[krove]

Per my news posting (which really was more of a discussion issue):

So what's the deal with the low number of active users in all the projects (with the exception to RC5-72)? After creating these neat-looking graphs to show quick statistics for each of our projects, I was curious to notice how low the blue bar (indicative of percentage of total users for the project that have been recently active) is on the various projects.

We need to get some competitions going, drives to get users to stay active and output for their respective DC project(s). Any ideas? Post them in the comments for this posting.

[The_Equivocator]

I guess I'll move my response from the site over to here:

Hmmm, I think that the reason a lot of the projects seem to have such a low participation is due to really old accounts in SETI and people having multiple accounts in F@H (I can claim to have created a few of them). Maybe if another quick stats bar was added that displayed the number of active participants per project as opposed to the percentage of active participants...

I think that a new bar such as that would be much more indictive of actual team participation.

[macdude22]

I used to do seti, searching for et is cool and all but not that usefull. RC5 is something that will be of benifit to everyone and that 1000 for the winner don't hurt none either.

[The_Equivocator]

Originally posted by macdude22:
I used to do seti, searching for et is cool and all but not that usefull. RC5 is something that will be of benifit to everyone and that 1000 for the winner don't hurt none either.

I hate to break it to you, but RC5 isn't really helpful... at all. It's a contest, plain and simple, trying to break an old security algorithm that no one actually uses for anything important.

If you would like to do something useful, check out the team pages on Folding@Home and dFold, both of which involve calculating the structures of proteins... a much more useful endeavor. D2OL is also a useful project, but I have not really looked into it. I concentrate completely on dFold and F@H.

[DrBoar]

Yes, left over accounts is a big reason. I started with SETI on a 6100/60 but it was so slow and the background switch was bad so I went to RC-64. With OS X more useful things like F@H came along so I switched to that and later went to D2OL. During this time I have been working in 3 different places and also changed computers at home. RC64 on a LCII is fun I have installled different clinets and have been allocated different nodes and left them here and there.

Then on top of that there are those clients that does not work with that OS or even in the background slows down a computer that from the start is marginal for a given application (games). So I think it is natrual that you have to try out different clients to find those that fit.

[bousozoku]

It's unfortunate, but not many people seem to care here. Folding@home almost constantly shows 48-53 active--there's another one this week, thankfully.

I'm not quite sure why people won't participate, but I had very little resistance (even for my G3/400!) getting into the top 100 for folding@home, but there have been 4 or 5 more people folding now than there were in April, when I started on this team.

People, show that you care and use those Macintoshes for something that could help humanity!

[Shaktai]

Well from what I have seen, it is not unusual for participation to drop during the summer months. A great many of our members are students, and for various reasons may not be able to participate full time all year long. Hey, even I had to shut down my boxes for a couple of days due to heat.

Now if we could find a way to make it more fun for those folks who don't have multiple computers, that might be good.

How about a competition that is for "single" computer users only? No farms or networks allowed? Or maybe a "macs only" competition. Multiple computers allowed, as long as they are Macs. No x86 or other platforms.

Hey, even reader50's super computer could compete then.

[jbcool]

Shaktai, in response to your previous post.

[OneMacGuy]

Originally posted by bousozoku:
It's unfortunate, but not many people seem to care here. Folding@home almost constantly shows 48-53 active--there's another one this week, thankfully.

I'm not quite sure why people won't participate, but I had very little resistance (even for my G3/400!) getting into the top 100 for folding@home, but there have been 4 or 5 more people folding now than there were in April, when I started on this team.

People, show that you care and use those Macintoshes for something that could help humanity!

I, for one, tried folding@home first when I was leaving seti. It certainly looked interesting. However, on my Dual G4/533 box, it ran as slow as dirt! I forget how long it took to finish a wu, but it must have been a week or more. Now my old Mac is not state of the art, but it is still pretty decent at most things and runs dfold pretty well. I think that their client is among the slowest I have tried. Maybe I should have tried it awhile longer and tried to figure out how to optimize it a bit better.

I really want a DC project with:

A solid client that does not crash.
Is stable enough to leave running on some of my clients cpus and not cause them trouble (with their permission of course!).
Does not require babysitting!
Runs decently on older hardware.
Actually does something useful (rules out SETI and RC5)
A long lifespan, not needing to be replaced with new versions often.

Is that too much to ask? All of these seem like pretty simple requests, nothing unreasonable. dFold does not seem like the place to stop and hang my hat for the long haul yet, should it be Folding@Home? At least seti met these criteria for the most part for several years, all except the usefulness piece....

Signed,
confused (and hot) in Georgia

[Shaktai]

OneMacGuy, I think you touched on a couple of pertinent points.

Distributed Computing has become common place enough, that people are beginning to expect a near commodity convenience from it. Or in other words to start it up and then "not have to mess with it."

The reality is that no distributed computing project is yet at commodity convenience levels. Not one of them. Distributed Computing is still a very leading edge technology. It is very much in the the frontier of computing and science, and as such all projects have inconsistencies, require frequent changes, and have to adapt and change on shoe string budgets. It has come a long way, but still has a long ways to go.

A lot of folks get into distributed computing with the expectation they just have to turn it on and watch the points rack up. The nature of the sciences that drive it though, require change, growth and adapatation, and all that on minimal budgets. They are far from mainstream and do not have the R&D budgets necessary to turn this into a commodity yet. While a few projects like folding@home have come up with some interesting results from the scientific perspective, not one has yet to come up with anything that can yet be classified as useful, other then for further study. There are no tangible (i.e. marketable) results available, that will take it to the next level.

Perhaps its my age, but my committement to Distributed Computing is not because I expect to get anything out of it, but rather because I believe it is something that will benefit future generations. It is purely a philanthropic effort. I donate time, money and resources from what I have for the greater long term benefit of all people.

It doesn't bother me if I loose a few hours or even days of processing time, because I know that is part of the development process. I expect it to happen, rather then demand that it doesn't. Quite simply, I don't expect it to be perfect, just to improve over time, and I choose to be a part of that improvement process. The teams, competitions, stats, etc, are just the icing on the cake.

[OneMacGuy]

Originally posted by Shaktai:
OneMacGuy, I think you touched on a couple of pertinent points.

It doesn't bother me if I loose a few hours or even days of processing time, because I know that is part of the development process. I expect it to happen, rather then demand that it doesn't. Quite simply, I don't expect it to be perfect, just to improve over time, and I choose to be a part of that improvement process. The teams, competitions, stats, etc, are just the icing on the cake.

I know I sound like I am still a little peeved about that last screw up they did at dfold, and I probably am. I put a lot of effort into whatever DC client that I am running and I guess I do have some expectations of quality standards on their part. I have done development for years in various database languages and even a rookie understands software testing before deployment, especially anything that M$ sells!

I put a lot more than just effort into it also. When I find a client that uses fast hardware and does not mind leaving their cpu's on 24/7, I sometimes make them special deals to allow me use their extra cpu cycles. It really hurts my efforts to then have the stupid clients start bombing their computers out within a week of install. I even sell some folks computers that I build at discount rates if they promise to run this "neat screensaver" that I leave on their computer. Good deal for both of us, I get the dc credits, they pay the power bills and they get a well built clone at a great price. I still have people running SETI that I do not even know how to contact and change over!

I switched from SETI so that I could feel like I was doing something worthwhile with my efforts. But I do enjoy the competition also. I also enjoy reading the thoughts of you younger guys on how to do things. Pretty good crowd of guys! Some pretty sharp guys, too. Someone always willing to help with an answer, amazing how much Scott knows! Not to metion Shaktai, Welnic, jbcool, Reader50 to name just a few!

I have enjoyed all of the forums, all of the competition, being a part of Team MacNN!

[jbcool]

Shaktai and Onemacguy you both have very valid points.

I understand that we are making a donation to the DC programs and may remove our resources from and DC program at any time. So we do expect that the clients should be as trouble free as possible since we are saving the research labs thousands of dollars each month. If there is and error made like d-fold made this last month then the should expect some repercussions for letting out a bad client. I made my statement by moving all of my machines over to f@h.

The stats being icing I disagree with and see them as more of a motivator / inticement for the dc projects to gain more spare cpu cycles from those willing to make donations from their machine.

I understand and agree with onemacguys frustration with the unstable clients. I also understand and agree with most of the points that Shatai brings to light.

Shaktai if I had a view out the door like you have I probably would be more melow in my senior years also. Must be the fast pace life here inf DAL/FT. Worth that keep me on edge.

Ok let me kick the soap box out and go do this some more.

Great people, great team, great help, great ideas.

[DrBoar]

On the soapbox!

As a scientist I see distributed computing computing as kindered spirit to the trend that more and more informantion becomes freely aviable. Raw data like DNA databases and tools as well as more processed stuff like scientific publications. A peek into http://www.tigr.org/ and many other sites I hope will at least show a glimpse of the vast fields that we contribute to

The 1800s was the breakthrough of mechanics with the steamengine in boats and trains and the combustion engine and the rollerbearing as some high lights. In the 1900 the electricity and electronics spread like wildefire for ever changing our life. 2000 will be the breaktrough of molecular biology and the remifications this has on the sciences of drug design, evolution, genetics, environment, medicin in general, agriculture and many other fields

Stepping down from the soapbox.
Now I will increase my share od D2OL by installing two clients on my dual box and kill that residual java stuff on my 7600/200

[bousozoku]

OneMacGuy:

I'm sorry if I sounded mean. Of course, you're right. The folding@home client is not extremely fast. Even the latest production code is not as fast as it might be. The Tinker code still seems really slow on a G4 but the Gromacs code (the newest) seems to age the G3 a lot because of the complexity of the work.

I guess I'm thinking about it from a big system standpoint--the kind of systems I've programmed and administered. I just set things and go and they work. I've yet to have a problem with the console version for Mac OS X.

Shaktai:

I remember a lot of people expecting to instantly be vaulted into the top ten! I also remember when I started that I spent a lot of time on the 0.6 point work units for folding@home.

Lately, I've noticed the dual 800 is really racking up the points (for a single machine), so I think there have to be people with better machines than mine who could be making the big points and be vaulted, not instantly, into the top 20.

[Chinasaur]

Hi,

You need front page pimpage. Ars has resorted to it to boost it's numbers..just recently against FreeDC in it's bid for #1.

A prominently displayed link or button on the front page of MacNN wouldn't hurt either. The Team MacNN forum is buried near the bottom of the forums list so it's out of sight. Get it moved up higher or get another link and watch your membership grow.

Best of luck


Chinasaur
FreeDC

[DAlex]

Originally posted by Chinasaur:
Hi,

You need front page pimpage. Ars has resorted to it to boost it's numbers..just recently against FreeDC in it's bid for #1.

A prominently displayed link or button on the front page of MacNN wouldn't hurt either. The Team MacNN forum is buried near the bottom of the forums list so it's out of sight. Get it moved up higher or get another link and watch your membership grow.

Best of luck


Chinasaur
FreeDC

Excellent points - I second that.

DAlex

[SciFrog]

Me too.

[Scotttheking]

Someone start writing.
I can get it posted if you can write it. Just make it in the format of news posts, small blurb, then full story. Macnn is setup for snippets, while ars is setup for full stories, but we can link to the full story from the snippet.