparing, ie. if format is correct, no upload errors, maybe file is empty because app failed! |
||
|
ID:
702 | Rating: 0
| rate:
/
|
||
|
Currently on vacation, but got a wireless connection in a hotel somewhere :-) Finally: application Charmm Success sucks ;) |
||
|
ID:
810 | Rating: 0
| rate:
/
|
||
|
Andre Kerstens
Forum moderator Project tester Volunteer tester
Joined: Sep 11 06 Posts: 749 ID: 1 Credit: 15,199 RAC: 0 |
||
|
We'll check these 0 credit results out as soon as possible. And of course we'll re-credit all of these results if needed.
Currently on vacation, but got a wireless connection in a hotel somewhere :-) ____________ D@H the greatest project in the world... a while from now! |
||
|
ID:
811 | Rating: 0
| rate:
/
|
||
|
We'll check these 0 credit results out as soon as possible. And of course we'll re-credit all of these results if needed. Don't worry about the credits, this is alpha, what's lost is lost. But make sure to have it straightened before going public. |
||
|
ID:
827 | Rating: 0
| rate:
/
|
||
|
Don't worry about the credits, this is alpha, what's lost is lost. I disagree. DC is sucessful because of credits. Full stop. Yes, this is an alpha project. Yes, credits are not guarenteed and may be lost. That's all implicit. But consider this: This is not only an alpha for the mechanics of the science project, but more importantly for how to deal with us "donators". The social testing. *Now* is the time to figure out the project's position on credits, and how to deal with unexpected issues like this. Now is the time to think it through. What's fair? And why? Do they need to make their decisions based on president with other projects, and why? Do they keep credit decisions on par with other projects? And why? Now is the time for the project to define their "moral center" with regard to issuing credits in return for our CPU time. This is the most important aspect of a DC project, because if they alienate us, there is no project. Once it "goes public", it will be far to late to figure that out. And without a precedence, they will appear capricious. That gets ugly. We've both seen that. So while they are testing with us, we are testing with them. We are here to help them do it right...right? ____________ Dublin, CA Team SETI.USA |
||
|
ID:
831 | Rating: 0
| rate:
/
|
||
|
suguruhirahara
Forum moderator Volunteer tester 
Joined: Sep 13 06 Posts: 282 ID: 15 Credit: 56,614 RAC: 0 |
||
|
Once it "goes public", it will be far to late to figure that out. 100% agree. The policy on credits should be established within the testing phase, because it's much easier to change rules (even if they were so radical). Also , when the number of the participants is limited, a consensus can be formed more easily. ____________ 
I'm a volunteer participant; my views are not necessarily those of Docking@Home or its participating institutions. |
||
|
ID:
838 | Rating: 0
| rate:
/
|
||
|
Once it "goes public", it will be far to late to figure that out. For my Intel Mac, the credits are evened out for the projects I am doing: (normalized as credits per 100 CPUseconds)
There is continuous updating to get the credits more or less equivalent across the projects and between computers based on the work done.
For my computer, Docking is giving quite a bit less credit than the other projects.
For my computer, Rosetta is giving similar or slightly less credit than average for the projects I am crunching at. |
||
|
ID:
840 | Rating: 0
| rate:
/
|
||
|
suguruhirahara
Forum moderator Volunteer tester
Joined: Sep 13 06 Posts: 282 ID: 15 Credit: 56,614 RAC: 0 |
||
|
Even if there are obvious differences for each project between credit per hour, I don't decide which projects I have my hosts crunch by concerning the relation. S@H has the high score, but I rarely crunch it because I don't have any interest in it. Whether larger amount of credits can get awarded or not doesn't motivate me.
|
||
|
ID:
841 | Rating: 0
| rate:
/
|
||
|
suguruhirahara
Forum moderator Volunteer tester
Joined: Sep 13 06 Posts: 282 ID: 15 Credit: 56,614 RAC: 0 |
||
|
double post...
|
||
|
ID:
842 | Rating: 0
| rate:
/
|
||
|
Once it "goes public", it will be far to late to figure that out. And without a precedence, they will appear capricious. That gets ugly. We've both seen that. That's just what I said: It's not important yet, but get it straight before going public. Of course credits are important in DC, and fairness with them as well. But now it's more important to point at faults and get them dealt with then to get credits. It's fine to get them, it's always better to get them in alpha as well, ubt if not, so what? Just make sure to watch a bit more closely for any irregularties with credits and mention them here. |
||
|
ID:
845 | Rating: 0
| rate:
/
|
||
|
Credits cannot be traded for anything therefore they are worthless. There should be no need for this project or any other project to develop a policy for giving out something that is totally worthless. Yet there is a need. Why?
|
||
|
ID:
849 | Rating: 0
| rate:
/
|
||
|
suguruhirahara
Forum moderator Volunteer tester
Joined: Sep 13 06 Posts: 282 ID: 15 Credit: 56,614 RAC: 0 |
||
|
Since credits are measureable in the boinc-sphere, it can be a milestone, and accelerates the competetion among teams and therefore advances the science (well, maybe I've said the similar thing on another board...). The case of the SAP of CPDN is a really good example. There has existed a competition (no officially though) between BOINC Synergy and UK BOINC Team for a month or so.
|
||
|
ID:
850 | Rating: 0
| rate:
/
|
||
|
Credits cannot be traded for anything therefore they are worthless. Just because you can't spend it, it's worthless? Wow, I guess there is no point in keeping score when playing tennis with a freind, or basketball, or using "money" when playing Monopoly. What's the point, right? ____________ Dublin, CA Team SETI.USA |
||
|
ID:
856 | Rating: 0
| rate:
/
|
||
|
You can't spend it, but it also doesn't cost the project to supply it. Things don't have to be of value to be valuable.
|
||
|
ID:
875 | Rating: 0
| rate:
/
|
||
|
Credits cannot be traded for anything therefore they are worthless. The point of a tennis match might be to determine who is the better player. In that case points serve a purpose. On the other hand some people play just for exercise and don't bother keeping score. Whatever the players agree upon is what's best for them. Likewise, the point of a monopoly game is to see who can become the wealthiest. The money (the points) serve a purpose there too. In both cases, participants agree upon what the game is all about. The point of crunching, however, is to do science, not to see who controls the most FLOPS. Lately it seems the point has become amassing credits which apparently proves A has more FLOPS than B. So what? What does that have to do with docking and folding? That's not the game I agreed to play and I resent the fact that my FLOPS and electricity are being wasted policing and enforcing the credit system and weeding out cheaters. If it cost me nothing to crunch I would care less. And if policing the cheaters didn't slow the science it wouldn't matter to me at all. But it does slow the science and I care deeply about doing better science and doing it faster. Suguru holds that the competition is good for the science. Is it really? We can see spikes in a project's production when 2 teams square off but when they do they ignore other projects to focus on just 1 project. Production at those other projects therefore must go down. While the total production of all projects combined doesn't suffer it certainly isn't enhanced by competition. So again, the competition does not serve the science. KWSN holds that the credits cost the projects nothing. That has been shown to be false in at least 1 case and it's probably false in many other cases as well. At SIMAP, for instance, the Holy Quorum of 3 was inviolable law, as sacred as the The Holy Trinity is to the Pope. Then POOF! one day the Good Code Faeries delivered code that made it possible to abandon the reliance on client-side benchmarks and POOF! they started doling out fixed credits and POOF! the 3 quorum was tossed out the window like a piece of fiery toast and a much more productive 2 quorum was ushered in. Well, that's all just too much POOFing and spoofing for me to buy. I don't believe in Code Faeries. It's obvious now the quorum of 3 had NOTHING to do with scientific rigor. It was all about preventing credit cheating. So, credits don't cost the projects? They cost dearly in lost production, 33% under a quorum of 3 if a quorum of 2 is all you really need. It seems 2 is all SIMAP ever really needed and it's more than likely the same at many other projects, 2 is enough. Then there is the cost of electricity and hardware paid for by crunchers, not the projects. The projects have become like bad welfare bums squandering the taxes you and I pay. They squander our resources and electricity and nobody holds them accountable. Somehow, a pail full of worthless points is supposed to make up for it. Horse excrement! Take a million of them to your utilities provider and just see if you can get a little knocked off your bill. Not worth even a Big Mac Worthless over at Ronnie Rotten's hamburger joint. I think it's high time crunchers did something about the waste and the credit system that fosters the waste. We need an across the board strike, 3 days of no crunching, maybe more, whatever it takes to either force the projects to sit down with the BOINC devs and get serious about revamping the credit system or else abolish credits across the board. |
||
|
ID:
880 | Rating: 0
| rate:
/
|
||
|
I think the credit system helps in many ways. First thing that comes to my mind is when a team looks for new members to improve in the rankings. Of course the system has its cons as well.
|
||
|
ID:
889 | Rating: 0
| rate:
/
|
||
|
suguruhirahara
Forum moderator Volunteer tester
Joined: Sep 13 06 Posts: 282 ID: 15 Credit: 56,614 RAC: 0 |
||
|
hi dave,
Suguru holds that the competition is good for the science. Is it really? We can see spikes in a project's production when 2 teams square off but when they do they ignore other projects to focus on just 1 project. Production at those other projects therefore must go down. While the total production of all projects combined doesn't suffer it certainly isn't enhanced by competition. So again, the competition does not serve the science. Please make sure that competition also motivates us to commit the project more. If one seriously commits the competition, then he perhaps try to bring his computer resource more out by, for example, upgrading CPU or RAM, or more simply, running PC longer. Doesn't it advance the science? Also, there's no need to help all sciences be advanced. Well, if scientists were forced to produce the scientific results in limited period, then they wouldn't depend upon distributed computing, which is sustained by volunteers, and purchase or borrow supercomputer by themselves... ____________
I'm a volunteer participant; my views are not necessarily those of Docking@Home or its participating institutions. |
||
|
ID:
891 | Rating: 0
| rate:
/
|
||
|
I think the credit system helps in many ways. First thing that comes to my mind is when a team looks for new members to improve in the rankings. Of course the system has its cons as well. Ummm... your users? It's the other way around, no? The projects are our users and since they use for free we are called hosts. I think if most of the hosts out there understood what their machines are really being used for they'd quit crunching tomorrow. Fortunately for the projects most hosts just crunch and don't bother keeping up with issues that affect them. Charities have been raked over the coals in the press before for wasting contributions and a few have not faired well. I think the projects ought to get their act together before some journalist with an attitude and a reputation for sticking up for the little guy decides there's a story that needs telling and a rip off that needs exposing. I encourage you to get your initial replication and quorum down to 2 no matter what it means to the credit system. |
||
|
ID:
894 | Rating: 0
| rate:
/
|
||
|
I encourage you to get your initial replication and quorum down to 2 no matter what it means to the credit system. So 2 is the scientific necessary amount for validation of a result. An experiment done twice is a valid foundation for a new theory? I simply don't know what's needed, could be one, could be 10, just don't let you decide this via credit discussion, but for scientific needs. As for this massive redundancy is a must, just not necessary on a WU by WU level, you can calculate credits on this redundancy, either WU by WU, or per batch of WU like Rosetta and Folding. |
||
|
ID:
899 | Rating: 0
| rate:
/
|
||
|
just don't let you decide this via credit discussion, but for scientific needs. If the science demands 3 then let it be 3, I have no problem with that. But if it's 3 just to reduce cheating then that is a massive abuse of the contribution made by crunchers and in that case I say throw out the whole credit system rather than let that abuse continue for even 1 more day. If 1000 credit whores leave because of it then too bad, it's their loss not the project's, the project will get 10,000 to replace them if they advertise their efficiency and their philosophy of spending OUR money to DO SCIENCE rather than spending it on titillating a pack of credit whores by converting cash into useless, worthless credits. I won't quit this project just because a credit system is in place but I have quit other projects (TANPAKU, Spinhenge) because they waste my CPU time and electric power far more than an alpha/beta project needs to. These days I favor SIMAP (quorum 2, replications 2) but I'll give D@H a chance and take into consideration it's only alpha. |
||
|
ID:
901 | Rating: -1
| rate:
/
|
||
|
Altruism got me to sign up my one PC. The competition/points/esprit de corps got me to
|
||
|
ID:
902 | Rating: 0
| rate:
/
|
||
|
Altruism got me to sign up a handful of machines I own, only 2 of which do much, the others are slow.
|
||
|
ID:
905 | Rating: 0
| rate:
/
|
||
|
Altruism got me to sign up a handful of machines I own, only 2 of which do much, the others are slow. You're *so* much better than me. Your message is coming through loud and clear. Funny thing is, my competitive crunching is just as valuable to the projects as your angelic crunching. Motivation is meaningless. A flop is a flop. Look, have you never participated in competitive team sports? Baseball? Basketball? Soccer? There is a reason many people do this beyond exercise. It's a great way to meet new friends, and teamwork can be its own reward. I think you are dead wrong about what motivates many (most?) personal crunchers. And that competition is critical. Without some way of keeping track, and measuring one's contribution, people lose interest. In my opinion, altruism attracts personal crunchers, but credits are what drive them them once here. This "gotta be on a team and kick butt" attitude so prevalent these days is a blight on humanity. Ah. So that's why I do it. ____________ Dublin, CA Team SETI.USA |
||
|
ID:
909 | Rating: 0
| rate:
/
|
||
|
Sorry you feel that way but gee, wasn't the point of your other poast to prove you're *so* much better than me? Your message is coming through loud and clear. Funny thing is, my competitive crunching is just as valuable to the projects as your angelic crunching. Motivation is meaningless. A flop is a flop. I agree, my message is loud and clear. Trouble is you just don't seem to be getting it. If we have a quorum of 3 rather than 2 just to keep your competitive crunching honest then your competitive crunching is wasting 33%. Look, have you never participated in competitive team sports? Baseball? Basketball? Soccer? There is a reason many people do this beyond exercise. It's a great way to meet new friends, and teamwork can be its own reward. Agreed. But does every single thing we humans do have to be organised into teams and a competition created? Why not just be on Team Docking and help kick disease butt. I think you are dead wrong about what motivates many (most?) personal crunchers. And that competition is critical. Without some way of keeping track, and measuring one's contribution, people lose interest. I disagree. I think you are dead wrong about what motivates crunchers and I think most would rather see the competition die than crunch 1 out of every 3 work units for nothing. Most people just get out from behind their 'puters and go bowling or join a softball league or something when they want to meet new people and kick some butt. You don't really meet people here, not really. Have you tried bowling or something like that? This "gotta be on a team and kick butt" attitude so prevalent these days is a blight on humanity. Ah. So that's why I do it.[/quote] Now it seems to sinking in. |
||
|
ID:
914 | Rating: -1
| rate:
/
|
||
|
No. Read it again. I was trying to illustrate the positive effect of competition on DC flops. I will say it again: Altruism gets people in the game, competition keeps them there and drives more flops. Not all people, obviously. But enough that DC would suffer without it. Your message is coming through loud and clear. Funny thing is, my competitive crunching is just as valuable to the projects as your angelic crunching. Motivation is meaningless. A flop is a flop. Duh. That's why I asked for exactly that in this very same thread back on 09/23. It's the 5th from the top. Read it for yourself. I don't recall seeing anyone here disagreeing with reducing the quarum, except the staff (at least for the time being). But that's not what we're arguing about now, which is the value of competition/credits to DC. Look, have you never participated in competitive team sports? Baseball? Basketball? Soccer? There is a reason many people do this beyond exercise. It's a great way to meet new friends, and teamwork can be its own reward. We are. I'm on the Docking@Home team, and I am also part of a sub-team. They're not mutually exclusive. [Balance of "I'm right, you're wrong", "You're right, I'm wrong" deleted] ____________ Dublin, CA Team SETI.USA |
||
|
ID:
916 | Rating: 0
| rate:
/
|
||
|
Me neither. I was just trying to show that altruism plays a big role here too and that DC won't die if the credits stopped tomorrow. As utility rates continue to spiral upwards more people are going to question the excesses. Each and every project would do well to trim excesses now, before people leave DC, rather than trying to get them back after they leave. If supporting the credit competition is what's driving the excess then the credits have to go for the sake of the science unless some other way of awarding credits can be implemented. |
||
|
ID:
921 | Rating: -1
| rate:
/
|
||
|
Andre Kerstens
Forum moderator Project tester Volunteer tester
Joined: Sep 11 06 Posts: 749 ID: 1 Credit: 15,199 RAC: 0 |
||
|
Don't speak for the staff please :-) I fully agree that we should go to a quorum of 2 if we can go to a quorum of 2 without compromising the results. For the moment, our scientist (who deliver the input data and process the output data) want to see 3 results agreeing before trusting the result enough. And that's the real job that we, boinc and all the other projects have to do: convince scientists that this system is a 100% reliable so that 2 agreeing results can be trusted; until that time we have to stick with three. But don't worry, behind the scenes a lot of footwork is done so that we can hopefully reduce the quorum at some point. Thanks! Andre PS I'm not a scientist, I'm a computer engineer :-) ____________ D@H the greatest project in the world... a while from now! |
||
|
ID:
937 | Rating: 0
| rate:
/
|
||
|
Wouldn't think of it. I was just repeating what you said in message 538 of this thread. I fully agree that we should go to a quorum of 2 if we can go to a quorum of 2 without compromising the results. For the moment, our scientist (who deliver the input data and process the output data) want to see 3 results agreeing before trusting the result enough. And that's the real job that we, boinc and all the other projects have to do: convince scientists that this system is a 100% reliable so that 2 agreeing results can be trusted; until that time we have to stick with three. But don't worry, behind the scenes a lot of footwork is done so that we can hopefully reduce the quorum at some point. I agree it is too early to switch to a 2 quorum system, at least for the macs. Different mac models returning different results is a real problem. And even a quorum of 3 doesn't really fix it. Let's say it is the mac mini that is creating the bad results. If you got a WU returned by 3 mac minis, they would all match, validate, issue credits...but still be a bad result. ____________ Dublin, CA Team SETI.USA |
||
|
ID:
938 | Rating: 0
| rate:
/
|
||
Message boards : Number crunching : Validator and Quorum
Database Error: The MySQL server is running with the --read-only option so it cannot execute this statement
array(3) {
[0]=>
array(7) {
["file"]=>
string(47) "/boinc/projects/docking/html_v2/inc/db_conn.inc"
["line"]=>
int(97)
["function"]=>
string(8) "do_query"
["class"]=>
string(6) "DbConn"
["object"]=>
object(DbConn)#60 (2) {
["db_conn"]=>
resource(120) of type (mysql link persistent)
["db_name"]=>
string(7) "docking"
}
["type"]=>
string(2) "->"
["args"]=>
array(1) {
[0]=>
&string(50) "update DBNAME.thread set views=views+1 where id=36"
}
}
[1]=>
array(7) {
["file"]=>
string(48) "/boinc/projects/docking/html_v2/inc/forum_db.inc"
["line"]=>
int(60)
["function"]=>
string(6) "update"
["class"]=>
string(6) "DbConn"
["object"]=>
object(DbConn)#60 (2) {
["db_conn"]=>
resource(120) of type (mysql link persistent)
["db_name"]=>
string(7) "docking"
}
["type"]=>
string(2) "->"
["args"]=>
array(3) {
[0]=>
object(BoincThread)#3 (16) {
["id"]=>
string(2) "36"
["forum"]=>
string(1) "2"
["owner"]=>
string(2) "52"
["status"]=>
string(1) "0"
["title"]=>
string(20) "Validator and Quorum"
["timestamp"]=>
string(10) "1160362880"
["views"]=>
string(4) "1775"
["replies"]=>
string(2) "54"
["activity"]=>
string(20) "4.5179505631299e-129"
["sufferers"]=>
string(1) "0"
["score"]=>
string(1) "0"
["votes"]=>
string(1) "0"
["create_time"]=>
string(10) "1158389083"
["hidden"]=>
string(1) "0"
["sticky"]=>
string(1) "0"
["locked"]=>
string(1) "0"
}
[1]=>
&string(6) "thread"
[2]=>
&string(13) "views=views+1"
}
}
[2]=>
array(7) {
["file"]=>
string(63) "/boinc/projects/docking/html_v2/user/community/forum/thread.php"
["line"]=>
int(184)
["function"]=>
string(6) "update"
["class"]=>
string(11) "BoincThread"
["object"]=>
object(BoincThread)#3 (16) {
["id"]=>
string(2) "36"
["forum"]=>
string(1) "2"
["owner"]=>
string(2) "52"
["status"]=>
string(1) "0"
["title"]=>
string(20) "Validator and Quorum"
["timestamp"]=>
string(10) "1160362880"
["views"]=>
string(4) "1775"
["replies"]=>
string(2) "54"
["activity"]=>
string(20) "4.5179505631299e-129"
["sufferers"]=>
string(1) "0"
["score"]=>
string(1) "0"
["votes"]=>
string(1) "0"
["create_time"]=>
string(10) "1158389083"
["hidden"]=>
string(1) "0"
["sticky"]=>
string(1) "0"
["locked"]=>
string(1) "0"
}
["type"]=>
string(2) "->"
["args"]=>
array(1) {
[0]=>
&string(13) "views=views+1"
}
}
}
query: update docking.thread set views=views+1 where id=36