• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

it looks like somoe Teams are planning a work stoppage

latest News

I have just recieved an email from Crunch3r, and he is suggesting that we wait for a while with any protest strike. He is in direct negotiations with Rom Walton and Dr. Anderson. We need to wait at least until the middle of next week before organizing a mass protest, so that the negotiations between them can take place. Crunch3r said that the whole of Planet3DNow team will also support the strike, IF it is necessary. So, for now, we need to just hold on until Crunch3r gives the word.

Regards, Daniel.

http://www.setiusa.net/phpBB2/viewtopic.php?t=2152

regards from Germany NRW
Sir Ulli

 
I don't think it's solely a "gimme credits" issue - IMHO it's more of a rallying point for a number of reasons, amoung them the one above plus the peceived treatment of Cruncher, and the negative stance on optimized apps. I rather suspect that every project is going to go through a similar turn, hopefully, to a much lesser degree of negativeness than this one has. To me, it makes perfect sense to develop applications to take advantage of the instructions available with certain processors that aren't available in others, such that some processors are "more equal" or "less equal" than others. As long as it's first and foremost the science that benefits, I don't see a problem.

The Einstein@Home project went through a whole turn of this but they ended up choosing to embrace the concept of optimized apps and actually hired the guy(Akosf) that was voluntarily doing the programming! 🙂 A win-win solution for everyone involved - always a good choice. 😉

 
Originally posted by: networkman
I don't think it's solely a "gimme credits" issue - IMHO it's more of a rallying point for a number of reasons, amoung them the one above plus the peceived treatment of Cruncher, and the negative stance on optimized apps. I rather suspect that every project is going to go through a similar turn, hopefully, to a much lesser degree of negativeness than this one has. To me, it makes perfect sense to develop applications to take advantage of the instructions available with certain processors that aren't available in others, such that some processors are "more equal" or "less equal" than others. As long as it's first and foremost the science that benefits, I don't see a problem.

The Einstein@Home project went through a whole turn of this but they ended up choosing to embrace the concept of optimized apps and actually hired the guy(Akosf) that was voluntarily doing the programming! 🙂 A win-win solution for everyone involved - always a good choice. 😉
Click to expand...

Just wait about 3-4 weeks and watch the fireworks at Einstein. The S4 data as of this morning only had an est 19 days left. Then will be the switch to the S5 data and I presume a new app. The new app has supposedly already been tweeked by Askof. So the credit hunting there will slow or stop.

Deja Vu anyone?

 
Originally posted by: RobertEJust wait about 3-4 weeks and watch the fireworks at Einstein. The S4 data as of this morning only had an est 19 days left. Then will be the switch to the S5 data and I presume a new app. The new app has supposedly already been tweeked by Askof. So the credit hunting there will slow or stop.

Deja Vu anyone?
Click to expand...

I really don't see the same thing happening at the E@H project because it's a wholly different environment there; Akosf is now working for the E@H project admins - optimized apps are being released by him and this is endorsed, authorized and encouraged by the project admins! It's a completely different situation.

The only down-side I can see is that if the message threads at other team's sites are any indication, there could be a large influx of users and demand for work that the E@H servers and infrastructure *may* not be prepared to handle. We shall see.


 
Heres what I see. Hope I'm wrong though.

A user can run an optimized app and standard client. In about 3/4 of those cases they will be granted 2-5 times their claimed credit. Nice exploit of the credit system.

Now when the new app comes out, assuming its highly optimized, the claimed vs granted credit difference will be nearly gone.

Now compare it with S@H with optimized app & calibrated client and we have a stiking similarity to the credit debate now. Crunch3rs issue aside.

Hope I'm wrong, but looks like deja vu to me. 🙁
 
But the thing you're not taking into account is the use of a quorum in the E@H project.

Currently, 3 (or sometimes 4) results are needed, all of which agree on the same result. Then the Claimed Credit values are added together with the result being that the Granted Credit is actually the average of the 3(or 4) results.

So if User A and User B are using un-optimized clients and submit results claiming 10 credits each, and User C who is using an optimized app puts in a result claiming 40 credit, the average ends being 20 credits Granted to each. In this scenario it would appear that the guy running the optimized app gets less what is claimed, which is true, however, User C does get a savings in time because his app IS optimized meaning he can complete more workunits in a given time than either of Users A or B.

Without the use of a quorum, yes, optimized clients give a distinct advantage over non-optimized from both a time and a credit perspective. But, thankfully, the E@H project doesn't operate this way and requires the use of quorum for results to be granted, which is why at the height of my project production I often had as much as 14k of Pending Credit - because those results were waiting for others to finish crunching and submit their results.
 
I am taking into account the quorum. In you example your numbers are backwards.

Using stock client and optimized app will result in short crunch times and lower claims. But in most cases considerable higher granted.

Take a look at this result of mine.

Or for that matter one of my hosts.

More often than not, I'm being granted 2-9x what I claim.
 
Ah, sorry I was thinking of a different project. 😱 But my point is still the same. Take your 1st link, the first user(presumably you, although it says "Anonymous") claims 7.28 because your app/cpu took so little time to complete it. User B claims 126.46, and User C is claiming 67.73; in this instance E@H drops the low number and the high number and Grants the one remaining, ie. 67.73 which shows that all is working exactly as it is supposed to be. You get a nice bonus for using the optimized app.

To which I'm still saying, so what?!?

The project is working EXACTLY the way it's supposed to with the full knowledge and blessing of the E@H project founders.

So color me confused about your point. 😕
 
Nope, your right on track with my point. Hang in there I can be a little slow at times. Yes, the host in the first link is one of mine. Nice bonus for me in the credits.

Now, take that exact same example, but replace the stock client with a calibrating one (trux or bonicstudio). Now instead of me claiming the 7.xx I claim what the stock app would have. But I've taken a fraction of the time. That would inflate my credit claims even more.

This case is a good part of what all the bad blood is about over at Seti.

Before enhanced, one could take an optimized app & calibrating client. Do the same WU in 1/4 of the time. Still claim and be granted the full credit. Basicly inflating their scores by a factor of 3-5x.

With enhanced for the most part that loophole was closed. People were pissed and felt ripped off that they wern't getting the credit they were used to.

Now throw in some overzealous mods, deafining silence from the project admins and egos the size of TX and a spark. Kaboom.
 
If people are allowed to get away with something for long enough, they tend to get VERY uptight when they can no longer get away with it any more. Just look at the illegal immigration debate going on in the U.S. right now for a good example... 😛


Personally, if other teams really think they need to go on strike from the SETI project, more power to them. The beauty of the BOINC system is that it is very easy to run any project you want to. If you don't like a project, just attach to a different one and life goes on.

Besides, if enough of the top teams stop production, it will give our team a chance to catch up. 😉
 
Okay, this makes a little more sense to me now.. the part I was missing was the part about the claimed credit being artificially inflated -- that I have never done in any project, and I can well understand why many would consider it to be cheating. The E@H folks would never endorse artificially inflated credit claims either.




 
You must log in or register to reply here.

TRENDING THREADS

Back
Top