Well first of all, sorry if this is the wrong place to post this, I am still a bit overwhelmed with the complexity of this whole "sharing regular pc's in the name of science/research". In fact, I only just started a few days ago and only have 20 credit points after doing one WU but I digress.

I actually first heard of folding@home and wanted to give it a shot (as it seems like a great idea!), but I was having difficulty installing it on my computer/had trouble making sense out of it all and then looked around a bit more and found out about BOINC and it just seems like a more user friendly application.

So what's up with placing Folding@Home here in the retired section? I mean, was folding@home once part of BOINC? If so why is it on its own now? Why isn’t folding@home placed under the BOINC client? I would have thought it would be better to have one universal client where all these sharing projects can be attained from. Is there rivalry between folding vs. boinc? I think they are from different universities but I am sure they can merge to make their projects better/bigger? Most (if not all?) the projects that are running on BOINC have the results go to the public domain right?

Anyway, any info would be appreciated.

Cheers.

Folding@Home predates BOINC; they did run some testing with a BOINC client, which is where the Folding stats come into play when looking at BOINCstats. However, they decided that BOINC wasn't the way to go, and have gone back to their own client. For the why's, etc., you would have to ask the Folding team.

Of course, there are probably other users here with more information than I--since I never crunched with the BOINC Folding client, and haven't done anything with Folding for a very long time (2002?).

Thanks for the very quick reply!

I managed to find this at the F@H forum:

http://foldingforum.org/viewtopic.php?f=15&t=8604&hilit=BOINC

I have to say, I am far more convinced that BOINC is a better client than F@H client, simply because of the easier set up and the multiple projects to choose from.

Hopefully in the future they can join under one client (either F@H goes to BOINC or vice versa) but in the meantime, it seems unlikely.

Cheers and thanks again!

Wow, they certainly seem to be opposed to BOINC for a number of reasons... None of them steadfastly convincing. Pity, I would consider going back to Folding if they had a BOINC project, and would have participated in the beta had I known about it at the time. Folding was one of the projects I bounced around to after SETI, before WCG and then finally BOINC and all projects thereunder. I had also considered using the Folding client for my ATI X1600, but I think I'll wait for a BOINC-based project to come out with a client first, since I have grown attached to BOINC as a means of tracking my contribution to/progress with DCP.

Although the contributors to that topic claim they have all the answers, they still insist that all BOINC projects must use project applications that are open source, which is hardly true. They also question BOINC's GPU power, which is rather confusing--since the power of an application on a GPU would come from the coding of the application, and not that of the BOINC manager.

Oh well, I'm not really active with biology projects right now anyway.

Very true noderaser, it seems they have little concept of what BOINC actually is. Reguarding their comments of comparing science with Rosetta http://boinc.bakerlab.org/rosetta/forum_thread.php?id=4469&nowrap=true#56733
Still Folding@home does beneficial science so I wish them well, I personally will not crunch for them unless they change their minds on BOINC

Nah, folding pre-dated the BOINC app, but not necessarily SETI@home classic. Back in the day, there were a few stand alone apps, seti being one, folding another, and world community grid had their own. But each of these ran on their own client.

The concept is much the same (peeps can use their computers to run something, when they're sitting idle), but getting all the stand alone clients to play together.

Well here's the downside for folding@home for me, now... The door's been opened, and people don't have to contribute to one project only. Fact is, someone can chose (perhaps due to interest, perhaps just as people chose to compete on multiple fronts, perhaps for whatever reason) to contribute to more then 1 project. Now for some projects this has a benefit (projects like LHC that didn't have a consistency of work are doable, without leaving CPUs starved for units; but for the user it also gives some freedom to process multiple things.

I'm sorry, but once this door is opened, I wouldn't want to be left having to crunch for 1 project only (any project); and any project that doesn't give me the flexibility to crunch multiple things (proprietary client with no projects list to chose from) will be the loser. Some might think "well if some can crunch something else; they'll spend less time on our project", but tbh the opposite can also be true "if someone else offers choices, then the potential contributor might chose to have the option to crunch other things also"

There was a time I ran the folding client and BOINC together, but it's also more administration, and was a bit more work on my part. And yeah, I might have a degree in computer networking, yeah maneaging software isn't a matter of can/can't do. It's more a matter I don't want to have to spend the time to micro-maneage CPU shares to make sure both apps get to run with whatever resource share I want, through doing the math and throttling CPU use appropriately, etc... Then (this was back in the stand alone client days), throw a third project in, and more local administrative tasks. Now, meh; it's more time then I want to put into setting it all up. So if folding@home decides to run a BOINC client, yeah I'll sign up (or merge my account in from folding). But until then, best of luck, someone else will (and is) crunching for them. Now if BOINC wasn't in the picture, then world community grid (aside from them having a BOINC project, also has multiple projects one can opt into); which would still leave the stand alone single project a choice one wouldn't have had to make.

They're not retired though, just continuing on their own.

The "retired" portion of Folding (as found here at BOINCstats/BAM) was the BOINC test they did a few years ago...

Wouldn't it be possible to use a dummy BOINC project to run Folding under BOINC. Possibly by using the BOINC wrapper.

There are of course several caveats but for a real Folding addict it might be a nice exercise to try.

It was suggested several times to yoyo to include it in yoyo@home, but he rejected the idea, because FAH has enough users already and doesn't need the support of a BOINC wrapper as much as for example evolution.

The quite anti-BOINC behaviour of the major participants in the support forum of FAH doesn't make such a move so likely as well.

Yeah, they did a closed/limited alpha test. But signups were extremely limited, the site for it was private to the point they didn't even want people to know the URL, and it never really hit a stage where, well tbh I'm not sure I'd really say that it as a project ever really and truly began Any mention to it came third hand or in response to queries some of us (myself included) made on their forums, and seemed more a token effort (for positive PR IMO), then something which was being geuinely pursued.

Even when a closed/limited alpha was being mentioned, they seemed to have certain thoughts against either going to BOINC or even making it an available option. My sense is that it was more a concession to some crunchers wanting a BOINC client for which it even got that far, rather then a genuine interest in having pursued such a client. Their position, rather constant to what it is now, is that they really always wanted to remain on their client. I do remember the limited alpha, along with attempts to hunt down the site, coupled with postings on their board which I'd personally made (both before and during said closed alpha). Following one URL that supposedly got leaked out; really only mentioned news about "yeah we're looking into it" but didn't seem to be an actual projects page with signups etc... It was a long time ago to take from memory however. I never really got the sense that this is something that sparked the largest interest there; outside not wanting to say straight out "no way in hell" when some of us were asking for such a client on their boards. At least this is the sense I was left with; and though I followed to see if this project would ever really hit even a visable state from time to time beyond mention of "yeah there's something we're looking into now", I also, in the back of my mind never really expected it would.

Partially due to this, I left off crunching FAH a fair time ago. They've got crunchers still, so are fine in that regard. I just really don't like being told "well if you're going to run our project,, you need to use a client that can only run it exclusively; else you must go through hoops to make sure all the projects get CPU time" when such isn't needed. To each their own in this regard, and to all how they want to allocate their own CPU.

To be fair to F@H, if memory serves me, their concern was the latency that the BOINC client introduces, namely that a project may be suspended to run another.

Their argument was that since each new WU depends on the results of an earlier one, they needed a quick turn-around time in order to keep the WUs churning.

I guess that one may conclude from such requirements that F@H needs to have the CPU for itself as much time as possible. IOW, it cannot (or doesn't want to) afford multiple projects competing for CPU time as BOINC does.

I also had the impression that Berkeley was unwilling to accommodate such needs in the BOINC client, which may have contributed to BOINC being shunned by the folks at Stanford.

It's really a pity, because F@H is a very well-run and worthy project. But, as others pointed out, managing multiple kinds of clients is too much of a pain when BOINC doesn't lack as well-run and as worthy projects.

HTH

I participated in the Beta-test as well. I wasn't originally invited, but sneaked in and got iirc the last slot in the secret part of the forum to discuss the test. Discussions of the test in the open part of the FAH-support forum were either moved to the secret part or flatly deleted. Users that came after me could thus only crunch, but not respond.

The wrapper technology was not quite developed than, they sent a bunch of dummy-WU to the clients that tried to contact Stanford for real work once they were "crunching" If no network activity was possible the WU idled for an hour to wait for contact. This was reduced to 30 seconds later, but it never really worked.
The Credits were just points from FAH, the conversion factor was never really established, it should gave been something about 2 (i.e. BOINC-credits should be double amount of FAH-points).
The deadline concept of FAH is not what's expected by BOINC, a WU should be sent back at least after 10% of the time to deadline we were told. But they as well refused to set the deadlines simply shorter for BOINC to accommodate their need for speed.
The account concept was as well not really compatible with BOINC. There was no such thing as accounts in FAH. You were identified by name and team, there was no check whether anything was double.

Finally the one person involved in the test in the lab, a graduate student, moved away and the test got abandoned. The final post by Vijay Pande in the secret forum on 28. August 2006 was:

I would argue that WCG is attempting to do something similar; by branding the BOINC client that they distribute (with WCG name and logo), new users aren't really introduced to the functionality of BOINC beyond the WCG aspect unless they find out elsewhere. That, and they are notorious for not playing like the other projects.

Could it be that FAH is more project-centric than WCG? Surely not!

As far as I'm concerned, it was all a PR wash; or a pleasant way of insisting "you give us all of your CPU time or you get no support". They never wanted to support BOINC, because they didn't like the idea people could run multiple projects. Thing is, in terms of public relations it can be good to put pleasant language to it, and give a face that will make potential volunteers happy. I never did believe they had true intentions of supporting a BOINC client.

And to their stated argument, Milkyway's algorithm is such that new WUs do depend on the state of one's previously turned in. In fact, the results turned in, generate new WU's based on their own blurb on their website for Milky work. Difference is, Milkyway went the route some suggested; just impose short deadlines. Taking away the freedom to crunch multiple projects was never considered a requirement from Milkyway's perspective.

All's fair though. They had chosen to be (IMO selfish) with potential donator's CPU time. In turn, I eventually ended up detaching from/uninstalling folding@home, and chose to not crunch for them anymore, at the requirement of micro-maneaging multiple pieces of software. They can chose to only run a stand alone client for purposes of not permitting potential crunchers a choice of what projects to run along side (perhaps using deadlines to their advantage also). But in consequence, if I'm left with the choice of "give us all of your CPU time, or none" I can chose the latter, and had

True, we're talking CPU time rather then money; but I'd say much the same if one charity made it a condition of contribution that a portion of any potential income can not be given to another charity, while they recieve donated funds as well, on grounds "well we might not get money as fast to advance our own project". It can be an important cause also, but in the end it would be like "OK, well under those conditions, I'm sure someone else would be happy to recieve said assistance and funds. I never did expect such a condition of "can't give to others also" to be part of the bargain I really just don't like such conditions or stipulations in giving, when it's mine to chose who to give to, etc... I just never could be very happy when it's put to me like that; when I'm the one being grascious enough to donate in the first place. Sorry but, well there's others who give, so they're OK without me...