Ars Tech
It looks like the leading producer for Ars Technica Team Egg Roll is nearly ceasing production, which is good news for us (bad news for project unfortunately). He is down to 12 processors but has been barely turning in work units the past couple of days. I've don't know this person nor have I been to their forums, but this will make it easier for us to overtake them. Unfortunately, though, less science is achieved with his departure if his production decline is permanent. Anyone know the scoop?
KingFish
KingFish
0
Comments
Troglodytes has some gripe about stanford or something. I don't know. It sounds kind of petty and goofy to me... Like, "Nyah nyah.. I'm gonna stop doing MEDICAL RESEARCH so that you make it easier for me to keep things running and get mega points" .. "Nothing like one of the top 6 producing folders in the world putting his foot down" one of those guys said.
I don't know the whole story, so I'm not going to judge, but if it smells like crap.......
He's saying "I'm willing to donate my time and my resources and my money in an effort to help a cause I deem worthy. Unfortunately, by your inability or unwillingness to supply me with sufficient work I am going to take my machines to another distributed computing project I also deem to be worthy but will allow me to help to my fullest extent. Once you change your system so that more work can be done (read: I am actually able to help you to my fullest extent) I will return my machines to process work for your project."
I don't care how inefficient he deems his production to be because of whatever he feels stanford is doing wrong - every single bit of folding that he DOESN'T do just hurts the overall project. He's making a "statement"?? What statement is there to be made? You either help with medical research or you don't... And he decided not to.
You say, "but this is medical, it is more important" and I am inclined to agree, but they are not my machines and they are not your machines. They are his machines and it is for him to decide whether or not SETI is just as important to him as Folding. In fact, it seems to me he might be inclined to agree with you and me that Folding is more important, but he doesn't seem to give it as much weight as you and I do because he also deems SETI a worthy alternative, whereas I do not.
But, whatever. I am not going to speak for him because I am not him. I have not been authorized to do so. I have never talked with him. I do not know his feelings. All I can do is offer up my interpretation in regards to his comments.
I do hope he comes back, though, because I, like you, think that Folding is more important than SETI, but it isn't my call to make. I do not own the machines. I do not pay the bills. I did not take the time to set them up. I do not maintain the systems. I haven't even been consulted for advice. Since none of that has been met I believe I have absolutely no say in the matter.
I do not agree with his judgement. Taken a little further one could argue that it would be reasonable to abandon the project just because your SSE-capable comps had a week of nothing but Tinkers.
I was put off by the whole tone of his post. His "...and Stanford gets a clue" comment was the icing on the cake. It sounds as if he thinks things would be better if he were running the whole show. Team 93 is successful because we are a team. Sometimes that means running the play the coach called, even if you don't like it.
Thank you, prof.. Those were exactly the words I was looking for. It was his arrogant and condescending tone that irked me. I could give a **** what some dude does with his computers. But he made it sound as if because he has 30 computers that he has some sort of "influence" over the direction of the project. It's the same tone that a very very rich person has when they are talking about city planning or something ... "Well I don't think that we need a park there. I think it should be a gas station instead.. And I AM 30% of the taxbase for this city, you know...."
If anyone wants to comment, at least understand the full background And Read his reply above.
In short, his 2 reasons were the inability to keep the machines fully occupied with wus and helping TLC to maintain the lead in that project.
What you preceived as important may be worthless to me
Personally I deem wu deletion to be far more harmful to the project, and a stroll through the top 50 teams stats will show this to be much more prevalent nowadays.
Look, as I've said - I have no issues whatsoever with what somebody wants to do with their computer(s). What I don't appreciate is the condescending tone he took. The whole "whereas..." bit was trite and unfunny. The whole "stanford gets a clue" indicates that he is exercising some righteous intellectual superiority over the entire project. I have a large folding farm, also in a "production" environment, and I run them very simply - no -advmethods, no -forcesse, nothing beta, just stable, tested, and normal. They may not run as fast as they possibly can, but they are still contributing valuable science to the project. If they do 1 WU a month because that's what stanford is sending them, then so be it.
But seriously, I am having a lot of problems with game #8093 in Free Cell - and heck I can't even find my Seti stats. :confused2 Jack
KingFish
There are calculations that can only be done using Tinkers right now, so we do them. I agree that I find dumpers even lower on the scale, but at least they are still working.
KingFish
I would take exception to that! When Stanford called me and asked me to buy a bunch of parts and stuff to start folding at home I said ok and I did it - but I didn't sign any agreement to do, although I may have agreed to the readme file ( I can't remember - I always agree ) I am here with this team because Project or Team leaders or Founders feel this way. That can and has caused some major problems. A little power can be a dangerous thing. Jack
KingFish
From what it seems, he doesnt like the 80% performace fraction thing.
The effectiveness rating is based on a minimum complete average within one 5th the turnin acceptance time. Tinkers do help folding, they can be used for whole new directions thta need lots fo boxes wporking to get an idea as to what if anything is good about the direction. they are also cheaper to generate.
This guy or gal is worried about points, Folding at Stanford wants results for a lot of hurting folks. Some of these studies are based on need to treat existing cases where options are avilable but the case has not been treated with confidence of success over 50-60%. Folding COULD save lives, will infuture, and could randomly do so immediately.
Some of the universities that sponsor are doing active medical research, and need more than last-gasp paliative-only treatment (called dampen the pain as patient is untreatable in reality). My father ended up in this last class, he was under hospice and doctor care and heavy morphine doses for months before he passed away at home. Our doctors came to the house. Dad could not get out of bed, could not be transported, toward the end. Nurses came to the house. Folding is a battle to end as much of that as possible as fast as possible. Points, smoints, the goal is why Folding even exists and the distributed Folding is Stanford run. They get to choose, they own the network, and IMHO this is GOOD move for speeding up actual medical results.
A P4, 1.8 GHz, can meet the minimum effeiciency rating for Gromacs with ease, so can a Barton. BENCH was on a PII or PIII box. My Barton has an average of .984... effectiveness, and I use it a lot. I will accept .970 bottom from it with some WU series I have gotten, best has been .989... It averages turnin in less than 2 percent of turnin time. So if boxes cannot do it in 1\5th of turnin time, etwas loss ist (something is WRONG).
I wish troglodyte luck, but trogs can be dense even in fantasy, so let him go if his boxes will not fold the way HE wants to versus how Stanford needs to have the clients fold these days with increasingly complex gromacs work. The skew is toward getting folks with faster boxes to get points faster-- That is reality, they SHOULD. The boxes are being rewarded for work accomplished, and the user for sponsoring their running of clients on boxes they pay to power and buy parts for, etc.
That is my dime's worth.
John D.