Slashdot Mirror
Could Distributed.Net Help the Mars Polar Lander?
Anonymous Coward writes "This official JPL press release
describes the current attempt to listen for faint signals from the Mars Lander. They get three windows a day, and it takes 18 hours to process data because the signal is so weak (if it's really there). Too bad they don't have a deal with distributed.net." Interesting thought. Is anyone at distributed.net or JPL interested in pursuing it?
we'll lend NASA some time on Distributed.net, in return, they switch to metric so they don't loose another space craft again... 8-)
-=- SiKnight
Hi. My name is Jeff Robinson, and I run BigCheese.com (currently in development), soon to be the Internet's only cheese-oriented vertical portal.
:)
I find it very interesting that the Slashdot community only seems to talk about applying distributed computing to abstract problems, like cracking encryption schemes, or ones that will never produce any results, like finding aliens. This is especially interesting in light of the many possible practical applications that distributed computing could have. Finding the Mars Global Surveyor sounds like a good application. Or, for example, I saw a story on the Slashdot front page about the government wanting Kevin Mitnick to turn over his encryption key. Why couldn't we just use distributed.net to do this? If there are any other suspected terrorists out there, we could use it to try and crack their encryption too. Or, for example, even in the cheese industry, there are plenty of problems (for example, in dairy equipment design, or the traveling salesman problem for cheese delivery) that could use the application of massive amounts of computational power.
Maybe somebody could implement an open source (in order to be community friendly) distributed computing mechanism so people could write their own modules and optimizations. And so what if the hacked clients aren't entirely accurate? A few mistakes can't be all that harmful.
I mean, it's not like a slice of Swiss has to have exactly 12 holes or anything like that.
Jeff Robinson
President & CEO, BigCheese.com
Primarily becouse the lander only has so much time.. And if this is going to be their primary communications outlet, they're going to have to be able to decypher the signals at *least* once per comm window.
And NASA already has the ability to decipher the garbage, on a normal PC. The code *should* be able to move into a d.net type of environment fairly easily..
Ehter way, *it can't hurt*..
-- I'm the root of all that's evil, but you can call me cookie..
And now the good news: This project exists and is in the works. It is real and called goes by the name of "Cosm". Check it out.
They don't have a client yet, but there is a CVS server with code that is being developed as you read this.
The little green men find out about this and immediately begin encoding a trojan horse in the decoy signal that NASA's been detecting.
As thousands of clients crash throughout the planet, Linux enthusiasts eagerly point out that their machines not only can process the signal but even identify the byte code signature of the
trojan, without any ill effect.
This seems finally like a worthwhile use for distributed.net's power.. sure, cracking encryption is cool, but necessary? hmm...
Hopefully, there'll be some serious cooperation between nasa and groups like d.net, I know I look forward to lending my processing power towards anything other than fading the start menus on my win2k box :)
TheSacrificialFly.This suggestion sounds more like the Seti@Home project - mass distributed computing power used to scan for signals.
The problem with the Seti project is that nothing is ever found. Nobody knows what they're looking for, and nobody knows what would be done if something were found. With a search for the Mars Lander however, everyone knows what the object of the game is. There's millions of quid/bucks worth of serious hardware out of reach for even those at NASA. Imagine the elation in the techno community if we actually found it for them again!
If distributed.net don't take it up, how about a Mars@Home project? I know I'd be one of the first to download.
insignificant sig
I'm an avid fan of distributed.net, have all my boxes cracking RC5, but it seems to me that SETI@home would have the implementation already set up for the distributed computing of radio waves.
Also, in all likelyhood the information needed would already be gathered by the time a server and client core was coded...
... If it only takes NASA 18 hours to proccess the signals and they have a limited number of signals to process, why the hell would they go through the trouble of setting up and coordinating a massive distributed effort? And people wonder where the money goes heh...
Listening for something that might actually be there? I'd donate my computer's cycles to that!
But really, I can't see any way that the existing d.net clients could be used to process signals, which means they'd have to write a new client and then redistribute it. That would probably take a lot of time. Plus, since the individual clients would be running on computers all over the world, isn't there the potential for error due to screwed up computers or deliberate tampering with the client? The d.net model is fine for something unimportant like RC5, but a sensitive multimillion dollar project might want a more well thought out system.
I wonder what the feasibility is of having a system similar to d.net, run by the govt(s), whereby a (small) monetary amount is given to end-users participating in the system. If it were calculated on a percentage system (you're paid for what percentage of the total processing power you supply), it would draw more users to the project and users would be more inclined to install the client on more computers. Projects would simply be anything of national or global importance: surely there are a number of projects like this undertaken internally, with the govts' own computers? Of course, the payment would be quite small; the main reason for doing it would simply be to help out. But it would be a nice bonus, and *would* attract more people. With this, substantial security issues would arise; especially if the projects are of a sensitive nature but with careful planning and a truly distributed system (so *every* client would have to merge in order to produce a readable image of the project -- if this is necessary of course -- and if the entire "keyspace" is encrypted BEFORE distribution, it should be impossible to break -- but doing sensitive projects would most likely draw ethical considerations, as end-users would want to KNOW what they're cracking; but it's just a side-thought :). What do you think of this? Any more pro's / con's? Another con I can think of is it could jeapordize d.net and assorted projects; people may prefer to give full processing power to the govt system rather than d.net if the monetary incentive is there. That would definitely be a Bad Thing, as d.net and the governments are likely to have differing opinions on projects.
s8n@planetquake.com / ICQ: 5089166 Riding the Rocket - Quake 2 Trickjump Competitions - http://www.planetquake.com/rtr/
What Link? The link to your stats? Are you saying that all your stats are bogus?
Hey Rob, Thanks for that tarball!
"Going to war without France is like going deer hunting without your accordion." - Jed Babbin
No. No. No.
Bloat in the Government isn't feeding the hungry or finding shelter for the homeless. It isn't even putting computers into ghetto Schools or removing graffiti.
Bloat is when founds are sought and allocated for those things but don't get there. Bloat is when you have a $50,000,000 "Urban renewal" project for downtown and this project establishes offices on the other, more affluent side of town. Then manages to run up $17,000,000 in "Administrative Expenses" before any actual work gets done on replacing the condemned buildings downtown.
Bloat is when constructing an office building costs a private business $3 Million and a smaller, simpler, less durable building costs the Government $8 Million.
In short bloat is not about what is done with our money but rather when *nothing* is done.
--= Isn't it surprising how badly I spell ?
Where did the insane guy get a cardboard box? From a refrigerator designed with space-age technology, of course!
Throwing money at the poor has been tried before, look at FDR's "New Deal" and LBJ's "Great Society", so, if that worked, we sholdn't have any more insane guys in cardboard boxes to worry about.
Technological advance is the only way to alleviate the effects of poverty, because you can't eat money. Money by itself is useless, it's the warm clothes, food, and medicine that matter. To produce more of those, at lower cost, we need a more advanced technology.
Then you will say: "let's invest in developing those technologies, instead of space exploration!" But it will not work. If you check any textbook on multi-dimension function optimization you will see why. Following the hill climbing approach blindly will get you trapped on the nearest local peak. If you want to minimize the cost of one particular technology, then you must pursuit other, even unrelated, technologies. If we had never done that, we would have now the absolutely most perfect stone axes imaginable, but nothing made of metal.
"I'm sorry, I don't see why this is a news story. Roblimo, It would take weeks for d.net to implement a new module, not to mention coding it..."
The point here is to get people thinking about the idea. This is not "News for Nerds" as much as it's "Stuff that matters."
Imagine a versatile, rather than single-purpose, "idle cycle" processing network that could be adapted rapidly to take on new tasks - like searching for low-power signals from out-of-touch spacecraft or comparing large numbers of telescope observations to help find small, moving images like asteroids and comets.
I'm sure there are many other potential uses for "idle cycle" distributed computing. I don't think it hurts to free up our imaginations now and then and brainstorm a bit about them.
If nothing else, a little speculation about the use of distributed computing to help NASA is a welcome relief from all the lawsuit and privacy and domain dispute stories that seem to make up a depressingly large percentage of the news submitted to Slashdot lately.
Remember the First Rule of Slashdot: "No matter what you say, someone won't like it - and will tell you so. Loudly."
- Robin
A) I also agree that slashdot is a great place for things that aren't news, but about discussing things that could exist or would be a good thing to exist.
B) I also agree that there needs to be some sort of idle-cycle eater.
Hence: cosm
Check out where the project is headed. They have a mostly done CPU/OS abstraction layer, and a utility layer is just coming into the works.
It is going to sport a neat OpenGL interface and should completely blow away anything else near it. Just think distributed.net with the flexibility to say that you want to be looking at the stars today, instead of rendering some feature film, or helping out with genetic stuff, or perhaps even a little crypto breaking. Or, make your own project, have it do all the fun stuff.
But the CPU/OS layer can work for anything. When the gui is done it could be used for many more projects than just cosm...
-- dieman - Scott Dier
yeah, and also they wouldn't be pointing the telescopes at Mars, anyway. Radio astronomy is about looking at stars and galaxies, not the planet next door.
--
Just FWIW, CPU cycles may not be what's limiting the data analysis - it may be tweaking the reduction algorithms and re-iteration, which requires a lot of human intervention. Back when I was an astronomer, this was 90% of the battle with certain sets of data - taking out unwanted dark currents, dealing with strange flat images, bias levels that changed with position on the sky. I almost went insane with one set of data from the Curtis Schmidt at CTIO. Ugh.
You mean the 'Struggling with budget cuts' NASA?
I'm sure they're swimming in supercompters, from the 1980's
Later
Erik Z
Democrats or Republicans. They are both taking us to the same place and they are not afraid of us anymore.
Been a bit of banter back and forth as to who has the most available computing power. Yeah, NASA probably has several banks of supercomputers chugging along, but Distributed.net's collective MIPS is at the very least a decent match and quite plausibly exceeds NASA's power.
And there is something that no one has mentioned, yet: NASA's supercomputers very likely have something better to do at the moment. NASA is a big agency, one with obvious needs for computing power. Who says that D net has to find the Polar Lander? Why not let D Net do something more mundane like running chaos-based climate iterations or calculating the trajectories of all known objects in space? (You could assign everyone a bit of space junk: Hey look, I got a part of MIR!)
Just a thought... we could break encryption or give SETI yet more redundant cycles. OR, we could do something more practical for NASA.
----
----
Am I the only one who thinks Microsoft is a misnomer? Perhaps Macrosoft would be a better fit?
That was the point the poster was making.
"The question of whether machines can think is no more interesting than [] whether submarines can swim" - Dijkstra
<rant>
It pisses me off to the extreme when the United States media, the government, and people like you state with some sort of authority that NASA has some sort of bad "track record." Let's look at what's going on here. NASA launched over a dozen missions last year. How many failed? I can think of two. But the prevailing attitude towards the pursuit of science in this country is one of apathy. New discoveries? Blah. New planets found? don't waste my time. Cure for cancer? Good for them. Multi-million dollar space missions fails? Now that is news! Because of this attitude exhibited by the likes of people like you, the media has made NASA look like a bunch of fools. Do you have any idea what goes into sending something to Mars? How could you possibly think these missions would be 100% successful? They can't be. What this leads to is a general malaise concerning NASA when it comes to the American public. So we end up with less funding for them and more funding for useless liberal fedbloat. I pray for the day when the Average Joe will be aware of the technical sophistication and sheer American Ingenuity(tm) that goes into NASA projects, and exactly how beneficial these have been to the United States, nay, the World as a whole. You are a victim of the media; or your own foolishness; or both.
</rant>
NASA is pretty famous for not having any computing power.
Distributed.net has a great record of getting new clients (like OGR) out in a timely manner.
I have a great bridge for sale. It's in New York, and has a great location. Serious inquiries only.
Not being a programmer, I am curious how long it would really take to implement a d.net client. I am not sure how the four different d.net clients now operate, but I assume that some portion of the code is reuseable and another portion of the code would be specific to the client. ...say MLDR. Then whenever the clients connect, and there are new data to crunch, those people who have setup their clients properly would automagically work on a few lander blocks.
Assuming that NASA would be willing to open source the signal processing code, is it possible that the d.net folks could 'plug' it into the clients? Since the d.net folks have already coded client switching into the current clients, they could conceivably createa fifth client
Just an idea....
In the immortal words of Socrates, who said; 'I drank what?'
I've just intercepted a signal from the Mars Polar Lander. Here it is..... "This program has performed an illegal operation and will be shut down." Oh well, I guess we'll never get in touch with it now....
So we end up with less funding for them and more funding for useless liberal fedbloat. I pray for the day when the Average Joe will be aware of the technical sophistication and sheer American Ingenuity(tm) that goes into NASA projects, and exactly how beneficial these have been to the United States, nay, the World as a whole. You are a victim of the media; or your own foolishness; or both
I smiled when I read this. Naturally many grandiose arguments can be made about the relative importance of discovering new planets circling distant suns (that is to say, confirming something we all more of less knew anyway) versus keeping, for example, several million impoverished families in warm clothes, food, and medicine for a year (that would be the 'useless liberal fedbloat' I suppose).
The fact is, that neither of these projects can go begging in a society that has long term hopes for itself. I'd agree with your general sentiments that these projects are important and deserve funding, but that's relative to our lifestyles. Personally I'm guessing my priorities might shift a little if my own physical survival were on the budget table for negotiation. The vague hope that someday humans will set foot upon the soil of a foreign planet seems rather unimportant when the insane guy in the next cardbox box keeps trying to steal your blanket.
Was it Dostoevsky who said "Boots are better than novels"? I always liked that quote.
-konstant
Yes! We are all individuals! I'm not!
-konstant
Yes! We are all individuals! I'm not!
It really worries me that this gets a score of interesting, when it bears no relation to the reality of the story.
The post on slashdot says 3 windows a day, the
article says 3, 30 minute windows per day. That means they require to process the data in roughly
7 and a half hours, even without assuming time to formulate a useful signal.
18 hours a day processing time means they miss 2 potential windows. If an infrastructure was in place that would allow distributed clients to be quickly assembled and spread this could be potentially useful. I doubt that this can be done in this case. This does not preclude it being useful in the future.
There is already effort by ex distributed.net people to put such an infrastructure in place.
Check out cosm for such a project.
Hmmm... good point. We (Americans) do bag on NASA quite a bit and don't give them credit for the amaizing accomplishments they do (anyone know how hard it is to throw a chunk of metal into space and have it land a few million miles away on target? Think about it).
*HOWEVER* NASA does have some problems. In Richard Feynman's book, "What Do You Care What Other People Think?", he points out a serious lack of communication between the engineering staff and the management with regards to the shuttle(s).
The management seems to believe its own press to the exclusion of the opinions of the engineers. That is one major sticking point for those of us who are proud of NASA's achievements, but think it could do a lot better with a different management attitude.
Before you flame me on this, note that I do understand the political reasons that NASA leadership over-sells various technologies (and, evidently, why they're sending missions to Mars instead of using the moon as a testbed...). It just strikes an ugly cord in me when I see well meaning managers hamstringing there worker because they just don't listen.
Comments? corrections?
"One man can change the world with a bullet in the right place."
- Mick Travis, "If..."
LOL! Yes, you're definately right on that account. I wasn't trying to really make an economical or political argument, but, rather, an observation of an attitude prevailent in the US. You do make a good point, though. Obviously, no society is or ever has been (or, IMO, will ever be) without problems. So how do we decide what's "necessary" for the government to address, i.e., poverty, crime, etc., and what's "extraneous, but useful and important" i.e. NASA.
Well, I haven't read that book, but of course NASA has problems. Every group of people has problems because, after all, we're only human. But shunning something because it is flawed most certainly will not fix it. Agreed, many things within NASA need to be addressed. I wouldn't even know where to begin, but I think your post conveys a similar attitude to what I was originally ranting about. Instead of the American Public whining about the fact that NASA has problems, we should demand that something be done about it. It's our taxes, so it is our agency; it annoys me how many people seem to forget that.
What about cutting money for missile defence research? It was an expensive waste of time in the 1980's, will be an even more expensive waste of time in the coming decade.
Any sufficiently advanced technology is indistinguishable from a rigged demo
--Andy Finkel (J. Klass?)
Isn't seti@home about filtering out radiosignals out of the cosmic noise? Well if this thing is sending, its signals should have been picked up already by the radiotelescopes, don't it? :)
Regards,