From carl.h

Ethan come on....

If he's taken sides, he's taken it for the overclaimers.

This as well please...

---

TY ;-)

---

Not all Czech`s bounce but I`d like to try with Barbar ;-)

Make no mistake This IS the TEDDIES TEAM.

From Saenger:

If he's taken sides, he's taken it for the overclaimers.

I understand this as a "Yes" from your side.
So you're not impartial, is that OK for a moderator over here?

Just posted by DB on backdating credits:

Second, To answer a question which came up on the boards--we will NOT be backdating credit totals. The new system will go into place early next week, adding on to the current totals.

---

Just posted by DB on backdating credits:

Second, To answer a question which came up on the boards--we will NOT be backdating credit totals. The new system will go into place early next week, adding on to the current totals.

That's just sad, but I probably have to live with it.
So all this overblown numbers will stay, and those who did the overclaimimg can continue to pretend to have done more work.
At the moment all stats here on Rosetta are more a bunch of random numbers then some measurement of work done. If this could be changed, it would be fine. If this could be changed even retroactive, it would be even finer imho.

It's a reasonable decision which I think everybody can accept, so let's go forwards with it rather than feeling bitter about the past (from whichever side of the fence people come).

---

Another post from DB on his journal:[snip]
Second, To answer a question which came up on the boards--we will NOT be backdating credit totals. The new system will go into place early next week, adding on to the current totals. [snip]

That's just sad, but I probably have to live with it.
So all this overblown numbers will stay, and those who did the overclaimimg can continue to pretend to have done more work.

Saenger:

The top teams want a fair/work based credit system. It is going to refelect the quality of the hardware ( in the cases they are like XtremeSystems and others) and the number of machines that have been engaded. a Kentfield , a Conroe, a Xeon, A Power Mac , running 24/7 will out produce a P4 or a P3 , That is a fact. At the end when all is balanced and when the tweaking that is needed to be fair , to recognize and acknowledge the diversity of the different CPUs and Os and Mobos involved, those who have the most powerful machines will have the best scores.

Another fact is that some teams have a culture where the most important thing is participation in the largest number of projects, while in other teams, the culture is to concentrate in few. Both cultures are valid and do produce good. What is wrong is trying to impose one worlds view on the other.

Hey, you seem to believe that competition is not good, I do . Let's agree to disagree. But please stop abscribing ulterior motives to those who believe like I do. I can guarantee you that that alone, will reduce the level of reaction; of anger that has come from me and those who believe like me. You dont call people cheats , time after time and expect a calm reaction.

We also disgaree on the issue of BOINC. To me and many BOINC is a tool, that is all. It is not revealed inspiration to be taken by faith. Sadly that good tool has a big security problem. Alas, every time people point that out; those who take BOINC as the only way to go clam up and say "Öpen Source, end of discussion" Sorry, it is not end of discussion: something has to be done with a sorce code that allows tampering not only in its credits but as it has been acknowledged in the science. And the problem with Boinc is way past the use of optimized clients.


My team abhors cheating as intensly as anyone one of you.

People seem to forget that the same day, we reached the top spot in the rankings , we got unimpeacheable confirmation that two of our members cheated by adquiring access to computers without the authorization of the owners. Right in the middle of the celebaration, our team leadership requested that close to 10% of our credits , close to 3.9 million credits be zeroed out. The ones that did not cheat (more than 450 members at that time) did not complaint: they agreed with our leadership and the next day we started crunching again.

It is aslo a known fact acknowledged by Dr Baker, that it was members of our team , specially me, that took to the developers the concerns that were reflected in these message boards that some claims were out of the realm of believability and that something had to be done. That is why members from various teams and individual crunchers worked with the developers documenting and comparing credit claims in a way that was not going to be disruptive to the functioning of the project. That is why nay cases of overclaiming were readjusted and you never saw a big spectacle made that could disrupt the project.


As to the old credits: To do what you have been proposing is to change rules after the fact: it is unfair. Under civil law something like that could be considered breach of contract and in our system of criminal laws is a violation of rights. For that reason alone: backdate was a no go from the begining. And if you had doubts about the potential for disruption, the last days are eveidence of that the mere idea of it is disruptive.

As to the optimized clients : The rules in this Project allowed them. Again, you are arguing heavy philosophy but, what is allowed in the rules, is legitimate. You can disagree with that: I can see why a person coming your crunching experience can have a very valid argument but those who have followed the rules as they were set from the get go have an equally valid claim , worst they have an additional equity claim: ex post facto application of rules is unfair . Calling people cheats when they followed the rules, inundating forums with several threads with the same argument doesn get you a recognition for your consistence of thought but it provokes anger at what some people will call obtusenness (sp) in trying to impose an patently unfair solution to what you percieve as a poblem.

Worst, the response to you and those who like believe in participation in large number of projects (with the effect of lower scores in the projects) elicited from those who believe in specilization in few projects and crunching for them full time ( which obviously brings larger scores) is the natural one: "who do those zero racers believe they are in trying to impose their way in OUR project" and that is what you got. Compounding that , the knowledge of what happened at SETI... which brought the natural reaction "these are the same popel that messed up SETI and now they are comming to mess in OUR project"

That a small recount of the starting of this flame war.

Sad.

That said.. as to your statement "...and those who did the overclaimimg can continue to pretend to have done more work." it is yo tht have to PROVE we did not. We dont have to prove you anything. In our system of law and justice those who make the accusations have to prove them, not those who are acused. Unless you can provide facts , all you have is your bnias and the innuendos that have been spouted . If you dont have actual work data to show that what you stated is true all you have is bias and innuendo.

Prove what you stated or else give it up. continue and the reaction will not be as calm as this has been.



Peace


Latter Edit

Note to Ethan: Sice I ws responding to Saenger while you posted . I did not see your post about him posting his comment elsewhere. So Right now I am at a loss: either I edit out my response , which would allow his post to satnd unchallenged and it has to be challenged. Or I wait till he moves or is moved.

So in this one : I will understand if I am deleted.

Saenger:

The top teams want a fair/work based credit system. It is going to refelect the quality of the hardware ( in the cases they are like XtremeSystems and others) and the number of machines that have been engaded. A Kentfield , A a Conroe, a Xeon, A Power Mac , running 24/7 will out produce a P4 or a P3 , That is a fact. At the end when all is balanced and when the teaking that is needed to be fair , to recognize and acknowledge the diversity of the different CPUs and Os and Mobos involved, those who have the molst powerfull machines wil have the best scores.

Another fact is that some teams have a culture where the most important thing is participation in the largest number of projects, while in other teams, the culture is to specially in few. Both cultures are valid and do produce good. What is wrong is trying to impose one worlds view on the other.

I mostly agree so far.

Hey, you seem to believe that competition is not good, I do . Let's agree to disagree.

Here I disagree.
I like competition as well, and that's why I want a fair credit system. The current one isn't. There is no way to really determine who has done how much for the project in the current credit scheme.

The new one will hopefully change that, and I think it's sad that the possible adjustment of the invalid current numbers will probably not take place. But I also think I can live with it, although I will still try to convince the project team that fairness is better.

But please stop abscribing ulterior motives to those who believe like I do. I can guarantee you that alone, will reduce the level of reaction of anger that has come from me and those who believe like me. You dont call people cheats , time after time and expect a calm reaction.

You don't call people whiners who only want a fair competition, and it will be far more calm.

We also disgaree on the issue of BOINC. To me and many BOINC is a tool, that is all. It is not revealed inspiration to be taken by faith. Sadly that good tool has a big security problem. Alas, every time people point that out; those who take BOINC as the only ay to go clam up and say "Open Source, end of discussion" Sorry, it is not end of discussion: something has to be done with a sorce code that allows tampering not only in its credits but as it has been acknowledged in the science. And the problem with Boinc is way past the use of optimized clients.

What other problems are there?
Open source is imho far better then closed source for BOINC (excluding the credit part for non-validating projects), as it allows it to be compiled to more systems, to be applied and adjusted to more projects requirement, and to have a more direct possibility of input by the big volutary community on the development of it.
The scientific application is another matter, here I fully understand the necessity of closed source in some projects, with the disadvantages of the smaller dev-population.

My team abhors cheating as intensly as anyone one of you. People seem to forget that the same day, we reached the top spot in the rankings , we got unimpeacheable confirmation that two of our mebers cheated by adquiring access to computers without the authorization of the owners. Right in the middle of the celebaration, our team leadership requested that close to 10% of our credits , close to 3.9 million credits be zeroed out. The ones that did not cheat (more than 450 members at that time) did not complaint: they agreed with our leadership and the next day we started crunching again.

It is aslo a known fact acknowledged by Dr Baker, that it was members of our team , specially me that took the concerns that were reflected in these message boards that some claims were out of the realm of believability and that something had to be done. That is why members from various teams and individual crunchers worked with the developers documenting and comparing credit claims in a way that was not going to be disruptive to the functioning of the project. That is why nay cases of overclaiming were readjusted and you nver saw a big spectacle made that could disrupt the project.


As to the old credits: To do what you have been proposing is to change rules after the fact: it is unfair , under civil law something like that could be considered breach of contract and in our system of criminal laws is a violation of rights. For that reason alone: backdate was a no go from the begining. And if you had doubts about the potential for disruption, the last days are eveidence of that.

As to the optimized clients : The rules in this Project allowed them. Again, you are arguing heavy philosophy but, waht is allowed in the rules, it is legitimate. You can disagree with that: I can see why a person coming your crunching experience can have a very valid argument but those who have followed the rules as they were set from the get go have an equally valid claim , worst they have an additional equity claim: ex post facto application of rules is unfair . Calling people cheats when they followed the rules, inundating forums with several threads with the same argument doesn get you a recognition for your consistence oif thought but it provokes anger at what some people will call obtusenness (sp) in trying to impose an patentky unfair solution to what you percieve as a poblem.

Worst, the response to you and those who like believe in participation in large number of projects (with the effect of lower scores in the projects) elicited from those who believe in specilization in few projects and crunching for them full time ( which obviously brings larger scores) is the natural one: "who do those zero racers believe they are in trying to impose their way in OUR project" and that is what you got. Compounding that , the knowledge of what happened at SETI... which brought the natural reaction "these are the same popel that messed up SETI and now they are comming to mess in OUR project"

That a small recount of the starting of this flame war.

Sad.

That said.. as to your statement "...and those who did the overclaimimg can continue to pretend to have done more work." it is yo tht have to PROVE we did not. We dont have to prove you anything. In our system of law and justice those who make the accusations have to prove them, not those who are ccused. Unless you can provide facts , all you have is ytour bnias and the innuendos that have been spouted . If you dont have actual work data to show that what you staed is true all you have is bias and innuendo. Prove what you stated or else give it up. continue and the reaction will not be as calm as this has been.

My prove is the use of the 5.5, wich is claiming too much credit.
As long as it's not clear who used this overclaiming client, who used the normal one, and who even used some more overclaiming client, I consider all stats here as meaningless.
There is obviously a possibility to get the record straight, to look at the work actual done as far back as February, burt it will not be done. That's something I still find sad. The overclaimers "won" this battle, but fortunately not the war, unless the new system will be adjusted to their overblown claims.

That is not a threat, it is a fact. People do loose patience as those who libel and slander them.

Peace

That' right. Would you please tell taht your teammates, who regularyily slander anyone who dares to disaggree?

Peace.

I want to describe the situation as I see it:

Rosetta doesn't need the validation of every WU because it's in itself heavy redundant (thousands of runs of the same molcule), so they had a quorum of 1.

They calculated the credit accourding to the original BOINC scheme as benchmark*CPU-time.

If people would all use the stock client, all would have got approximately the same credits. Some are a bit disadvantaged because of the suboptimal benchmarking, like Linux.

There were so-called "optimized clients" on the market, that were developed for the use with optimized applications over @SETI. They overcome the disadvantage of the optimizations combined with the credit claiming process by artificially inflating the benchmark to calculate the same amout for an optimized WU as with the original application on the original client.

The use in other projects was explicitly discouraged by the developers, and usually it didn't matter anyhow, as most projects use as validation tool a quorum of 3 or 4, so off-average claims would be discarded.

A problem pops up if the project uses another kind of validation, like Rosetta, CPDN, Folding... and so doesn't weed out obvious overclaims.

So the situation now is this:
Probably most users used the stock client, got normal credits, and were happy.
Some used the "optimized" client, got more credit then normal, and were happy as well.
Some (I think very few) even went further, and compiled their own, even more "optimized" client, and were even more happy.

This came to the light as some teams encouraged the use of the "optimized" client and the irregularities in the granting process got exposed.

Now there is no way to tell how much anyone really has done for the science, as all credits are tainted with possibility of being just artificially inflated.

It was probably not the intend by many users to artificially inflate the credits by the use of the 5.5 and thus make the competition worthless, but it had this effect.

And to make one thing clear: This applies as well if you only look at Rosetta as a solitary project, but of course more if you look at it in the context of the whole BOINC environment.

AFAIK there is a possibility to get the records straight, and to let the cerdits mirror the real input in science again as far back as February. This should be done to get the competition straight (although it's of course a waste of ressources for the science;).

Edit: speelink

Saenger, we said that for weeks.

What I still don't understand is why they wouldn't compete with the same weapons. If everybody uses the same client, they surely are still at the top. And then everybody can realize their involvement in the project in terms of users, or machines or whatever.
Now we are in a sort of fog.

Equal credits and thus a healty competition, that's we ask since the beginning of this tragedy, until the word cheater was used and .....

If one team used them, another team would have to use them to keep even. Let's not go back on how we got the credits we have now as I think everyone understands.

---

...I think everyone understands.

I'm a bit unsure about that, as some still claim the 5.5 did "level the playing field", while it did the opposite.

Now there is no way to tell how much anyone really has done for the science, as all credits are tainted with possibility of being just artificially inflated.

Question : You have 2,540 credits exactly how much of the science have you done.

It`s not facecious, you seem to imply that the credit is a measure of the science, I see it as measure of what your machine has done compared to mine, a measure of the science cannot yet be assessed IMHO.

To further state you want to compare and compete against my 4 machines that (were) running rosetta 100% of the time seem`s a bit ludicrous and to be frank, without being offensive, a bit of a lie. I can see no purpose to you even looking at the figures. Now if you are talking as the head of a team this may differ.

XS pushed for and the project has come up with another measure. To retrospectively remeasure all points would be a big task, we know this is now not going to happen. What stands is what stands, to continue the optimised/not debate is foollhardy and alters nothing. Even if I now happened to go over to your side and agree with you, there is nothing either of us can do.

Please remember it was Jose and XS that pushed for this fairer system and they obviously, judging by David Bakers post`s in XS forums, carry some weight ( like it or not; no he didn`t visit Teddies either)

So Jose has got you a new fairer system which I think you wanted, anything else such as backdating is not going to happen.

---

Not all Czech`s bounce but I`d like to try with Barbar ;-)

Make no mistake This IS the TEDDIES TEAM.

...I think everyone understands.

I'm a bit unsure about that, as some still claim the 5.5 did "level the playing field", while it did the opposite.

What the client did or didn't do isn't relevant, nor are the motives of those who used it. It wasn't forbidden by the lab and the science results were valid and helped the research. I only wish to stop this line of posting since it doesn't help where were at now after Dr. Baker has stated that there will be no back-crediting.

---

I'm a bit unsure about that, as some still claim the 5.5 did "level the playing field", while it did the opposite.

Old ground saenger and saying it is not going to rewrite history, is it ?

---

Not all Czech`s bounce but I`d like to try with Barbar ;-)

Make no mistake This IS the TEDDIES TEAM.

Jose,
I didn't participate in the Seti war, I leave Seti just before to start crunching for Rosetta. Just a remark, I proposed several time to compete on the same level or create various ranking for various type of crunchers. I only got sarcasms... then things slip. But OK, I had my standard BOINC point of view and I tried. If it is not possible, I give up. You're probably beter.

...I think everyone understands.

I'm a bit unsure about that, as some still claim the 5.5 did "level the playing field", while it did the opposite.

What the client did or didn't do isn't relevant, nor are the motives of those who used it. It wasn't forbidden by the lab and the science results were valid and helped the research. I only wish to stop this line of posting since it doesn't help where were at now after Dr. Baker has stated that there will be no back-crediting.

Would you please do everyone a favor, and at this point discourage the use of the "optimized" BOINC clients going forward?

---

BOINC WIKI

Alright. . this thread is closed. The lab has stated that past credits won't be recalculated, which was the topic of this thread.

Since I can't actually lock the thread, future posts will be deleted.

John, in the new system it doesn't matter what client users have. . the credit claims get averaged out and everyone gets the same work credit.

---