DC - how would we like it?

[Spacehead]

Great thread Peter :thumbsup:

Great posts everyone!



I've mostly just skimmed over these post so far but i'll re-read them later, but one things comes to mind:

I'm most familair with S@H so i'll only speak to that.
The "Do we really know what we are crunching?" question has come up before, if not here, at other forums. For my part i guess i've just come to assume that there are enough smart people out there running this project that there's no abuse(if that's the right word) going on.
Look at all the universities, corperations, goverment agencies etc. that have teams running this. Look at the sponsors, corperate & not, on the home page. Look at the funding thru foundations & grants.
I've just always assumed, right or wrong, that someone out there would look into a project before they donate thousands or tens of thousands of dollars.

 

[petrusbroder]

OK mates ... have you seen that this thread has been pinned?
It seems that it is of intrest for quite a few people ...
Lets continue in the same friendly spirit: discuss, think, post ... and think again.

Thanks for all the good input! :thumbsup:

 

[petrusbroder]

Originally posted by: Spacehead
Great thread Peter :thumbsup:

Great posts everyone!



I've mostly just skimmed over these post so far but i'll re-read them later, but one things comes to mind:

I'm most familair with S@H so i'll only speak to that.
The "Do we really know what we are crunching?" question has come up before, if not here, at other forums. For my part i guess i've just come to assume that there are enough smart people out there running this project that there's no abuse(if that's the right word) going on.
Look at all the universities, corperations, goverment agencies etc. that have teams running this. Look at the sponsors, corperate & not, on the home page. Look at the funding thru foundations & grants.
I've just always assumed, right or wrong, that someone out there would look into a project before they donate thousands or tens of thousands of dollars.

I agree. S@H is the "great-grandmother" of DC-projects and I feel quite sure, that all the donors and sponsors and funding providers check the results, the administration and the use of the monies. If all projects had a similar history all would be quite good.
I also think that corporations and universities can not afford to throw their money away on something they do not think is reasonably OK.

But S@H has one big "advantage" in this question: there is no way (as far as I can see) the results will generate money for any body. It is pure science. A similar reasoning is valid for Einstein@home, LHC@Home, Euler and some others.

How is this for those projects that may generate a product which will be sold? I am not only thinking about projects searching for drug candidates but also for the projects which simulate design of ICs or of software, or similar. Or those projects which map the internet.

What I am trying to say: What if (is this not the grandmother of all loaded questions?) a DC-project finds a really marketable product which becomes a top-seller and generates a lot of money for a company, which may or may not have been involved in the DC-project?

This is a hypothetical question. But I think that some projects should have (or have) thought about this ... what are these thoughts?

 

[Wolfsraider]

I have had a few days to mull things over,

I want to add a few points that are not addressed here.

My reason for doing dc is the fun of it, there are only work units and friendships, stats and milestones.

I want my crunching to remain fun.
Scrutinizing something just because you can doesn't make it right.There has been no sign of transgression, no hint of wrong doing. If it ain't broke don't fix it.

Now for the Side I want to address, Choice.

I used the term witchhunt before.

lets say the reasearch is done, and the results are in.

Lets also say a teammate wants to crunch a project on the bad list, how many times will he hear through pm or otherwise that he is crunching for a "bad" project, how many tempers will flare, arguments start, or attacks be made, simply because he chose a project that he/she felt was fun.Will he change projects from the pressure or will he stand his ground?

How many other forums will start to question the validity of the projects, how many projects will have to spend money trying to prove their worth, the same money we save them by crunching for them.

Pretty soon we may see the projects start to buy their own supercomputers and not need our help.

I don't know if you know this:
The Rebel Alliance has started a Rebel Alliance Team for Dimes, as we were not able to crunch as a mini team on anandtech under that project, We are in it for the fun, and will do our best to make it fun, but you will always have those that want to rule the way things are done, that want to decide for all the acceptable way to crunch, the right project to crunch, and who is allowed to crunch. I am against this, TeAm Anandtech has a great reputation for the diverse projects we all can crunch, Tas, Taboo, etc... were created to boost production, but if they are told they cannot crunch together, or if they are told that they can only crunch certain projects, what do you think they will do?

I doubt that this will remain a fun place to crunch if this is put into action.

If you look at the dhep project or the problems dpad had or how lifemapper made folks look at things differently, you will see the "picture" I am refering to.

Lastly, have you looked at production quotas lately, seti is down to half as are many projects, its just not as fun as it used to be, too much bickering, too many folks not following the live and let live rule.

Do you think this scrutiny will make this better somehow?
As I stated in pm to you peter If this does go through I will not stay here as it won't be "fun" (this isnt a threat folks, it is my choice, Dc is supposed to be fun.)

Mike

 

[emjem]

Whoa! I've got 7 systems running FaD so after reading through this thread I flipped over to the FaD home page to see just who it is that I'm contributing to. Found this quickly and right up front:

"Treweren Consultants is a small specialist software and consultancy business that focuses on innovative drug design technology for the pharmaceutical and biotechnology industries that reduces the cost and timescale for drug research."

And this: "The initial funding has come from Treweren Consultants but income from partnership agreements to exploit the results will allow Find-a-Drug to become self-funding."

I just don't like that words 'business' and 'expoit' mixed together. Am I working for a software consultancy business? I have to admit that I sure as hll don't understand where my contribution is going or how it is being used. And from what I do know as of now I wouldn't send them a check.

I don't want this thread to be diverted to a discussion of FaD so maybe we need a sticky thread for each DC project where we can post our findings. That way we can work as a group to do our own research of the researchers. Who knows, maybe some of the 'researchers' will find it worthwhile to post.

 

[Insidious]

Wow :shocked:

When I read the comments added by Wolfsraider, I was shocked to learn (at least this is what I think I read) that the Rebel Alliance had been excluded from a DC because it was more than one individual??? I would like to hear details of this censorship (PM or in the thread if appropriate). I think it would be outreageous to exclude a group of individuals while welcoming an individual with a group of machines (so many times not even owned by the person who claims the points)

Secondly, I would like to say:

C'mon folks... You can't seriously believe much of any group would put the time, effort and resources into developing a DC project if there were no opportunities for monetary gains.... I don't think the outrage is warranted about the developers hoping to earn something for their efforts. I see nothing wrong with that at all.

-Sid

 

[petrusbroder]

The thoughts coming up here are now getting even more interesting.
But the debate heats up too. When there is heat, a flame may come.
That has not yet happened.
Therefore, just on a very respectful and friendly note, I would like to submit:

1. Please remember, we are debating here and should be friends. We crunch for the same team - the best there is: Team Anadtech.
2. Please re-read the first and the other posts before replying.
3. Please post your replys after making sure that you yourself would feel OK if an opponent had published it.

And to continue the debate ...

 

[Assimilator1]

Originally posted by: flyted
Seems like some just don't care about the integrity of a project, just doesn't interest them. That's OK, I think they are in the vast minority. Others would be interested to find out about the details of what the project is really made of and the ethics of the program, but don't have an interest or time in doing the research. This group I would bet would be in the majority, and probably read these threads more than participate. Almost all newcomers would look at our findings since they are less invested and more willing to keep an open mind rather than someone who's done a project for years only to find out its not what they thought. Still, a lot of current DC'ers would look and see how their project stacks up against the rest and either voice their concern to management or leave for a better run one. And some will stay with the current project no matter what the results, thats fine also. I don't think anyone is out to make people change what they want to run, only present them with the facts and let them decide for themsleves.

Yep I'd be one of the 'interested but dont have time' people ,I would definatly read up any info dug up on any of thr projects.

On a personal basis re S@H I'd agree with this :-

Originally posted by: Spacehead
I'm most familair with S@H so i'll only speak to that.
The "Do we really know what we are crunching?" question has come up before, if not here, at other forums. For my part i guess i've just come to assume that there are enough smart people out there running this project that there's no abuse(if that's the right word) going on.
Look at all the universities, corperations, goverment agencies etc. that have teams running this. Look at the sponsors, corperate & not, on the home page. Look at the funding thru foundations & grants.
I've just always assumed, right or wrong, that someone out there would look into a project before they donate thousands or tens of thousands of dollars.

But I'd happily read more info about it

 

[petrusbroder]

Wolfsraider: YGPM!

 

[petrusbroder]

Thanks, Wolfraider for your post. IMHO this is a valuable addition to the discussion.

Originally posted by: Wolfsraider
....

My reason for doing dc is the fun of it, there are only work units and friendships, stats and milestones.
I want my crunching to remain fun.
Scrutinizing something just because you can doesn't make it right.There has been no sign of transgression, no hint of wrong doing. If it ain't broke don't fix it.

This is a very important point with which I at least agree: The crunching should be fun. I will not continue crunching if the fun goes. There is no intention to take away the fun from anybody. Nor have I seen any signs of transgressions or wrong doing.

So: if there is no problem, then a comparison of the different projects should not matter. Nobody should be afraid of a comparison if all is well. That is healthy competition (which - as far a s I know - is one of the more important aspects of the american way of life). A comparison gives a good basis for choosing a project. And then - at least I - would feel even better, because I know that my choice is based on knowledge.

Originally posted by: Wolfsraider
Now for the Side I want to address, Choice.
I used the term witchhunt before.
lets say the reasearch is done, and the results are in.
Lets also say a teammate wants to crunch a project on the bad list, how many times will he hear through pm or otherwise that he is crunching for a "bad" project, how many tempers will flare, arguments start, or attacks be made, simply because he chose a project that he/she felt was fun.Will he change projects from the pressure or will he stand his ground?

Nowhere has there been any proposition to do anything other than to collect data about the projects - and if you check the proposed points, not even data which is hidden. Only data which is openly available.
Nowhere has there been any proposition to create a "bad project list". The comparison or rather the data will be a good foundation for finding the favorite project. I for myself am very much against a "bad-project-list" because that takes away the freedom of choice.
Nowhere is there any proposition that some kind of board or group or individual should make a judgment based on the compiled data. That is the responsability of each and everyome cruncher.

However: each reader of the collected information will have the chance of forming his or her own opinion about the projects.

Originally posted by: Wolfsraider
How many other forums will start to question the validity of the projects, how many projects will have to spend money trying to prove their worth, the same money we save them by crunching for them.

I do not know if forums will querstion the projects, I think that there are some doing it already, but again: if there is no transgression or no wrong doing, why should the projects bother to prove their worth? It is the science which will be important, and how the clients work, how the stats come along, how the feed back works.

If the science is bad, than we should not crunch for them. Good science proves itself good in time, bad science is sooner or later exposed as such - please see all the stories about fraud in science: most of it has been uncovered within a few years.

Originally posted by: Wolfsraider
Pretty soon we may see the projects start to buy their own supercomputers and not need our help.

I don't know if you know this:
The Rebel Alliance has started a Rebel Alliance Team for Dimes, as we were not able to crunch as a mini team on anandtech under that project, We are in it for the fun, and will do our best to make it fun, but you will always have those that want to rule the way things are done, that want to decide for all the acceptable way to crunch, the right project to crunch, and who is allowed to crunch. I am against this, TeAm Anandtech has a great reputation for the diverse projects we all can crunch, Tas, Taboo, etc... were created to boost production, but if they are told they cannot crunch together, or if they are told that they can only crunch certain projects, what do you think they will do?

I doubt that this will remain a fun place to crunch if this is put into action.

Nobody should tell anyone what and where he or she should crunch. That is the choice of each and everyone cruncher.

Originally posted by: Wolfsraider
If you look at the dhep project or the problems dpad had or how lifemapper made folks look at things differently, you will see the "picture" I am refering to.

Lastly, have you looked at production quotas lately, seti is down to half as are many projects, its just not as fun as it used to be, too much bickering, too many folks not following the live and let live rule.

Do you think this scrutiny will make this better somehow?
As I stated in pm to you peter If this does go through I will not stay here as it won't be "fun" (this isnt a threat folks, it is my choice, Dc is supposed to be fun.)

Mike

Mike, I would be very sorry to see you go. Please note, we are discussing this openly and there has been no consensus about anything, nothing is decided. And DC must remain fun, or - as I said before - I will stop crunching.

About the production for Seti@home. It may very much be as you say, Mike. But some contribution to the decline in production should be be attributed to the many projects that now exist and adress interesting questions. I also think, that the technical trouble with the transition to BOINC and shutting down Classic contributes to the declining production.

I hope these arguments are seen as what I want them to be: a contribution to this ongoing discussion. If anybody feels offended, please let me know, so that I can clarify my points better without giving offence.

 

[BlackMountainCow]

I don't think that you offended anybody, Peter. It's just a natural thing that people react in a defensive - and as result to that sometimes offensive - way if their common habits are questioned in a certain way. I myself took a hold of building a dedicated cruncher due to this discussion. I just want to be clear about the DCs I'm doing.

No offense here, but Mike has always been a very strong about his principles and that is a good thing. Peter never tried to impose anything on anybody, like a black-list of "unworthy" DCs. After all, we're just talking/discussing about the future if DC in general, not the specific future of TeAm Anandtech. It's up to each and every single cruncher to decide what he or she is doing and what kind of consequences - if any at all - he or she will draw from our discussion.

Nobody wants anybody to leave. Yes, in my opinion, the fun of DCing has decreased a bit over the last few months. But I guess that's just how it is with a marriage for example: you can't always be on cloud #7. Sometimes #5 or #3 will have to do for a while. Only if that kind of feeling persist for quite a long time, then it's time to think it all over.

 

[Wolfsraider]

Originally posted by: BlackMountainCow
I don't think that you offended anybody, Peter.

Nope no offense taken

No offense here, but Mike has always been a very strong about his principles and that is a good thing. Peter never tried to impose anything on anybody, like a black-list of "unworthy" DCs.

I am not saying peter or anyone has done this, I am saying that it would open the door to this

Nobody wants anybody to leave. Yes, in my opinion, the fun of DCing has decreased a bit over the last few months. But I guess that's just how it is with a marriage for example: you can't always be on cloud #7. Sometimes #5 or #3 will have to do for a while. Only if that kind of feeling persist for quite a long time, then it's time to think it all over.

I agree Christian.

Maybe I am not saying it right. (about the second part)

Mike

 

[petrusbroder]

OK, guys, OK, :roll:
Then lets add some other questions to the board:

How can the fun in crunching be increased? or:
What is needed to increase fun in crunching for as many crunchers (specifically) and as many others (generally) as possible ?

BTW: Just got thinking: Am I a cruncher? What is a cruncher? Anybody attached to a project delivering results? or is there a more specific definition?

and: What is a crunchee? a just bitten potato-chip? a just munched jellybeen? a just crushed peanut? :laugh:


edit: spelling (had 6 spelling mistakes in one post - :disgust: )

 

[Insidious]

LOL :roll:

I think when you run DC, you are a Cruncher and when you mis-configure those MoBo drivers and fight BSODs for a day you are a Crunchee

New topic (?)

Lately, there have been some major changes on the project I crunch. I don't like some aspects of it, but I like (very much) their strives to move forward and learn not only the science of what they are doing but the DC nuts and bolts (methods) as well.

Here's the question:

How much is too much for a 'work unit'?

lately 'my' project has developed some methods that creat BIG work units. So big that many crunchers don't have the horsepower to do that part of the project. (there are still lots of other work units that take more traditional amounts of resources and are crunchable by all)

These units are so big that they have offered up big bonus points for working them and have even gone so far as to begin to pick and choose what users can receive them based upon their hardware resources.

So now we have a situation where not only are some folders are given ALOT more points in a given amount of crunching time.... others are feeling left out and discriminated against. Total bummer for the project as crunchers are leaving because of it. (others, like me just whine and become annoying to have around )

So folks, what is too big?
What maximums should a project shoot for when they are breaking up their project into little pieces for us to process?

I'm sure these projects put lots of effort into answering this question... what do you think?

-Sid

PS: No one has offended me and I hope I haven't voiced my opinions in a way to offend anyone either. Especially petrusbroder, GleeM and BMC. I couldn't tell for sure from your responses, but please know you have my respect and I hope not to loose yours)

 

[JonB]

I'll toss my response into the heap.

1. Should DC projects go commercial? If so, how should the contributors be remunerated?

I won't participate in a commercial process, so I could say that "I don't care" but I do care, since it would siphon off the volunteer participants like TA and DPC.

2. What do we - the crunchers - expect from the projects? Papers? Aknowledgments? Good information? Honor? Money? "Real-world"-results?

I expect regular status updates and statistics. Weeks or months with no visible change or progress is sure to make me start looking for another project.

3. What makes a project meaningful to crunch?

For me, it must do something to advance scientific understanding of a problem that isn't directly related to making money. Pure science, not commercial science. That is one of the reasons I've stayed away from drug research and leaned toward encryption and physics.

4. Do we really know what we are crunching?

When I turn in PhaseRotB results in DPAD, am I really computing the best ratio of tritium and deuterium for a better hydrogen bomb? Am I finding a more lethal version on anthrax instead of a vaccine? Wow.

What it boils down to is: Research and Trust, but perhaps mostly trust. I do a little reading and research, but mostly I form an opinion from message boards and comments from the operators, get a "warm fuzzy" feeling and begin to participate. If the trust disappears, so do I. I stopped RC5 and OGR for that reason alone.

5. Also, do we know how the results are used?

See #4. If I don't trust the project, I don't turn in results, so it doesn't matter how they use the results. I suppose that if I became convinced that a project was really a disguised hijacking of computer results for some obscure military weapon or terrorist activity, then I would be morally obligated to stop as many people as I could. Thankfully, nothing like that has happened to me.

6. Also a question about the people behind the projects arises: who are these people? What info about the projects should we require?

Also answered in a few places above. Research and trust. Stay in contact with the project forums and statistics. Listen to the problems other people have.

 

[GLeeM]

Originally posted by: Wolfsraider
Lets also say a teammate wants to crunch a project on the bad list, how many times will he hear through pm or otherwise that he is crunching for a "bad" project, how many tempers will flare, arguments start, or attacks be made, simply because he chose a project that he/she felt was fun.

This already happens to some extent. Mostly from newer people and we patiently teach/advise until they become a TeAmmate - see it a different way, that is. Well anyway that's what you all did to me And I thank you in advance for your continued patience.

 

[petrusbroder]

What is too big a WU?
That questions has several different layers:

1. Should speed be rewarded?
2. Should dedication (i.e. staying a long time) be rewarded?
3. Is it the work (in this case: speed by time) which should be rewarded?

4. Should people be excluded because of their hardware?
5. Should anybody (no matter what hardware) be able to crunch all kind of WUs?

You know, I started writing a tetative answer and got stuck almost immediately. These are hard questions, because you want people feel appreciated for the work they do. And how do you compare the work produced by different people under very different conditions? What may seem easy for one may be very hard for an other.

It is a matter of balance: speed (and that means most of the time money) vs. stick-to-it-ivness? Fun vs the honor of being top ranked?

IMHO the best way is to split the project into two categories: those who can crunch those big WUs for themselves and those who don't want to or who can not in their own. These are different systems, and should not be compared with each other, and there should be no competions: it would be like comparing an oil tanker with a speed boat: they have different strenghts.
On the other hand: may it be so, that the project wants to save work (i.e. the splitting of a large set of data into smaller chunks) and thus appreciates the "big WU crunchers" more because they make the life easier for the scientists.

I still think, that two mathematical operations should have the same reward even if they are performed on different computers and take longer time.

OK, this answer is anything but clear, but lets get the discussion under way!

 

[Wolfsraider]

Bump cause there is good info here

And the babes went this way in their thongs.

oops looks like they are headed for the pool

Me<----runs toward the pool

"CANNONBALL"

:Q

 

[petrusbroder]

LOL @ WR!

 

[Bleep]

Well here is my take on the whole thing from the perspective of one that has quit crunching on the Anandtech team and why I did.
I did about 1/2 million rc5 units. I dont remember how many seti units but a lot. The backbiting on the board and allocades that the BIG crunchers were doing with other peoples machines received were somewhat responsible. The fact that I contributed stuff and money to persons that were building big crunching machines only to have my money and the project just go by-by. A bad choise on my part, but the people responsible for what I considered a out and out theft were never abmonished by any member of the community. There was and is to much importance put on stats rather than the science or whatever is involved.
All scientific research will eventually lead to a product of one sort or another so somewhere down the line someone will make some money from the work but that aside most all the projects have merit and one should not be held supreme above all others.
Have I been using my machines for DC? yep just not on any team and the one I have chosen I have researched very well before I dedicated my time and money to the project.
This may not be the kind of post you guys are looking for and I have not been here for about a year but just dropped in to see what was happening and see that nothing has changed Boinc is still screwed up.

Bleep

 

[petrusbroder]

Bleep, I do not quite understand your post.
I can understand that you left the team if you felt that ...

The backbiting on the board and allocades that the BIG crunchers were doing with other peoples machines received were somewhat responsible.

You also write :

The fact that I contributed stuff and money to persons that were building big crunching machines only to have my money and the project just go by-by. A bad choise on my part, but the people responsible for what I considered a out and out theft were never abmonished by any member of the community.

That I don't understand. Would you please explain more exactly what you mean? What do you consider an "out and out theft"? Since I have not been on this board such a long time I have no history about this. If you do not want to post it you could PM me. I have no intention to re-start an feud or fight, but I am quite curious ...

Start edit: Some readers have informed me that you may be alluding to a certain Beowolf cluster (which I BTW do not know anything about. If that is the case, I can understand your frustration. If you want to discuss that, it may be a good topic for a thread by itself. End edit

Then there is the importance - or lack of importance - of the stats. For quite a few people the stats are a means of competition and of keeping tabs. If that helps the projects to get WUs crunched is that so wrong? I mean, not all people can keep their interest awake because of the sense of accomplishment for the project per se. Since the processing of WUs is important for the scientists in that project why not use a competition to maintain crunching power?


PS. Those who contribute, please remember the the first post in this thread. DS

 

[Wiz]

I remember giving of my time & effort to help a guy once only to have him take what I did for him and go somewhere else. I'm sure you remember that, and if not shame on you.

(did I say once, heh that would be cool if it were only once - however the times I've helped people who stayed & made a big difference to TA far outweigh the nuissances of the kind I mention above - far, far, far outweigh them!!!)

 

[petrusbroder]

I hope I understand you correctly, Wiz, that what you are saying is that trust and work can be well placed and may feel very good (second paragraph) while trust and effort misplaced leaves some (very) bad taste for a long time (first paragraph). The behavior shown by the guy who left is a bad form of cheating, perhaps not what one would call directly "illegal", but still a bad very bad attitude.

Although I do not remember that specific episode you allude to - probably because it was before my time (I have been posting in this forum for less than a year and most of my reading and posting has been done since march 23, 2005) - I certainly can sympathize with the feeling of being cheated from a lot of work ...

Behavior which represents cheating is bad ethics.

OTOH: Would you please explain why I - or somebody else - should feel ashamed of something I/we did not take part in or did not know about?

Edit #2: I have recieved info by PM - and understand things better now. The PM will however remain exactly what it is: private.

Edit #1:Spelling/grammer

 

[Wiz]

I guess I was intentionally being a bit cryptic in order to not come right out and call people names. If you don't think you did anything wrong then you are most likely not who I was addressing. Everyone still on this TeAm (as far as I can tell) has my utmost respect and gratitude for being TeAm mates.

If you are one of those I was referring to and you still don't feel you did anything wrong then I pity your lack of concience.

No one else need to feel any shame - TeAm mates are all regarded as highly valued friends.

I was speaking of the few bad apples who have come here from another team and acted like friends seemingly just to get help or information from us and then took their leave of us, using what they gained here against us.

There are others who honestly came, asked for help and told us up front they would not be joining our team, they were honest and there are those of us who have helped them either because we were just generally helpful people, or hoping they might consider joining us at some future time, or merely to help the project they were doing whether they joined us or not.

Honesty is the key.

People who return once a year just to snipe at us are seriously to be pitied, but not to be taken seriously. Those I speak of are team dividers, not team builders. I would not put it past other teams to devise a plan to put some members on an opposing team with the intention of driving down morale & dividing that team. It's happened here before, it may happen again.

 

[petrusbroder]

Thanks Wiz, for the clarification.

Originally posted by: Wiz
...
Honesty is the key.
....

In this we agree. Honesty - whether from team mates or from projects and their coordinatiors - is the one and only basis for ethics and for correct behavior - and in the long run, for the success of a project.

/me: sitting down in a sofa, taking a :beer:: cheers, team mates of TeAm Anandtech!/