Tinker (NO s at the beginning Paleeze!)

[witenoiz]

:bigggrin: I rezalize the thinking behind the anti-Tinker folders :woowoo: but shoot-fire boys and girls I just love them :usflag: My little farm is folding 17 right now and in a few days the results will be an added 1,276.20 credits. I have been a member of this wonderful group for less than 3 months :rarr: Oh, I know it seems longer, :bigggrin: Athlons and Tinkers are a match :sawed: What if I had these suckers Over Clocked to the MAX :vimp:
I need a Logo Stable Farms - We Fold :celebrate Our mechanic is the Maytag repair guy! :cool: Jack

[mmonnin]

But in the same time you did that 1.2k you could do 1.6k+ with gromacs.

[primesuspect]

Who cares? Stanford needs tinkers folded just like they need gromacs folded. Thanks for your contributions, witenoiz. So, are you gonna lay the specs for your farm down on us? Inquiring minds want to know ;)

[mmonnin]

I am doing one now and have done several in the past few days. The project just gets WAY more out of one gromacs WU compared to one tinker WU. A LOT MORE.

[a2jfreak]

Just imagine the kind of points Jack would be posting if all his machines used -advmethods and -forceasm/-forcesse--depending on which client is used.

[shwaip]

Just because protiens are worth less points doesn't mean they don't need to be done. Folding is not about the points, but the results of the work we complete.

[a2jfreak]

True. That isn't what I was attempting to imply (if you were responding to my post). I was just thinking it would be neat for Jack to start to give the big dogs some competition in the points dept. :)

Just because protiens are worth less points doesn't mean they don't need to be done. Folding is not about the points, but the results of the work we complete.

[mmonnin]

More points usually mean more work done for the project.

Something new is about to happen in this area but I am not allowed to spill it.:) My lips are sealed.

[a2jfreak]

/me dangles as Tootsie Roll Tootsie Pop infront of mmonnin.

Those lips still sealed, buuuuuuuuuuuuuddy?

[mmonnin]

Hehehe actually look for an announcement concerning something else tomorrow. It prolly wont effect most of you guys tho.

[Straight_Man]

And, some of the Gromacs build on Tinker work. Tinkers can be used for first trials.

Gromacs do the multi-variable playing and Tinkers can do first-look-at-idea tries.

Points feed competition, which gets more work doen if new folders are not discouraged by low pointage. If it works for people turning in more, that is good-- if not, competition can lead to folsk turning away from folding right after they started or because thye do not know how. I fold for this team instead of my own, because you folks taught me how, and watching the OC discussions made me learn something about OCing. right now, temps are too high to OC my Barton, I will probably have two P4 boxes within two months as the P4s are more stable at higher temps. And, they ROCK for Gromacs folding. Were I not on this team, I would still be folding, but pure points do not motivate me worth beans.

I am trying to point out what works, not what does not get more smaller boxes doing base work for Gromacs-capable boxes to continue. There are lots of Universities with smart folks who license the cheapest way to get base research done, and folding is one of the majors for Cancer research and altzheimers and even the avian cancer that in humans results in a nasty fluish thing that puts older folks in hospitals.

What Witenoiz is doing is to help do some of the base research. What the big fast boxes will get is more and more complex things, seen that trend all the time I have been folding. Gromacs is one way to take combo treatment ideas and pre-evaluate them so what is pumped into folks trying to kill somehting deadly is better pretested instead of, as with my dad, he ended up having to be used as a guinea pig as much as have a real chance to live. He did not live. He was :screwed: by something that was not of human fault in a blameable sense.

He got chains of recurring tumours, they worked very close to spinal column right inside back of his jaw, adn surgery would have meant removing part of jaw and since tumour was large and also nested right next to spinal column of neck, in front, part of his tongue and part of his mouth would have to be removed. He was evaluated, thought a good choice for a near-tumour infusion with chemo that was much stronger and toxic than the nomral chemo or radiation could do without killing him literally with the treatment itself. This was a research project. The place also had a skull-base surgeon for the alternative.

Thus, I fold so this does not have to happen to someone else, and I want something better SOONEST! I still :cry: inside when I think about that as I helped care for him at home for a long time. I still remember him near the end when he was confined to a hospital bed, had to be elevated at a 45 degree angle and his tongue was swollen from tumour growth inside and behind it. We went to injected Morphine to keep the pain minimized. He was unconscious for the last month. He planned his own funeral 9 months before dying. My grandma also died in my arms, literally....

She had lung cancer. She went fast, thank God, and lived to her 90's. That was mom's mother.

Dad died about 3 years before his mother did, he was 71, she died at 102. His death broke her will to live. Hhe stopped eating mostly, minimum food we could talk her into eating.

I am doing what I do simply because I also give folding feedback, and when I asked for something like an effectiveness rating simply because slower boxes were NOT as effective with Gromacs as bigger boxes, I am sure others also did. Not for more points, but so more complex scenarios could be worked through faster by combining things that worked partially to see if combos work.

Folding is not about me, though conceivably research done in folding will lead to things that might help me some day. It is about finding ways that are not 50% likely to NOT succeed having to be sued as there is no alternative for them at the time they get massively hit with major problems humanity does not know enough about to know how to treat. I want more results in faster, that is my goal, but they must be valid results.

BTW, this is not a RANT, I am not looking for sympathy, I am not hyper-sad -- time (and faith) has let the massive ache be replaced by real here and now life coping, but the remembrance is still painful-- THIS is a major motivator for me. I WANT BETTER than my dad got. ONE HECK of a lot better, available and cheaper, FASTER working to kill or prevent, but with pinpoint accuracy as much as possible. Others feel the same way-- I know this.

Call it the "other" motivation some folks have to fold. Feel free to fold for whatever reason you choose and whatever makes you fold is good, but the end result will not just be points. Nor will it be instant, though enough branches of research are workign that things might coalesce in surpisingly sudden ways for some things.

John D-- who wants his folding boxes to be :smokin: with results that mean something to someone who needs them BAD while fighting for his\her life. :D

[witenoiz]

I will work on "Why I Fold" tomorrow. John D aka Ageek pretty much nailed it though! You guys got to lighten up a bit. I just do what is sent to my boxes and blades and noticed all the Tinkers in progress. Since I use the no-nonsense console version it is pretty much agreed that I can't add any flags! Hey, it was a lighten up comment - I think this is a fun team and I am in here for the long haul because you guys aren't messing around with 17 or Bust or Chess Brain or Seti or Find-A-Drug or on add infinity to the list of DC projects. Folding is my thing - and I do love Gromacs - 3 a day on my P4 2.8 then another Tinker. Say La Vee - I'm in the mood so I'm off to that other thread now. Jack

[RADA]

Tinkers? Gromacs? :scratch:

Dunno what that means, (Best caveman voice) They send, me fold! :D

[muddocktor]

UH Jack, you most definitely can use the flags with the console version; all I run is the console version myself on my farm. If you feel happy with the way you presently are running the clients, that's fine. I just wanted you to know that you can use all the flags with the console, just like with the gui client.:)

[a2jfreak]

Mudd is correct. Also, you can run the console version as a service too. Check out bluetentacle (http://www.bluetentacle.co.uk)

[TheBaron]

you aren't using any tags? slap the -forcesse tag on a bunch of your machines and i bet you'll see a serious improvement! -advmethods helps too :-D

[qparadox]

Sshh he's catching fast enough already ;). J/K you definitely want to try -forceSSE. It *might* cause problems on some of the athlons but the new beta core seems to have fixed most of those problems.

[witenoiz]

UH Jack, you most definitely can use the flags with the console version; all I run is the console version myself on my farm. If you feel happy with the way you presently are running the clients, that's fine. I just wanted you to know that you can use all the flags with the console, just like with the gui client.:)
I have never seen an example of how to do this - I can't get the same screens that give you the option. I must be missing something - I WILL try almost anything once. Send me very detailed instructions - I would love to try. I don't know how to post a picture of my desktop with the Fah properties box. The only tabs it has are Options Font Layout and Color. Jack

[mmonnin]

Make a shortcut of the Fah exe or use the shortcut in the startup folder or any other shortcut.

Right click the shortcut and go to its properties.

In the target line add "-advmethods -forcesse -verbosity 9" so it looks like this.

C:\FAH directory\FAH.exe -advmethods -forcesse -verbosity 9
If there are "" around the target path put the tags after the last ".
"C:\FAH directory\FAH.exe" -advmethods -forcesse -verbosity 9

There must be a space between the exe and the -advmethods.

Got any more Qs, ask.:)

[witenoiz]

Thanks mmonnin - I think that does it.
So far so good! I did all of the Intel stuff and will work on Athlons in time. Now, another question, I think the answer is yes. Do I have to start the program again? I don't think I have to re-boot. Jack

[mmonnin]

Yes the client must start with those settings to use them.

Intels could really use the advmethods and the Athlons could really use the -forcesse.

If you have EM3 I would sugges trying the EM3Calculator to see the points per hour difference. All you gotta do is click on the WU and enter your frame time. It will give you the points per hour/day/week. I think it might work if you get the Calculator and the EM protein file.

[muddocktor]

Jack, once you have started the client with the flags in the shortcut, you should be good to go. Here are a couple of pics I made of me setting up a console and also adding the switches:

[witenoiz]

I'm getting there - I never would have figured how to get to the screen I needed by using a shortcut properties route - I use FahStatsLog and it looks like I need to wait on a few of these out and do a little more tweekinghttp://www.short-media.com/forum/attachment.php?attachmentid=6993&stc=1 Jack

[mmonnin]

Wow negative 1 percent. Now that is good man.

[witenoiz]

That -1% is really holding me back :rarr: Actually that is not running, it is on the new HyperThread box and I haven't figured out how to get 2 to run without slowing the whole process down. The one running on the same box has all the flags you told me to put on it. Slow but sure I will get it right. But HAY :) I've been a Team ShortMedia member 2 months 22 days :rockon: 42nd place aint too shabby :Rocker: Jack

[mmonnin]

Heh, Ive been a member for 2 years ~5 months and I am in 20th.:rolleyes2

[TheBaron]

wittenoiz, that 2nd console will NOT work correctly on the hyperthreading box unless you have a -local flag attached to it and both consoles have a different machine ID. if you DO have it set like that and that 1% is staying there, i would suggest deleting the work unit and restarting the console

[witenoiz]

wittenoiz, that 2nd console will NOT work correctly on the hyperthreading box unless you have a -local flag attached to it and both consoles have a different machine ID. if you DO have it set like that and that 1% is staying there, i would suggest deleting the work unit and restarting the console
:vimp: Answer to The Baron - both instances are set up correctly, you are looking at the one on my slave drive and it is Machine ID 2. I have found that if you don't double-click on the little red cog to start the program it won't run. That is the case with Machine ID 1. That console will always show a -1% unless I double-click it. :topic: I even tried to send this message without entering any text - that didn't work either. :sawed:Jack

[mmonnin]

Use 2 instances of the consle. You cant really get two GUIs to work together. At least one has to be a console. The second I might add.

[edcentric]

I wouldn't mind the tinkers as long as my old 900P3 wasn't the one getting them. On my non-SSE TBird box I can about hte same points with tinkers or gromacs. The P3 dies with them. They run OK on my XP boxes, but just not as much production as with Gro.

[mmonnin]

Ive gotten like 4 or 5 in a row now on my Slowest Athlon system because the performance fraction is above 1.0 for some reason. It was like at 4 and now its down to 2.6.

[profdlp]

Can someone explain this performance fraction business?

:confused2

[mmonnin]

Its the percentage time left before the deadline as a fraction. I dont remember what deadline it is actually. I think it might be the deadline before you recieve no more points and not the one posted on the Point Summary Page.

So if you finish it with 90% left before the deadline you will have a rating of .90....I get .98 and up on my Athlon systems so its not that hard to get above 80%. 80% is where the Assignment servers will start giving you gromacs because Stanford doesnt want any delays on returned WUs.

Somehow I got mine messed up and was way above 1 which should be the max. It prolly was because I changed my time somewhere in there. I may have gotten a WU, reset the CMOS and the date reset to like Jan 03 which would make it seem like I finishes the WU before I got, thus the greater than 1 perf fration.

[profdlp]

Ahh.

So this is not bad: Updated performance fraction: 0.924888

That was from the worst of my "good" computers (I'm afraid to check the K6-2)

[mmonnin]

Yep that will do sir.:)

[a2jfreak]

Updated performance fraction: 0.000001

(I'm afraid to check the K6-2)

[Al_Capown]

Where are these numbers coming from?

[profdlp]

Where are these numbers coming from?
Updated performance fraction: 0.000001
Well, a2jfreak's came out of thin air (though it may not be far off...) :bawling:

If you set " -verbosity 9" as a flag it will show up in your log file. :thumbsup:

[mmonnin]

Or if you have EM3 it has it as well. Highlight the name, and click on the queue.dat icon and its the last bit of info. Thats how I see mine.