The Four Big Folding Sins

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

The Four Ways an Honest Folder Can Go Wrong with [email protected]:

1) Post up clients or Work Units on the net, for others to download and fold.

This is a violation of Stanford’s license and EULA with you. It also invalidates any work units that are returned, because the audit trail is broken. If it’s extensive, the project will have to be redone.

2) Store completed, not yet returned work units, for any reason.

Oooh! Makes Stanford livid! Dutch Power Cows and a SE Asian team did a little of this, some years back. Just a little, but you should have seen Stanford’s hackles go up, up, up!

[email protected] is entirely time-sensitive – Stanford needs those completed work units back ASAP.

3) “Cherry pick” work units, so you fold only the ones you like.

Changing [email protected] clients for better PPD (Points Per Day) is fine, but don’t delete any type of work units your client’s server assigns to you. For example, I want more points, so I change from the classic one core client, to the SMP client – that’s what Stanford wants you to do. Now fold whatever SMP work units you are assigned.

However you do it, and however you rationalize it, Stanford has had trouble with folders who “Cherry Pick” their work units, and has programmed the servers to help prevent it.

This is a sore spot with Stanford [email protected] You don’t want to go there.

4) Fold on computers you don’t have permission to fold on.

May constitute theft of computer services, and tarnishes the good name of Folding at Home. We’ve had one member lose his job over this, despite verbal permission. Get a written OK if it’s not your system.

For any of the above, your points may be zeroed out by Stanford.

Originally posted by Adak, in this thread.


Leave a Reply

Your email address will not be published.


  1. Awesome!!! An extension of Adak's post and made a front page article. Now my dilemma... link the original post, or this new article post.
    What say ye seniors (there seem to be a lot of us now-a-days)? Since we've now been "empowered" with front page capabilities (thanks I.M.O.G. and deadly), should I be linking things in the FAQ and Intro to threads or to front page articles?
    Will the article remain a good link?
    I vote for the article it will be good for the long term, as a link. It uses better fonts and colors.
    If the article will only be temporary, then use the forum post as the link.
    I think the article link will be permanent. I would say use the article link, as I think the article looks a little more organized than the forum post.
    K... links have been added to the Intro and FAQ. :)
    Check them out... let me know if you think they need any adjustments.

    The links are working perfectly, h. This is some good information for everyone new to folding, and some of us "old timers".
    Has any benefit come from [email protected]? Just curious. I haven't seen anything relating to actual outcomes due to this process.

    Right off, a few things:
    1) how to go about creating the world's most powerful supercomputer for analyzing protein folding, from a distributed computing network.
    2) verifying the results of FAH clients simulations. This is no easy task, and may involve months of folding before enough results are returned to make a judgment possible.
    Keep in mind that in the early days of FAH, the P3's and such, did only about 1/4,000th of a second in the folding simulation, per 24 hours of folding.
    3) the FAH peer reviewed papers (a very serious matter for any scientist), have detailed lots of protein structural info. All of FAH's data is available to researchers, btw.
    4) before FAH, the drug companies did their own protein analysis, but that data was (and still is) proprietary, and shared with almost no one. It is also very incomplete. Since FAH began, several other protein analysis distributed computing projects, have begun. They have seen the benefits.
    5) FAH began as pure research. Forced, since it had collected no data yet. Now, with all the faster PC's and GPU's, and PS3 clients, it is starting to zero in on some of the critical proteins that cause disease, and others that protect us from disease, using that data. Particular attention goes to those diseases that are directly protein folding related (as far as we know). Alzeimers, Parkinsons, Huntington's, and the human version of Mad Cow Disease (Crutzefelds-Jacobs Disease and it's variants).
    Will FAH ultimately lead to cures? You can't tell. That's how it is with science. You can see that most first world countries have established some protein folding project of their own, and you can guess that they're not doing that because they expect to waste their time and money.
    Until we learn a great deal more about how our critical proteins fold (and how they can be influenced by prions, etc., to misfold), we can't hazard a guess how important the research might be.
    When it comes to projects like FAH, if we (the public), don't support it, it simply won't be done, or be done well. What we'll get for our efforts, we don't know. What we'll get for not trying, we know all too well - next to nothing. So we try.
    I think another folding sin should be folding on unstable systems. I'm not sure how much data verification is done on the work units, but if you've overclocked your system but it is not prime stable for extended periods of time (24 hours in my opinion) then you're probably spitting out bad results.
    Sure you still get the points but the reason to fold is to help out a cause not just to rack up points.