BOINC isn't helpful to science?
9 Dec 2006 21:24:09 UTC
Topic 192171
(moderation:
I have a computer science student friend who pokes fun at me for taking BOINC so seriously. I just had a conversation with him where he claims BOINC doesn't help science or scientists! I'll quote the coversation, in part, below (with his name removed, of course). What can I say to convince him? A project scientist chiming in would be neat too ;)
ME(4:06:19 PM): as a computer scientist you should think DC/boinc is neat!
HIM(4:06:21 PM): you're just lucky i don't try to explain how boinc actually doesn't help anyone
ME(4:06:31 PM): but it does!
HIM(4:06:34 PM): not really
ME(4:06:37 PM): lies
HIM(4:06:48 PM): like i said before, it helps computer science a lot more than medicine
ME(4:08:52 PM): this is one of the two i run: http://boinc.bakerlab.org/rosetta/
they MUST be getting something useful from the project or they wouldn't do it!
ME(4:12:11 PM): again - WHY would rosetta (for example) make and maintain a project for NO REASON?
HIM(4:10:30 PM): but when you look at all these projects, what are their results?
HIM(4:10:37 PM): for rosetta, i can't find any
HIM(4:22:54 PM): but that research, at the moment, is not really being applied, and as far as i know, aside from the climate change one (the results of which have been disputed) nothing practical has come out of it
BOINC isn't helpful to science?
)
Most of the projects have not yet produced tangible results. Einstein is an engineering project at the moment, and the result is the Large Hadron Collider at CERN. ALife was part of a PhD thesis on artificial life, and as such when the PhD thesis was granted, the project ended.
BOINC WIKI
RE: I have a computer
)
Well, I guess it depends how (im)polite you want to be! :-)
On the one hand:
Ask him why he's sitting on top of the industrial stack, enjoying it's benefits ( including his own health ) and popping the dumb question "what have the Roman's done for us?".
More productively:
Suggest that he do some homework ( don't hand him anything, he sounds like one who needs to do some genuine thinking & searching here ) on (A) the history of BOINC and (B) the history of science generally.
Also perhaps:
Explain that, by definition really, BOINC is applied to science research that is actually quite hard. The problems require the combined resources that BOINC enables, to even to begin making some headway into! Those problems are not 'linear model thinking' and possess an intrinsic complexity. A read of Ian Stewart & Jack Cohen's book "The Collapse Of Chaos" would educate him handsomely.
If he's keen:
Come in to the lion's den ( a board of one of the BOINC projects ) and ask around! Warn him that he'd have to do better than pure gainsay to be taken too seriously though!
Cheers, Mike.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
In the first place, many
)
In the first place, many research projects never produce "practical" results. That's not to say the work is worthless, it's just that we may not know what to do with the results at the present time. The main thing is that we increase the knowledge base and in so doing increase our overall understanding of how, and why, things work. Many of the current projects may only produce results of interest to a hand full of people in a specific field, or, one of them might produce a result that could change our lives in ways we can't even imagine at the present time!
Does BOINC help science projects? Certainly it does! While it is far from perfect, BOINC allows any project the chance to tap into huge amounts of computing power. How many of the current projects would be up and running if each one had to waste the time and money writing programs to harness that power, or, if they had to pay for that processing power? I'd guess that most wouldn't be. Instead they can use BOINC and spend the time and money on the actual project. That's pretty much a no-brainer.
From what I heard, people
)
From what I heard, people engaged in grid computing don't like BOINC. They call it "populistic". Shame.
Tullio
RE: From what I heard,
)
I hadn't heard that .... but, as usual, what you conclude depends on what you assume.
The challenge with research, by definition, is that you don't know what you don't know. You can't arrange comfortable boundaries and boxes, you do it out of curiosity and gain the context/goals later.
BOINC certainly has a sociological aspect/dimension which I find fascinating, and is deserving of research in its own right. If nothing else it is one humungous nerd get-together! Some are not so nerdy too .... it can have a rowdy crowd flavour also. [ Hence my role is safe for now ]. :-)
Where else could you interact genuinely and earnestly with a vast range of humankind - well of the subset that do distributed computing anyway!
Need I say more! BOINC rocks! :-)
Cheers, Mike.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
There is without a doubt
)
There is without a doubt people here doing this that cover the entire scale of types of humans on this planet.
I have ran into many different types of people over the years on the web but by far the BOINC people have a bit of every version of humans that have electricity and computers.
The main goal is that I get lots of credit doing Einstein and LHC and of course a Noble Prize in Physics and to be known as one of the greatest physicists in world history and also a big stack of that is as tall as I am.
(of course I am pretty much kidding)
The usual way of quantifying
)
The usual way of quantifying science these days is in terms of numbers of papers in peer-reviewed journals, preferably high-impact ones like Science, Nature or The Lancet. This may seem perverse when you consider "blue-sky" projects like SETI or Einstein, whose success seems to depend on finding an elusive something, but managers and, crucially, funding bodies need something to count.
Does anybody know anything about how many papers have been generated by BOINC projects?
RE: Does anybody know
)
There is a list on the BOINC main page. I used it as a bibliography for a paper on BOINC I am writing for "Automazione, energia, informazione", the magazine of Associazione Elettrotecnica Italiana. Of course each of these papers has a bibliography too.
Tullio
I think I'd also point your
)
I think I'd also point your friend to a couple of other BOINC projects:
MalariaControl seems to have a very applied focus, and I believe will be out of Beta soon and producing multiple "real science" papers.
UFluids is being completed as part of a master's thesis (or is it a doctoral dissertation??) from which you can expect journal articles to be produced.
Personally I think his basic
)
Personally I think his basic premise is faulty.
Knowlege and understanding is the goal of Science. Engineering produces the practical applications of the Science.
For example, if and when we find a gravity wave here at EAH, what's going to be the immediate practical application of that? However, verifing another prediction from Einstein's theories would certainly be considered another triumph for Science and thus well worth the effort put into doing it.
In any event Science and Engineering are closely related. Sometimes "practical" needs drive the Science (like Malaria Control), other times Research needs drive the Engineering (like here at EAH and LHC). Either way results are produced, and BOINC is one of the tools they used to achieve them.
Alinator