SETI@Home Transitions To BOINC

SeaDour writes "The team at SETI@Home have finally released their highly-anticipated new client software based on the BOINC (Berkeley Open Infrastructure for Network Computing) software platform. This new platform promises transparent version upgrades, more efficient work unit distribution, and the ability to seamlessly integrate other distributed computing projects that are also using the BOINC standard. For now, SETI@Home is allowing both the Classic and BOINC clients to run, but eventually they will shut down the Classic data server and force everyone to upgrade. You can read more about the transition here."

Waste

[mphase]

SETI seems like a bit of a waste of energy compared to Folding at Home. It's not that I don't believe in extraterrestrials or anything, I even think that SETI is a pretty worthwhile project but compared to curing some of the ailments folding works on...well yeah.

SETI running out of Work-Units?

[SB5]

A very long time ago, I heard that SETI@Home was running low on work-units because their client was so popular that they were just burning through them... Did they restructure it? What happened. I remember when I heard that I started downloading work-units that were taken by the dishes more recently then I had been seening too...

Beta Means What?

[stecoop]

The team at SETI@Home have finally released Bonic

On Bonic web page: Status BOINC is under development. We are conducting a beta test of BOINC using the SETI@home and Astropulse applications. The public release will be announced on the SETI@home web site. Several other distributed computing projects are evaluating BOINC.

Bonic has been "released" for use for a long time; I thought when a release annoucment arrives then the product is no longer beta. So which is it - Released means ready for use or does it mean Please beta test now?

Re:Waste

[Anonymous Coward]

Maybe the new BOINC software will allow you to split your computing time between SETI and Folding?

Re:Waste

[CatLord42]

Maybe the new BOINC software will allow you to split your computing time between SETI and Folding?

Whoever modded the parent down should rethink their decision.

Folding at Home [stanford.edu] seems to be another distributed computing project, just like SETI. I haven't RTFA-ed, but the original post says that BOINC will allow multiple distributed programs to run. At worst, this is redundant, but it is definitely on topic for this particular part of the thread!

Will new client screen out 'cheaters'?

[Siergen]

I recall reading that some SETI contributors had found ways to artificially raise their rankings for number of packets processed (forget how they did it). This angered some contributors whose high rankings were based on real results, and who were now being knocked from the top spots by the 'cheaters'.

Does the new client include methods to block the methods used to spoof the current SETI@Home client?

Source Available

[eeg3]

Interestingly enough, the new client has the option to compile it yourself. The old client didn't have this option, or atleast it was very difficult to find, _if_ it was available. Now maybe it can be ported to archs that were previously unsupported.

Re:a great joke

[Kiryat Malachi]

SS1 wouldn't have sidereal motion, it isn't high enough. Also, there's the fact that its only up there for a couple minutes. But basically, it wouldn't "look" like a signal from extraplanetary sources.

The Moon, well, they would say "It's coming from the moon". I suspect there are ways to tell if someone is bouncing it off the moon... like the fact that it would be an on again, off again signal in synch with the rotation of the planet.

To successfully hoax the SETI program would require a *lot* of effort and smarts.

I don't do pushed software upgrades

[Anonymous Coward]

"BOINC transparently and securely downloads new application versions. This lets us upgrade and extend SETI@home without requiring you to download and install new software. "

Well, if I can't turn this feature off, they've lost my cycles. I don't even allow my OS vendor to perform automatic downloads of "new versions" of programs.

For those with the tinfoil hats, the Patriot Act could be used to force Berzerkeley to download random "interesting" ware for the Feds, and keep quiet about it under penalty of law, under the umbrella of looking for terrorist activity. This ain't Java playing in a secure sandbox either.

File format is not XML: why not?

[LightStruk]

Will the format of input and output files change?

Yes. The new format is XML-like (though not legal XML).

Anybody familiar with the rationale behind this decision? The sample file [berkeley.edu] is indeed very close to legal XML. If it is so close, why not go the last mile and make it legal?
Well-formed XML facilitates communication and interoperability, because standard [sourceforge.net] XML [microsoft.com] parsers can grok it, making it easier to write new implementations that understand the same XML format.

Re:Ah, Seti@Home

[Johnny Mnemonic]

It found that you can ask home users with more computing power than they personally use to donate their compute cycles, if they find the project interesting enough and your work is Very Embarrassingly Parallel.
Furthermore, as broadband becomes more popular, the work will not need to be quite so parallel. And as more devices have actual CPUs and go online, you could ask more of even more appliances--for example, one could reasonably run BOINC on their Tivo or Xbox.

That, as it's been said, is an important discovery in and of itself. The world is more lacking of VEP compute problems than CPU time, apparently, but maybe that can be changed; and maybe that can be changed on a problem that is important. Part of designing a VEP task is thinking about the issue differently and configuring your compute interactions differently; now that SETI has demonstrated the possibility, and has been expanded by BOINC, perhaps it will attract more interest and spur adoption of VEP worksets.

Re:Interesting...

[Anonymous Coward]

Even worse -- and the reason I won't be transitioning to BOINC -- is that it apparently downloads and installs new versions of the applications without user knowledge or approval. It's my system and I insist on maintaining my delusion of being the one in control of what gets downloaded and installed when.

Not quite ready for the unwashed masses?

[stefanb]

I had heard about the eventual switch-over some months ago, but never found the time to play around with the beta, so I took the opportunity now to install the client and check it out.

On Mac OS X, all went well, and my PowerBook is munching on it's first unit, fans spinning. However, when I tried to start the client on a Sun box at work, it failed with "ld.so.1: ./boinc_3.18_sparc-sun-solaris2.7: fatal: libstdc++.so.3: open failed: No such file or directory." A quick Google confirmed my suspicions: the client is linked against the GCC stdlib, which is not a standard part of Solaris. Now, that's easy enough to fix if you've worked with Solaris before: just go to sunfreeware.com, and find a suitable binary package to put on.

However, someone not knowing about Solaris, GCC, and sunfreeware.com might be a bit stumped. And the boinc/setiboinc boards reveal that quite a number of beta testers are confused about this, not only on Solaris but also on Linux. It's not completely obvious which GCC/libgcc packages contains libstc++.so.3 (as opposed to .2.x or .4.x).

The real kicker is that I couldn't find any hint of this problem or a solution on the site. I probably looked in all the wrong places in the last half hour... And I couldn't find a feedback form or email address either. This definitly needs to be improved if they want people to move over to boinc.

Can you disable automatic updates?

[Hibernator]

In the transition FAQ [berkeley.edu] it says

BOINC transparently and securely downloads new application versions. This lets us upgrade and extend SETI@home without requiring you to download and install new software.

which makes me wonder if users can disable that. I don't want anybody installing software on my computer without my approval, thank you.

The FAQ didn't answer that question--does anyone know?

Re:Waste

[NanoGator]

" I even think that SETI is a pretty worthwhile project but compared to curing some of the ailments folding works on...well yeah."

I help SETI because it's drastically underfunded compared to the types of things folding would cure.

Re:Waste

[kalidasa]

Thing is, it's pretty easy to get money from charities, governments, private philanthropists, and other institutions to fund medical work. How easy is it for Seti to get that money?

the be all end all of posts

[Darth_brooks]

Lets see what we can cover:

BOINC isn't nearly as usful to society as Folding@home, AIDS research@home, help feed starving disabled puppies in war torn african nations@home, etc.

BOINC != Seti@Home. BOINC is a step up the ladder from Seti, it provides the infrastructure for multiple projects. *you* choose the project to attach yourself to and contribute time to. In an ultra-perfect hippie world, Folding@home would use the BOINC infrastructure. Instead you get to help out who you want.

I ain't trustin no Berkeley hippies to silently install no black helicopter, tinfoil hat disablin' technology on my system.

Then don't use it. If you ran seti, you really had no way of knowing what was coming down the pipe now did you? You opened up a nice big gaping connection into your system while trusting that the work units weren't poison pills and that Berkeley's infrastructure hadn't been comprimised. Run the client on a non-critical machine, put it outside your firewall if it makes you happy.

Scientific progress goes BOINC!

You're very clever. You're the only person that ever thought of that.

Aliens will enslave the earth when we make contact!!!!!

You really shouldn't have rented Battlefield Earth.

Re:So long, SETI@Home..

[rritterson]

While you might be right, I hardly believe that a large number of users are running their computers 24/7 just to appear on a semi-obscure top 100-list.

Also, I followed your link and I like the new system much better- it awards credits based on CPU time/clock rate instead of just number of work units completed. Thus one credit will be more uniform across all platforms. What's wrong with that?

Re:Waste

[The Lynxpro]

"SETI seems like a bit of a waste of energy compared to Folding at Home. It's not that I don't believe in extraterrestrials or anything, I even think that SETI is a pretty worthwhile project but compared to curing some of the ailments folding works on...well yeah."

A waste of energy? Its an exploration of a scientific question for folks interested in hard science. How is that a waste of energy? That sounds like an argument people use when they claim that money spent on NASA should be spent on fixing the problems of "the real world" such as poverty.

Perhaps if mankind finds 100% proof (through SETI) that intelligent life exists out in space, us humans might actually try to live in peace with one another. Is that exploration a waste of time? Certainly with peace we could free up resources towards tackling diseases that plague our population. Then again, the counter argument is that most medical breakthroughs occur during conflict. Maybe we should be looking for hostile space aliens then...

By the way, you can use BOINC to choose what resources you want to spend on various shared distributed processing programs, such as between SETI and Folding. At least the Beta version did...

Re:Ah, Seti@Home

[deglr6328]

I've recently decided to uninstall SETI and leave the search after staying with it from the begining in '99 because one small graph in an article of Scientific American has kept sticking in my mind. Around 2000-2001 SciAm published an article that included this [mit.edu] graph of SETI's search results (negative, natch) for the galaxy.

Even back then you can see that a large portion of the interesting parameter space has been excluded; it's been 3 years and not a peep. SETI's negative result is very, very important but it feels like it's time to move on.

BOINC SETI@home - Second System Syndrome?

[wintermute1974]

I find the BOINC client to be far from user friendly. I could figure it out and get it working but I have my doubts that the average Joe could.

I have to agree. It was with some sadness that I uninstalled the old SETI@home client before installing BOINC. The old client was compact, quick, and friendly. In contrast, the BOINC interface seems cheerless and industrial.

If tonight had been my first experience with the SETI@home project, I would have uninstalled it completely and told all my friends to avoid it. I refuse to keep any program that crashes my system when I try to use its basic functions.

That said, I really like SETI@home, and I'm willing to stick it out with the new BOINC client. I only hope the most egregious bugs are removed. Ever since I was a kid, I have wanted to contribute to the search for extraterrestrial life. Since I didn't grow up to be a professional astronomer, I would continue to gladly contribute my spare clock cycles even if the SETI client was much worse than it is now.

I think that SETI@home does important work, but I worry that BOINC might become a classic second system, with plenty of new functionality and configurability, yet big, cumbersome, and bloated in comparison to the original version.

people who define themselves by not believing

[exp(pi*sqrt(163))]

Christ! It's just a sig, not a definition of anything. I've used countless different sigs over the years. This particular sig is merely a response to the countless bumper stickers I've read over the years pronouncing that God exists, we must obey Jesus, Darwin is wrong and countless other articles of religious faith.