Slashdot Mirror
SETI@Home Version 3.0 Client Preview
zAmb0ni writes "We have posted a preview of the upcoming version 3.0 client for SETI@Home. The preview is based on the beta version 2.70 in limited release. You can check it out here: Team
Ars Technica Lamb Chop." I wonder if it will run on this pci thing we mentioned yesterday.
Ahem.
Who would buy a CPU *solely* for SETI performance?! Maybe my priorities are bass-ackwards, but I really would love to meet the person who spends hundreds of dollars *solely* to process little chunks of data that have about a one in infinity chance of containing evidence of intelligent life from another planet.
Is it just me, or is that line really sad?
- A.P.
--
"One World, one Web, one Program" - Microsoft promotional ad
"Remember when the U.S. had a drug problem, and then we declared a War On Drugs, and now you can't buy drugs anymore?"
This is not a trivial question. Firstly, SETI@Home uses the Aricebo Radio Telescope. This is a very nice dish for radio astronomy, but useless for SETI work. It's far too small. The smallest useful dish or array will be the hectare array, being built by the SETI Institute. Aricebo will only be able to detect signals from nearby stars that are: (a) at LEAST as strong as our most powerful RADAR, (b) SUSATAINED for a substantial period of time, (c) containing information on a carrier wave, (d) orbiting a planet or star, with no compensation for motion
In short, leakage (the most likely sort of signal to be found) will be invisible, actual RADAR type devices will be screened out (too short a duration and no information content), and any civilisation advanced enough to WANT to locate other civilisations by sending deliberate signals are likely to be filtered, by being screened out as local interference through a lack of doplar shift.
Methinks that SETI@Home is ingenious, but is using the wrong telescope. And it'll be finished before the RIGHT telescope has been built & put on-line.
As for "Open Sourcing" SETI@Home, it was, to start off with. The original UNIX client was GPLed. Hardly anyone bothered to do anything with it, and so they closed the source & shoved it over to a commercial house. Don't blame them - look to yourself first.
Having said that, SETI@Home's attitude has been somewhat attrocious. They've been going on about security, when that was never the cause of them going non-Open Source. Progress was. And part of that is their fault. They refused to set up a CVS repository, did VERY slow (and low-quality) releases, and basically impeded themselves at every turn. They should have done a damn sight better than that. Yes, there were only a few people there, which is EXACTLY WHY they needed to use CVS, rather than relying on manually testing every e-mailed patch, and rolling a fresh tarball by hand every few weeks or months.
Honestly, if SETI@Home has shown anything, it's shown that we should be less worried about intelligence "out there" and rather more worried by the lack of it down here.
--
You are a fucking moron.
Why do you need guts? Slashdot is anonymous. And I like being flamed every once in a while. So, here goes:
The reply to my post was pure garbage. My SETI username is the same name that I put in for software, and at the time there was an outside piece of advertising spyware installed on my system. The top part goes, though. I have been able to crash my client with specially-constructed (i.e. random) TCP/IP packets. Don't know if that's bad use of sprintfs or just failure to account for bad incoming packets.
I've had a question about this for a while, and none of my friends can seem to answer it. So, I'm going to let the Slashdot give it a shot.
Is there any way to use additional processors or co-processors on a PC (x86) to run SETI @ Home clients?
For example, I have a Voodoo 2 (12 Megs of RAM) in my Linux box. Would there be some way to write a SETI client that uses the Voodoo's processors to run additional SETI clients/threads? This situation is pefect because unless I'm using a 3D program (Quake 3, Unreal Tournament, etc) the Voodoo is doing nothing.
The only reason I ask this question is because computers do math, it's a fact, just math all of the time. Why not have the SETI client use some of the great FPU (Floating Point Unit) on the graphics processors of a Voodoo card? Have the Voodoo do it's math on the Voodoo's processors rather than the PC's CPU.
Also, I don't see why this won't work for other things besides Voodoo cards. Any card that's strictly 3D, some NICs have a small co-processor for checksumming, or even a way to set the prioroity on the SETI client using a 2D/3D card (so 2D performance doesn't suffer when the user is using normal windowed applications).
"The 85 I fear they don't got a clue."
Perhaps your friends would like to see our successful (as you admitted yourself) capitalist nation turned into yet another socialist failure-experiment, but I for one wouldn't. If you have any illusions about the grandeur, oops, I meant the squalor that most Russians enjoyed under their "enlightened" socialist regime, read We The Living by Ayn Rand. And, by the way, ad hominem criticism, as you implied in your statement (i.e., people who aren't left-leaning politically are stupid/uneducated), is not a valid form of arguing. Try again. -t
This is clearly aimed at Geeks-with-Cash, not at the average compuser. If I actually had any money, I'd buy one. I don't even run the Seti client, I do the distributed.net stuff, but this thing is just so cool that I'd definitely buy one if I could. After all, talk about bragging rights:
"Hey Bob, did you hear about that Seti project where people use their computers to help search for extraterrestrial radio signals?"
"Yup. Pretty cool, eh?"
"Yeah, I decided to go ahead and install it on my workstation. Seems like a great project for a geek to help out."
"Yeah, I installed a special multiprocessor vector processing unit in my computer to work on Seti all the time; it runs on a PCI card, and beats the hell out of a P!!! 500 even though it doesn't use any of my CPU's cycles to do the work, it's all in hardware. Took these military surplus vector processors and..."
Now, that's some nice geek bragging rights, my friend. Talk about exotic hardware. I just wish they'd do something like this for distributed.net, since there have to be a few embedded chips which would handle crypto-cracking pretty well.
That brings me to my #1 desire in an exotic PCI card: hardware-based encryption. I want a card with an embedded processor(s) to handle a very strong combination of crypto specifically designed for encrypting hard drives. Wouldn't it be amazing to have a PCI card which registers to your BIOS as the primary hard disk controller, and then prompts for password information before bringing up a boot menu allowing access to your real hard drives and operating system(s)? Imagine, with a dedicated card like that the entire system could be encrypted with almost no overhead, since the card would handle all decryption/encryption and leave the main CPU(s) free. The only slow down on such a system would be the slight delay in routing I/O calls through the card, but I'm sure it's technically feasible to do such a thing. IBM does something similar in the hardware of some of its big-$$$ RISC systems. Now, a card like *that* would be sweet, and if implemented right with good drivers virtually fool-proof.
"The 85 I fear they don't got a clue."
What I've found is that the client's TCP/IP code is easily overloaded. If you can fake a TCP/IP packet in response to its connection, you can initiate a buffer overflow in the client. Boom - instant security hole.
Not only that, but I suspect that the server has the same sloppy coding. I didn't want to try it, because I don't like crashing public servers, but it would be very possible to take down SETI@Home, or even to get root, if you were l33t.
You've been warned.
You guys are looking for aliens in the wrong place. I watch their sitcoms all of the time on my TV that I rigged with a radio antenna. The call me on my cell phone all of the time (stupid pink kid needs to leave me alone and quit trying to grab my phone though).
We're all different.
Eh...
this is sort of haphazard science.. are they going to re-crunch the old workunits? just say "oh well"? I would think if they were interested in these new tests, they should have built them into the original client. Think of all those work units that will either (a) have to be re-processed or (b) not contribute to the analysis of the new data..
:)
or maybe there is a (c) it will be very easy and not take much effort to get this data from the old units.. i don't know enough about the data to say.. i just crunch units. i'm in the top 10%
wish
---
I'm looking for ya, dad!
I never jumped on the SETI@Home bandwagon for a bunch of reasons. The program seemed to be very poorly written, at least compared to the d-net client. I was already involved in the never-ending RC5-64 project. But, the biggest deterrent, was the simple fact that SETI@Home didn't need my processor time. They had 6 billion Windows lusers running their screensavers and a very finite amount of data to be crunched.
Has this changed recently? It seems like just as many people as ever are running the client. Since they only get x amount of telescope time per week, do they actually have enough new work to hand out to all their users?
I love the idea of searching for extraterrestrial intelligence, but I think they've got it well covered. I'm sticking to d-net and optimal golomb rulers. At least I feel as though my processor time is helping to achieve a goal.
We must respect evil, and we must make evil respect us.
I think that they should include this feature in the next version:
When your computer does find intelligent life, it opens up a chat window.
First of all I'm not a big fan of SETI@Home. I can see it's value and use but the ET angle on the project didn't motivate me as it did other. No biggie, personal preference.
The point I'm interested in is that essentially the 'search' is over. Unless things have changed since I last heard, they are just re-analyzing all of their data from as many different angles as possible. This is a good thing from the standpoint of science and verification but it would seem that a good deal of the excitement and newness of the project is gone. What they need after they exhaust the data that they have collected is to get access to the southern hemisphere. More of the milkway can be seen from the southern hemisphere (and thus a higher liklihood of 'interesting' signals) compared to the nothern hemisphere.
And that PCI card is a piece of junk unless either the SETI@Home cpu power has dropped off (they were crunching stuff in real-time) or you insist on improving your stats. It didn't seem from the Ars article that the new client was slower than the current client.
I've wondered if we have enough bandwidth now to have a near earth orbit object search program developed. Something much more immediately useful than SETI, and frankly, more important. As anyone who's seen those numerous asteroid collision disaster specials knows there are precious few people and little money to look for these objects.
Fsck cluebie moderators. I'll say what I want, offtopic or not. And fsck having to qualify every bloody statement just
Well, yesterday, I was checking my SETI@Home progress, and something odd seemed to be happening. The display started shifting rapidly and the graphical part seemed to be scrambling around instead of updating in chunks like normally. A few seconds later, the imagee swirled and coalesced into a message: "Welcome, Jacob, we've been expecting you." I don't know what happened next, but my power went out. That's when things began to get _really_ weird.
First, the walls started glowing blue, even though the power lines were dead. Then, I saw two shimmering images in front of my eyes. At first, they were just swirls of dust, glimmering blue from the light. But that dust soon formed into two blue creatures with three arms and three eyes. Apparently the standard theory about the green face, large eyes, and pointy chins was wrong. I didn't have a chance to speculate more because at that point I was knocked out by a strange device held by one alien. I do not know when I awoke because the clocks stopped when the power went out, but when I awoke, they were gone and in their place was a message: "There are exactly seven hours until the destruction of the human race. Do not attempt any communication with the outside world and you will be saved."
Well, I wasn't going to listen to them. I now have exactly seven minutes left on my battery. I plea for you to listen and heed my call, even though it may already be too late. Even more than I fear my detection, I fear for the human race. They will come for me any minute now; it is up to you to stop them. I saw their critical weakness right before being knocked unconscious. It is--
--
Friends don't let friends misuse the subjunctive.
How about an Arstechica Slashbox? That way all of their stories could still be posted and we could get on with discussing something original. This story was posted on Arstechnica yesterday.
Hmm, still doesn't look like the seti@home client can buffer more than one unit at a time...
/. then?
what a great bit of code. seriously, run dnet instead, if only for the client.
why don't new dnet clients get announce to
Yes, maybe the /. crew should read at least the highest moderated posts before cross-linking.
Anyways,
OT:
to continue the arguments as to why this is a hoax, I can add the following:
As the picture shows, there's no seperate PCI interface chip, so the PCI interface would have to be implemented inside the CPU.
Although possible, I doubt that the army would be surplussing CPU's with integrated PCI bus interfaces. These would be pretty new devices, especially in army terms (like, they are still using leading edge 386's a lot)
Any PCI board designer knows that you can have only two PCI devices per board maximum, if they are directly glued to the PCI bus. So the board with six devices would require a PCI bridge of some sort. Which does not appear in the picture.
The FAQ states that only one PCI board is supported per system. It is pretty much impossible to design a board that can not work when an identical board sits in a different PCI slot. That's just the way the bus is designed. It's like saying that you can not have two systems with the board in the same house.
The only reason why you couldn't have two boards would be driver related, which is not the reason they give, and could be easily solved.
Breace
Supposing that you sent erroneous data to them, they would want to flag you as being unreliable. I would imagine that this is the purpose that that serves. I doubt that they are getting demographic information from that.
We're all different.
Eh...