Hello..I came here for some help on Rosetta project

[Verwijderd]

I'm Movieman from XtremeSystems.
As you may have seen there's been a lot of comotion over at the Baker Labs forum(Rosetta) brought on by a small group that came from SETI to Rosetta. They made up their minds that they were going to change everything around to suit their ideas of how things should be.
The people here that know me know I'm not one to sit back and be quiet about things I disagree with.
What I want to do is this:Send a message to Baker Labs that they will hear so clearly that maybe they will just start to think that we are not 10,000 individuals that can be ignored but a group that can make a statement of their displeasure on a problem.
If you don't want to join in on this I will understand and there will be no hard feelings.
I've just posted this same message at Free DC and XS has already started to shut down.
It will take a few days for us to shut down totally but it will happen. The other option is to buffer your work units and shut off the network part of Boinc..
I figure one week total should send a message they won't forget and it will tell them that we can work together as one huge entity.
Thanks for your time,
Movieman

[ Voor 6% gewijzigd door Verwijderd op 18-08-2006 16:24 ]

[Witlof]

Is this about the optimized clients and/or credits system? I don't know the whole story so please tell me (and maybe us) more. A short version is OK or some links to topics concerning this subject.

Found some info in the yesterdays DPCH. Click

[ Voor 45% gewijzigd door Witlof op 18-08-2006 16:37 ]

[ColdFusion]

It would be helpful if you would enlighten us with some information where this whole problem is about

[Verwijderd]

yes, this is about the clients but this is mostly about a small group of very vocal people trying to tell a project that they will change everything or they will post it to death..and thats what these idiots did until baker Labs caved in. I want to say to Baker Labs that guess what pal, we are the ones that do the work not these guys with a 20RAC that have a score of 2000 for a years time.
Any help you guys can give to this will be appreciated.
Yes, I am mad..54 but still the Irish temper!
Movieman

[Verwijderd]

it`s funny: i cannot understand an irish guy talking, but even an irish guy writing i can hardly read.

[Verwijderd]

I'm sorry for that. I thought it might be an issue. The world acts as if it revolves around english and it doesn't..I can speak french,spanish,latin and a little german but Dutch baffles my mind.It looks like a variation of English,French and some german but I could be totally wrong.
Movieman

[Verwijderd]

Verwijderd schreef op vrijdag 18 augustus 2006 @ 16:53:
I'm sorry for that. I thought it might be an issue. The world acts as if it revolves around english and it doesn't..I can speak french,spanish,latin and a little german but Dutch baffles my mind.It looks like a variation of English,French and some german but I could be totally wrong.
Movieman

More like a bit of german, with german and some....german.
Can't detect any french accept the words like, trottoir etc.

[Verwijderd]

ok, I am learning. BUT my german is limited to counting 1-20 and I might make a few errors doing that. I've looked at German but it looks a lot different than Dutch.
Movieman

[ColdFusion]

Well as for now I just don't have any opinion. Maybe it is just a few vocal people who try to get things their way. But are these bad things then? I do run Rosetta, but I hardly ever read the Baker Lab forums. Only thing I came across was new credit system, which I think is a good thing.

So with the information (none) you're giving me here, I can't really judge about the things that are supposed to be bad (?). Perhaps Kristof can help here.

[Verwijderd]

ColdFusion schreef op vrijdag 18 augustus 2006 @ 17:07:
Well as for now I just don't have any opinion. Maybe it is just a few vocal people who try to get things their way. But are these bad things then? I do run Rosetta, but I hardly ever read the Baker Lab forums. Only thing I came across was new credit system, which I think is a good thing.

So with the information (none) you're giving me here, I can't really judge about the things that are supposed to be bad (?). Perhaps Kristof can help here.

I'll try and explain as best I can. It's not a change to a work based credit system. I've been after that for months. The one they are testing right now has more errors than the Titanic has holes in it.
Run one work unit 4 hours and get 23 points, run another 4 hours on the same machine and get 86, a third and you get 63..That's from my own DX3600 system.
This big issue here is power. These guys have forced changes based on their idea that everything should be exactly the way they want it and anything else and they will yell and scream till they get what they want. THAT is what I want to stop. Besides this, they contribute next to nothing to Rosetta.
This is a power struggle plain and simple and the more machines I can turn off the better I can make the point to Baker Labs..
Movieman

[Bio1ogics]

Hey Movieman, nice to see your face around here.
Is it a good idea to get the spokespersons for DPC, Free-DC and Xtreme systems to collectively send an e-mail about the concerns directly to David Baker and/or his team?
Our person here is Kristof (his name at dutchpowercows dot org).
Topics for that mail:
Give David a nice mathmatical example.
Ask David where this will end: Will this group next demand certain drug-targets to be calculated?
Or give their own targets to be calculated?

These are grave concerns which David should recognise also.

For those wanting to click something
Movieman's Free-DC thread

[Kristof]

(Message in English, so everyone (also non-DPC members will understand. If any DPC member doesn’t understand, please ask, and sorry if there are mistakes in my English….
Even een boodschap in het engels, zodat Movieman ook kan volgen. Moest je iets niet begrijpen, vraag het gewoon, dan wordt het vertaald  )


To give you all a explanation:
• As you know Rosetta has launched the idea to start with a new crediting system. The current system has some negative points:
- There is a difference between AMD and Intel processors, a difference between Windows and other OSes, etc
- There are normal clients and optimized clients (these optimized clients were first meant to give a more fair score to eliminate the points above. But they grow and asked more and more credits, so it became less fair.
• The new system should give a solution for that. Every WU has an amounth of credits, and you’ll get rewarded that amount, when you send in the result. So it doesn’t matter if you crunch 1 hour on that WU, or several hours. You’ll get the same points. The difference is that you can crunch more WUs within the same time with a faster PC.


So far so good. The idea was good, but as usual the implementation.

• The amount of credits attached to a WU has to be in proportion with the amount of time you spend on it. So if you crunch for 4 hours on a WU, you should get the same credits if you crunch on a similar WU also for 4 hours with the same computer. First test have given the result that this isn’t the case.
• Also some people want that they will reset all credits earned until today, and start over again with the new system (cfr TSC/D2OL)
• Others want to go on with the new system starting ‘today’, but keep the current credits
• And another group wants to recalculate the past earned credits, via the calculation of the new credit system.

Now the point of Movieman:
On the forum, there are discussions ongoing that get the lead of users who have crunched only a few 1000 points on Rosetta. They are shouting that:
• The use of optimized clients (like we, DPC, do use, but also XS, FreeDC and others use) is cheating
• The currently earned credits should be deleted and that they have to start over with 0 credits etc

The problem of Movieman is that the project team seems to listen to the small users, and won’t take into account the comment of the big crunchers (XS, DPC, FreeDC, …).

[ Voor 98% gewijzigd door Kristof op 18-08-2006 18:27 ]

[Verwijderd]

Bio1ogics schreef op vrijdag 18 augustus 2006 @ 18:05:
Hey Movieman, nice to see your face around here.
Is it a good idea to get the spokespersons for DPC, Free-DC and Xtreme systems to collectively send an e-mail about the concerns directly to David Baker and/or his team?
Our person here is Kristof (his name at dutchpowercows dot org).
Topics for that mail:
Give David a nice mathmatical example.
Ask David where this will end: Will this group next demand certain drug-targets to be calculated?
Or give their own targets to be calculated?

These are grave concerns which David should recognise also.

For those wanting to click something
Movieman's Free-DC thread

That's an idea that might be well worth doing..JUST our names on that email might make them think twice. I sent david kim a plainly and frank email last night on my feelings which he should be opening right about now(9am on the west coast of the USA right now)
Teddies has already shut down. I'm debating contacting other teams.
Maybe if they see the equalivant of 1 million points a day disappear they will open their eyes.
At the VERY least they will know that we can act together and there is strength and power in numbers.
Movieman

[Kristof]

Verwijderd schreef op vrijdag 18 augustus 2006 @ 18:11:
That's an idea that might be well worth doing..JUST our names on that email might make them think twice. I sent david kim a plainly and frank email last night on my feelings which he should be opening right about now(9am on the west coast of the USA right now)
Teddies has already shut down. I'm debating contacting other teams.
Maybe if they see the equalivant of 1 million points a day disappear they will open their eyes.
At the VERY least they will know that we can act together and there is strength and power in numbers.
Movieman

Movieman, you agree with my message?
And eurm, please contact me by e-mail also (the e-mail you send to David etc)
you know my e-mailaddress I guess

[Verwijderd]

you got a good point there movieman, i was reading yesterdayevening the post here on tweakers.
and i was a bit shocked, i was a year ago active on TSC project, and the pult the same stunt as the guys from bakerlab.
for me it was the reason to left the project.
as you can see my output on rosetta is very low as i was planning to leave rosetta for another project.
so i join youre action and shut it down for at least a week.

[Martijn13]

Even een kort berichtje in het nederlands: alle mensen die nog niet weten waar dit over gaat, moeten even deze draad lezen.

Thanks for everyone's support

[Verwijderd]

Kristof schreef op vrijdag 18 augustus 2006 @ 18:28:
[...]

Movieman, you agree with my message?
And eurm, please contact me by e-mail also (the e-mail you send to David etc)
you know my e-mailaddress I guess

Kristof: EXcellently said! I just read your message in the other post and it's perfect.
Lets forget points for a minute and just talk work units.
I'm guessing that combined the DPC,Free DC and XS do at least 50% of the work done daily on Rosetta. Thats just my feeling, correct me if I am wrong.
If that number is correct can you imagine the effect of the 3 teams shutting down all together.
I know it can't be done in one day but over several it could be done.
The idea of the email has merit but I think it will fall on deaf ears but that doesn't mean we shouldn't try it.
Let me see how this plays out today and get back to you tomorrow. It's only 1pm here now and 10am on the west coast of the USA where baker Labs is located. They are still having their first coffee.
Thanks,
Movieman

[ColdFusion]

Isn't it an idea to perhaps make a petition like poll where the above mentioned points can be voted on? And that this url is passed on to everyone. That way you get a lot of votes, people tend to click on something rather than type up their opinion. With all the votes from the Top 3 teams, I think quite a point will be made.

[Bio1ogics]

Just a C&P from XS-forum, by David Baker

Dear XS crunchers,

Your combined efforts have been really important to what we are trying to accomplish! I'm sorry that the discussions of the new credit system have gotten so out of hand, the last thing of course we want is to alienate the people who have been contributing so much. I didn't anticipate this at all.

I think that after we make the transition (there will be no backdating) everybody will be happier as the credits will more accurately report what a computer has accomplished and for computer hardware experts interested in maximizing performance comparisons will be more meaningful.

I think that the main problems have been not with the proposed system itself but with all the discussions it has provoked. I'm thinking we should just remove these threads from our boards. What do you think?

thanks again,

David

So it seems, they've opened their mail-box Movieman! ;-)

Plus, no backdating

[Martijn13]

Bio1ogics schreef op vrijdag 18 augustus 2006 @ 19:20:
Just a C&P from XS-forum, by David Baker

[...]


So it seems, they've opened their mail-box Movieman! ;-)

Plus, no backdating

Just wait what happens, we'll see afterwards. He has left the forums already for now.

EDIT: kan iemand die morgen de DPCH post even een linkje of verhaaltje maken naar/over deze thread? Het is denk ik wel belangrijk als mensen dit weten!!

[ Voor 15% gewijzigd door Martijn13 op 18-08-2006 20:22 ]

[ParaNoiMia]

Martijn13 schreef op vrijdag 18 augustus 2006 @ 19:34:
[...]


Just wait what happens, we'll see afterwards. He has left the forums already for now.

EDIT: kan iemand die morgen de DPCH post even een linkje of verhaaltje maken naar/over deze thread? Het is denk ik wel belangrijk als mensen dit weten!!

Je mag ze ook zelf posten hoor, ook als niet dpc-er

[Martijn13]

ParaNoiMia schreef op vrijdag 18 augustus 2006 @ 22:19:
[...]


Je mag ze ook zelf posten hoor, ook als niet dpc-er

Doe ik als ze er nog niet zijn morgenochtend vroeg

[Verwijderd]

I just put it all out in the open on the Baker Labs forum. Nothing hidden, plain and truthful talk.
maybe this will say to them how we feel but I have a feeling that the only thing they will understand is seeing the contibution from all of us drop to zero.
If the 3 teams stop sending back data for 3 days, that will make a statement that no one can ignore.
Thanks for listening,
Movieman
here's the link:
http://boinc.bakerlab.org/rosetta/forum_thread.php?id=2121

[ Voor 9% gewijzigd door Verwijderd op 18-08-2006 23:29 ]

[Verwijderd]

Guys, I just had enough. 200 posts in a thread and a mod deletes it all. I am done with rosetta for good. Thanks for listening to me today and I wish you guys the best of luck.
Movieman..

[Crack]

Deleting a tread on a forum is of course a big no-no. I didn't read it but a offending topic should only be closed.

If this is all the cause of the new credit system, then I don't understand the problem realy...
I've tested the new system on RALPH@home: The new system is still being tweaked, but the principle seems fair to me. Nothing's final yet, and some bugs are fixed already. The difference isn't that big anymore: Now it's only 10%, that's a lot better that a standard linux client compaired to the 'optimised windows' version in the old system.

The remark that big crunchers' opinion is more important because they crunch more units is absolutely rubish. The dedication to the project of a smal cruncher that particpates is as great as someone with 100 computers.
If you aren't happy with R@H and want to pull your machines out, that's up to you. There are enough other projects that you can join.
Crunch on.

[Martijn13]

Crack schreef op zaterdag 19 augustus 2006 @ 03:19:
Deleting a tread on a forum is of course a big no-no. I didn't read it but a offending topic should only be closed.

If this is all the cause of the new credit system, then I don't understand the problem realy...
I've tested the new system on RALPH@home: The new system is still being tweaked, but the principle seems fair to me. Nothing's final yet, and some bugs are fixed already. The difference isn't that big anymore: Now it's only 10%, that's a lot better that a standard linux client compaired to the 'optimised windows' version in the old system.

The remark that big crunchers' opinion is more important because they crunch more units is absolutely rubish. The dedication to the project of a smal cruncher that particpates is as great as someone with 100 computers.
If you aren't happy with R@H and want to pull your machines out, that's up to you. There are enough other projects that you can join.
Crunch on.

It wasn't the Credit problem. I understand they're still tweaking it. It was the fact that people are scamming the project forums, people that have no rights to say anything whatsoever (well, they have the rights, but contributed close to nothing to the project), and the lack of communication with BakerLabs. Now that Baker Labs finally starts to understand what is going on, they are trying to get out of trouble. THAT is what we don't like. Have a look at the latest DPCH for more info.

[Pim.]

Let's keep the discussion in this thread, that way we have just one English topic (on this Dutch forum) and the topic keeps in sight insted of fading away like an DPCH does

[Verwijderd]

Pim. schreef op zaterdag 19 augustus 2006 @ 11:02:
Let's keep the discussion in this thread, that way we have just one English topic (on this Dutch forum) and the topic keeps in sight insted of fading away like an DPCH does

Sorry if I posted in the wrong place. If I did it was by accident and accept my apology.

Crack schreef op zaterdag 19 augustus 2006 @ 03:19:
Deleting a tread on a forum is of course a big no-no. I didn't read it but a offending topic should only be closed.

If this is all the cause of the new credit system, then I don't understand the problem realy...
I've tested the new system on RALPH@home: The new system is still being tweaked, but the principle seems fair to me. Nothing's final yet, and some bugs are fixed already. The difference isn't that big anymore: Now it's only 10%, that's a lot better that a standard linux client compaired to the 'optimised windows' version in the old system.

The remark that big crunchers' opinion is more important because they crunch more units is absolutely rubish. The dedication to the project of a smal cruncher that particpates is as great as someone with 100 computers.
If you aren't happy with R@H and want to pull your machines out, that's up to you. There are enough other projects that you can join.
Crunch on.

The thread delete was just the final straw to me. The main thing that pissed me off was the 4 weeks of daily hearing myself and my team called cheaters by people who in real terms contribute nothing to Rosetta and absolutely nothing being done about it by the moderators.
Think on this a minute: Every person who visits the baker forum sees in print that your team is cheating. That thought leaves with them and diminishes your team's accomplishment in their eyes.
These are the things that made me mad and decide to boycott the project.
It is not the points. I have campaigned for a work based points system for months.
These benchmarks are a joke and we all know that.
XS output dropped from 515,000 to 380,000 yesterday and a lot of the machies are remote.
Our #3 guy windforce is out of the country on vacation.
If we don't settle these issues with Baker by Monday I expect to see our output drop to under 150,000 as DDTUNG pulls his machines.
Beleive me, we don't want to stop Rosetta but I can't stand by and let this happen with no support from Baker. By saying nothing, he gives the "whiners" credence.
About the small crunchers: I have no issue with a small guy, what I have issues with is a guy like that T.Castro and that guy Saenger that have rosetta installed but essentially don't run it.
I checked saengers stats: Since last May he has run rosetta a total of 8 hours and yet he is posting on the rosetta forum many times daily. His agenda is not Rosetta BUT in making Rosetta into what he wants it to be. In my eyes a guy like that hasn't earn the right to make complaints. It's not the one PC, It's the one PC that does nothing.
Thanks guys for having me here and the warm welcome.
Movieman

[ Voor 16% gewijzigd door Verwijderd op 20-08-2006 09:41 ]