Slashdot Mirror
Adam Beberg Leaves Distributed.net to develop Cosm
BigJim.fr writes "Adam Beberg announced
on the Distributed.net list that divergence of views on the
further goals of Distributed Computing Technologies Inc (DCTI) have
led him to go on developing independently Cosm, the distributed
computing engine that was known as V3 to members of
Distributed.net."
Is that as far as actual significance of the calculations made you might as well just run a couple of random number generators on the machines and wait till one of them hit some random number... In a way thats what your doing anyway.
Was that really the idea behind distributed? Waste countless resources just for the slight chance some team might win some money in a decryption contest and be a popularity contest for teams battling for the fame of who can dedicate the most computers doing useless tasks?
Now that SETI project or that project to find sequences of numbers who have special addition properties (forgot the name) at least have some merit.
THe purpose of decryption contests is to prove that even a brute force attack using spare CPU cycles could defeat the inadequate encryption systems which the US government believes are sufficient to protect our privacy.
Does anyone know what tasks this new project will accept. I think that distributed computing is cool and has a client running in the background right now, but I'm just mildly interested in these encryption contests and would much rather have my computer perform some more usefull tasks, like the other possible usages mentioned on distributeds homepage.
/Tord
Dnet is a great project, and most of the guys (nuggett, dbaker, silby) seem cool on irc (except all the times dbaker kicked me for making fun of "Topanga"), but...
People having been begging for the client source for 2 years. Don't worry it's coming with v3.... Sure it is... I understand that they don't want some goober screwing everyone else up, but it seems to me that advancing distributed computing is far more important than whether or not rc5-64 is finished. But it's their code and they should do what they damn well please with it. I just wish they hadn't said it was gonna be free if it's not gonna be in a reasonable amount of time...
Since RC5-64 will take years, then the whole purpose you state (showing that current encryption schemes are inadequate), actually works against us. Lawmakers are going to point to DCTI and talk about how it "took the whole power of the Internet 4 years to crack a simple RC5-64 message". In their eyes, this would show that RC5-64 was more than "good enough".
It has long been time for Distributed.net to move on to other projects and to open up the source so the coding work could also be "distributed".
distributed-net will lose lots of people to Cosm. People will prefer open-source free software instead of the closed-source almost-free semi-insecure distributed-net client (which is so untrusted it ran under nobody/nogroup in Debian).
And if you really want to proove it do it with a specialized machine wich at least does not act in the favour of the US governments case...
Saying look this box wich was produced for X million dollar can decrypted DES in a couple of days is convincing proof. Saying the same if the result was achieved by thousands computers networked worth countless million dollars does not strike me as an especially convincing argument.
What hardware and resources are needed can be shown perfectly well in theory, if you have to produce an example at least make it striking... this one would hardly impress me if I didnt know anything about it...
ohhhhhh all those thousands of computers could only try 7% of all those keys in all that time... my gosh then that encryption must really be good.
"Waste countless resources"? The clients I've seen only run when the system is idle. Most systems spend most of their time running through a "do I have anything to do? Nope. Do I have anything to do? Nope (etc)" loop. Contests like these put resources that *would* be wasted to good use (or at least somewhat non-useless use).
Theres also natural resources being used... ie power.
Note that computers are always getting more powerful, so that as time goes by the amount of resources dedicated to cracking will increase exponentially. I would be surprised if RC5-64 took more than 2 years to crack. You can't just look at the current rate of progress and extrapolate it linearly; it's a nonlinear function of time.
Before we broke them, DES and RC5-56 looked almost impossible too.
We have now adequately proved that 56 bits is not enough. Now that we have, it's irresponsible to spend our CPU cycles doing it again - anyone who is listening, heard the first time.
We should be spending our cycles on something of more value to humanity. I like Mersenne primes; others like SETI; either, or any of several other things, would be better than further decryption contests. There was value in doing it once to prove we really and truly could do it. Publishing a scientific paper saying we can do it isn't the same. (It's the same theory as India and Pakistan doing nuclear tests: if you actually set off a bomb, people pay a lot more attention than if you send a memo to the UN notifying everyone that you could set off a bomb if you wanted to.)
But now that we've done it once, there is very little value in doing it again. We're not really proving anything anymore.
...out of 2 years of development time.
Check out the latest note from Nasby:
> "With our new forward momentum and renewed enthusiasm..."
Jeez, what a crock.
Good judgement comes from experience, and experience comes from bad judgement.
- W. Wriston, former Citibank CEO
I hope that SETI At Home takes advantage of Cosm; they haven't been too slick about reinventing the wheel, so far.
Cosm should also be rather helpful merely for people who want to do their own distributed computations on a small set of home or school computers, e.g. training neural nets, ray tracing, genetic algorithms, artificial life, etc.
To date, people have had to do home grown distributed solutions over and over, which is pointless, painful, and wasteful. So I hope Cosm becomes everything it promises.
Professional Wild-Eyed Visionary
i've always like dnet and run rc5/des for a long time! now if my computers can be part of a large distributed computing service i find it cool :) :)
for SETI of for another purpose
--
"Science will win because it works." - Stephen Hawking
The idea is to do it because it can be done. It is a game.
As a side benefit those guys worked out many issues with distributed computing which may apply to other areas like SETI.
I wonder, if COSM is what the v3 project was going to be then multiple modules will be available to run different tasks. With these different modules you could get a computer working on any project you choose, but if it's going to be like the SETI client thing it might require lots of data to be sent VERY regularly and make it almost prohibitive to the home user. (Esp. in England with non-free home dialup internet access)
;)
If think you'd have to be careful not to have too many projects requiring lots of data to be moved, that is one of the nice things about the RC5064 project, the data involved is small and basically negligable, unlike a 3MB / night download with SETI.
Just my tuppence worth, yes I'm English
Delphis
Adam Beberg (duncan) wrote:
it has become apparent that the goals of DCTI have changed considerably over the years, and are no longer the same as what they were.
David McNett (nugget) wrote:
It has also become clear to us that Adam's goals for Cosm and distributed.net's vision of its future differ enough to justify this parting of ways. Adam is very motivated about seeing the system he's designed, Cosm, implemented and put into production. True to its name, distributed.net is more focused on seeing what can result from a truly open and distributed continuing development effort. While each of these respective approaches is viable in its own right, attempting to co-mingle them has proven to be counterproductive.
What's the division about? It's clear from these carefully worded pieces that they've decided to split, and they've put an amiable face on it. However, the announcements are so vague I can't tell what the real issues are.
One of the problems I've always had with the d.net project was their closed decision-making style. On the one hand, they've build this wonderful thing for running the DES/rc contests, and made it fun to participate. I think that's really nifty. On the other hand, they've been advertising 'v3' for a over a year now, with its plug-in architecture promising a wide variety of clients to choose from, and an open interface so one may write one's own. But it never arrived, partly, it seemed, because Beberg wouldn't let anyone else work on it. Sometimes I felt like they didn't want to allow any other clients because they'd lose people from the rc effort, which is what they're really interested it. I think this might not be such a problem given rc5's microscopic bandwidth, footprint, and tolerance for latency compared to alot of other distributed computing projects If you think seti@home is bad, try cg rendering, or scientific simulations! Even the Mersenne prime client is more efficient with a large (>~16MB) memory allocation.
Nugget speaks of "a truly open and distributed continuing development effort." Their hot new client is the OGR project, and still closed source. Beberg is at least publishing a programming interface, but hasn't specified a license yet.
I've always been bothered by d.net's interest in using my processor for their particular project, paying only lip service to giving (control) back to the community they created. I've always been bothered by their failure to grok open source development.
I guess what all this boils down to is that I'd like to think that either Beberg or others at d.net have seen some light in this vein, but I can't tell which of them it might be. Comments?
Is there any such protocol (public, private doesn't count) and does it work on a wide selection of problems? I mean can one machine connect to another at port x and send it a list of arguments telling the other computer about itself and what it's doing and the net address to find the same component to execute or data to share or something. Would be really cool if done right. Rather than needing a centralized server you could have data passed between individual machines in interlinked clusters.
At what price learning? At what cost wisdom? The price is a man's peace of mind, and the cost is his life.
I mean, Adam is the name I've most associated with d.net... then again, maybe that's just cause he's been one of the more vocal folks.. ehehe.. not saying 'the quiet folk' don't contribute, let's face it, the entire idea behind d.net is the 'power of many' as opposed to 'the significance of the one'.. why does this sound like a Star-trek episode gone awry? :)
In my opinion the idea of massively distributed computing is the biggest thing to hit supercomputing since transistors. Cosm is just a natural progression from distributed.net. Coordinating thousands or millions of PC's worldwide and all independently owned is a non-trivial problem, and that is why distributed.net is such a point design built for a specific task of cracking keys. Cosm looks to me to simply be the next step, a more generalized platform for web-based supercomuting. I will support it regardless of what problems it is used to solve. It is the concept that is important to me!
the Prank Institute Because a reason why never beats a why n