Slashdot Mirror
ProcessTree Gets Its First (Paying) Client
myosin writes: "ProcessTree (sell your unused cycles for $$$) will soon announce it has its first paying client. Here's the [preliminary] announcement :)
From their site:
'The good news is that as soon as the end of the month, some of the suppliers of ProcessTree will be able to earn money with their computers. The bad news is that this only applies to a limited number of participants that will be hand-picked according to the requirements of the project.'" And joe points out (from that same announcement) that "The job is a quality-of-service monitoring system that allows real-time checking on the performance and availability of Web sites. We will initially start with 25 locations, for each of which we plan to have a number of suppliers to provide a 24/7 coverage from each location. Each such group in a location we call an 'ideal machine.'" However, they prefer to trade shekels only for cycles on your permanently connected Windows-based machines for now, thanks.
Just thought that I would be the paranoid one here. Maybe its all some elaborate scheme to install DDoSware on unsuspecting windows machines?
Mwahaa haa haa.
-"You'll have plenty of time to sleep when you're dead."
Now I finally have a way to pay off that Cray I bought on eBay.
Even so, I know 3 windows users that leaves their computers on 24/7, compared with about 10 linux users, thats with permanent connections.
I can see a big problem at Schools and Companies networks when all of a sudden there is an increase in network traffic.
If that's the case, my girlfriend would probably force me to get it done to my car so I can't pick up chicks!
;-)
I think I'll keep quiet about this one!
The idea of open source is not that everybody reads all sourcecode they run. The idea is everybody (that can) looks at some source they use once in a while (usually when fixing/improving something), and then would notice if something was wrong. While the percentage of users doing this has certainly dropped with linux/BSD's new-found popularity, there are still enough capable people out there who do this.
**Yawn**....So you're trumpeting remote admin services *nix has had for years, or touting its great advancements over Win9x....woooooooooo
---Most Definitely not a Karma Whore---
Haha, yeah . . . theft is cool.
Next time I see a bike on the street that's not locked up, I think I'll steal it. They deserve it.
And my neighbor doesn't keep his lawnmower locked up, either. Sure, it's on his property, but I think it'd be cool and 733T to teach him a lesson by stealing it and pawning it for beer.
What if the task that the client is doing is illegal? For example, it can portscan computers of some Super Secret Agency, or it can work on cracking of materials protected under DMCA... Will the owner of the computer be able to prove that he didn't know what is going on? Even if so, will he ever get his computer back? I don't think so.
From a technical standpoint, distributed computing is damn cool. I don't think anyone here would dispute the technical brilliance of not-for-profit projects such as the SETI network. However, this will not make money.
ProcessTree has a client on-board, but have not yet decided on the compensation rate for the "suppliers". This is where the problem starts. There is not a rate that will work.
If the rate is too high (which it inevitably will be in the beginning, to get "suppliers"), the company will be quickly drained of resources. Whatever ProcessTree is planning to charge the client (can't be much, with the price of processing power these days) will not cover "supplier" compensation, project development time, network bandwidth, administrative overhead, etc.
The only solution to this problem will be to reduce the rate paid to "suppliers". This reduction will continue as ProcessTree attempts to refine their cost structure. Eventually, it will get to the point that "suppliers" don't think maintaining the ProcessTree client is worth the hassle, for a mere buck and a quarter a month. Goodbye, ProcessTree.
To make money, a business must be able to solve an existing problem better than existing means. This project will fail because 1) it will be cheaper and easier for clients to maintain their own Beowulf cluster of monkey brains and 2) the rate paid to "suppliers" will be too low to matter. CPU time is cheap. There is not a significant problem to be solved.
Some computations can be made in such a way that you can check the work that other people have done.
For example, if I am trying to find giant prime numbers, I can distribute out individual numbers to check to all the workers. Each worker can then test to see if it is a prime, and can then convince me that the number is not a prime by giving me a factor of the number. I can check this much faster than I could test the number for primality.
Attacks against this can take the form of the attacker claiming that the number I have issued him is a prime, causing me to expend my own (limited) compute time on checking this claim. Coupling this with, as epaulson said, double checking, or even triple checking, this can be minimised.
So I can still benefit (but not as much as in a fully trusted scenario) from a distributed untrusted computing infrastructure.
Oh, come on! I submitted this story to Slashdot a full six or seven hours ago, and it was summarily rejected.
-aardvarko
webmaster at aardvarko dot com
I just have a quick question as to the actual project. In what way does this differ from Keynote's service. I am a contract engineer and my company works very close to Keynote and we get full coverage throughout the globe of the companies websites as well as our competitors. Seems like the same thing, and Keynote is quite affordable for it's services. Is this a case of reinventing the wheel or something new?
Dacels Jewelers can't be trusted.
Simulations, optimization problems, risk analysis of some sort? I assume data mining is out of the question. What sort of calculations are not suited to this kind of very distributed processing?
Time wounds all heels
Saddam doesn't want anyone to speak up against his regime because that would allow his opponents to rally the population against him. Revolutions happen when groups of rebels get together and spread propaganda against their oppressors. If you punish everyone who dares speak against you, you can stop rebel movements from forming. So it's impossible to have free speech in a totalitarian state. In a democracy, on the other hand, information is spread freely, and this allows people to decide which government suits them best.
As for people being locked up for saying "unpopular stuff" in democracies, what kind of stuff are you referring to? Death threats? Hate speech? Copyright? As always, your rights stop where other people's rights begin, and speech is no exception. People are locked up for saying things in democracies only when saying those things hurts other people.
Of course you can have a totalitarian state and disguise it as a democracy, but I'm speaking of states which really are democracies, such as the one I'm living in.
"Any system can work"? The states that are most prosperous and have the most personal freedoms (US, Canada, western Europe, etc) are all true democracies. So I guess any system can work, but democracy has shown itself to work better than all the others.
As for communism, we've had communist regimes for almost 100 years and all of them have sucked. We have a lot of "good examples" of communism. How much more suffering is necessary before everyone accepts that communism doesn't work?
By your logic we don't have any examples of decent countries run by blacks (to pull a random example) so we should just assume that black people can't run a proper government? Or would it just show that the countries with black leaders are in poorer parts of the world with a lot of religious and economic problems that would prevent any country from doing well under any leader?
I think it's the same with communism. There's no reason communism requires secret police, or anything else that Russia had. Many non-communist counties are brutal totalitarian regimes so it's not like communists have cornered the market (is that a pun in this context?)
Any state can have politicians who want to stay in power and who fear revolution. Look at Clinton/Gore, and the Bush clan... We just watch them a bit better, but if we didn't I think we'd find our freedoms slipping away just as fast as in any other country where someone takes absolute power.
that would probably have something to do with me being an Australian.
How we know is more important than what we know.
There are several players in this space now that have some serious VC backing and it will be interesting to see how it shakes out. I'm no expert, but it seems to me that an Open Source / Free Software play here could really be interesting about now.
--Seen
"I used to be a dilettante. Then I thought I'd try something else for a while."
Probably because there's 10 times more Wintel based machines for them to pick from than *nix boxes... I doubt the corps that have Onyx's and E10000s in the back room will let process tree's software anywhere near their machines.
When the high end is unusable, the next step is to use the commodity platform, which would be Wintel. Yes, linux and *bsd may be a bit more stable, but to use those would mean to limit themselves to a small subsection of the PC market... Plus, i believe they're only distributing closed source binaries, which further impeeds them from the diehard GNU market.
We're not talking mission critical here... Any machine can crash out for a few hours here and there, because there are so many other machines ready to take it's place should that happen... Kind of like a RAID 5 composed of computers...
Who's the client? I could tell you who it is, but then I'd have to kill you :).
All the l33t h4xx0rz use Windows though.
I signed up months ago, and did get an email from them yesterday asking me to update my info. I run W2K, which is about as stable as Windoze gets (I like it, actually...up for two months, with the only reboots being to install new hardware...). I've also got a 24/7 LAN connection (I love my school :). My machine is pretty powerful, lots of RAM, etc. If this doesn't fit the bill, I'm not really sure what would.
www.code-fix.com
First, it looks through your cookie file.
Then, it logs into any e-commerce sites you have cookies for, and if they're the good kind that saves your cc info it buys all sorts of things!
Hm. That would be interesting. You could easily write a virus that runs around buying stuff... The credit-destroying, economy-enhancing virus.
I'll call it... AMAZONLOVESYOU.
wow.. so all those windoze boxen that I have owned can earn me dollars? Cool.
How we know is more important than what we know.
yes.. if you were to rent the computers and run them 24/7 you could claim their usage off your tax at the end of the year, so essentially cycles cost nothing.
How we know is more important than what we know.
heh.. yes.. their first customer is the Cult of the Dead Cow.
How we know is more important than what we know.
Holy mary mother of god, that tripe was moderated up?
... YMMV
For one, it's not the vast majority of "linux people" who have "skill". The people with the real knowledge are a tiny minority. But all it takes is one person to write a script to get around it and hundreds will download it. Now, do they have "skill" or not? I think the answer is painfully clear.
For two, people who try to cheat systems like that are, if not acting illegally, total assholes. If you've agreed to lease your computers time/bandwidth for something, what does that say about you as a person if you just turn right around and try to cheat your way out of it? It says "Hello, I'm a sneaky fuck who can't be trusted with a plastic straw much less any responsibility or a computer. Please tie me up in a burlap sack and hit me with tire irons, then throw the corpse in the river. Better yet, burn it."
Then again, that's just the impression I get
Was I the only one thinking that the name 'Gamma Flux' sounded stupid and cheezy until i realized that it's actually and appropriate name for a project dealing with radioactive radiation?
--
You can still use Linux - Just run a virtual Windows 9x using an emulator such as plex86 or vmware!
Don't forget that there are moral objections as well as ones to the safety of your box. Say that I'm a pacifist, but my box is being used to create terrain maps for missile guideance systems. Or to do particle simulations of the effects of a nuke.
Vintage computer games and RPG books available. Email me if you're interested.
Though they are only picking a few machines to run this test you should still - Sign up anyway - , it dosent hurt. Just think if this really takes off, you could leave your machine(s) running all day and possibly make just enough money each month to cover the electricity bill you ran up!
So, demand mileage when you sign up.
Here the ProcessTree partner statistics.
As you can see Linux is under 2%.
I thinks this is why they chose windows.
Yep, I'm their target market - college student. I have a very fat pipe, don't have much money, don't pay for electricity, and will do anything to make a couple bucks if it doesn't work.
Unlike the "pay to surf" people, these folks are offering to pay for something of value and you can't cheat w/ a little program that pretends to move the mouse every 3 minutes.
Maybe they're afraid that the linux people have too much skill and will be more likely to try and cheat the system so they're concentrating on the windows community?
__
. . . They've failed to turn a profit as the selected computers failed to stay running long enough to finish the project by the given deadline.
I'm wondering why they're choosing windows-based machines for this task. I agree that it makes a lot of sense to hand-pick computers depending on the requirements of the job, and I understand that this job requires 24/7 connections and all. However, it seems to me that it's easier to find a *nix machine that's truly on and connected all the time than a windows machine... especially in the realm of home computers, their target audience, most windows machines will be running 9x, which simply can't be considered 24/7. Is it because they have more development time invested in / more control over the windows software? Or is it just because that's what most of the people signed up for the service have?
I've had this sig for three days.
What do they use? I am assuming it is some EXE, but I couldn't tell from the article. If they used VBscript or JavaScript, the performance would be a lot slower than a C app on ONE computer (AFAIK)
:wq
The reason they are only using Windows seems pretty obvious to me. Windows has the largest desktop marketshare, and I'm willing to bet that the majority of people who have signed up at that site are actually running Windows.
So, logistically, if they really intend to be churning out projects in the future, it makes sense. It takes much less time to develop for a single platform per project.
What this also means to me is that depending on the size of the project, different pools of machines will be used.
-lb
The only downside to the project...they contact you when they need your processing...you're not always processing....if you are out looking for some extra cash...dont count on it.
Couldn't pay me enough to run arbitrary code for anonymous third parties.
Vintage computer games and RPG books available. Email me if you're interested.
Who needs SETI@home when you can just smoke a big fat bowl and talk to aliens anyway(s)?
permanently connected Windows-based machines
Presumably Windows would have to go without a reboot for 24 hours a day?
And it still crashes
It's not difficult to safely run untrusted code.
It's all a matter of how secure your sandbox is. In fact, it doesn't matter whether your code is Java or in native code - the only reason all these
distributed pay-you-for-cycles companies are choosing Java is not because it's more secure, but because they've got someone else to blame when a security hole is discovered. It doesn't really matter if it's the JVM or something else that's overwriting my files.
VMware, plex86, User-Mode-Linux - they're all on the right track: trap on the privilegded instructions and safely emulate them. Every bit as
safe as Java, but without the speed penalty. (This assumes of course that VMware, plex86, and UML have all the instructions handled safely, which can be hard to do on the X86 instruction set. But I'd guess that Sun's JVM has at least as many bugs and security problems as VMware does.)
What you couldn't pay me enough to do is to let anonymous third parties run my code. These distributed-processing people need to take a lesson from the online gaming community: Your clients will cheat, and they are smarter than you. The only way to catch them cheating is to re-run the computation, and don't bother with encryption or obfuscation, because remember, your clients are smarter than you. Also, don't even bother to try and keep "sensitive" data hidden from the people running your computation: you can't. Only send things to the remote clients that you don't care if they see, because they're going to.
The problem with re-running your computation is, of course, that it takes more Compute Time. Sure, you can just check up a percentage to take a statistical guess, but how much is your computation worth to you? If you're willing to pay for it you're better off to check everything again, because processing power is "cheap." Even still, for my serious calculations I'll do them in-house, thank you.
-Erik
Hello there, Armin Lenz from Distributed Science here.
Our decisions to launch with Windows client software first is a simple one: Linux clients are in the works but are under a contractual deadline, which is hard to meet by itself. As we provide updates, Linux clients will be rolled out as well, in particular because Linux boxes are considered 24/7 reliable (unless they are dial-up, of course).
A large number of the locations chosen for the first run are in Asia and South Africa. If you have a suitable machine and were not called upon directly, don't worry, the project is supposed to expand to several hundred locations over time and if you are in a sizeable city, chances are good that we'll have use for you there.
All this is independent of processing jobs that require computing time or storage capacity. Projects of this kind will be rolled out later, we are responding to customer demand first and try to get as much of the ProcessTree network utilized so the suppliers can earn some money.
ProcessTree - Isn't it time your computers started paying for itsel
Ahh, that's what I want. For my car to be a billboard for random corporations, for my friends to mock me for it, and get paid not nearly enough for the heckling I will endure.
Maybe you can use that $400 a month to get a prostitute, because I don't think you'll be picking up chicks in that car.
BilldaCat
This sounds like a great start for the program. I hope they secured their patent before they went public. I had the chance to write up a similiar patent proposal over a year before they started up. My bad for not following through.
My next point is their payment structure. From what I saw of it, their plan looks very similiar to a Multi-Level Marketing Plan, not unlike Amway's. It is admirable that they want to reward referrals but I'm not sure that they fully researched how people will try to take advantage of this. Maybe they need to reward referrals up to a maximum? That way you get more money from actually doing the processing than by referrals.
For a MLM, Amway's plan makes the most sense, especially compared to some of the stranger plans I've heard of but I'm not sure if it is appropriate as part of the business model for a distributed computing company.
We the Sheeple...
1. There are more Windows machines
While true this project needs a limited number of hand picked 24/7 reliable systems.. There are more 24/7 Reliable Linux boxes that are not otherwise employed (Not part of a larg IT network) and are just "personal systems" that would want in on this.
2. Linux users won't install strage software.
Distributed.net proves Linux users will (by in larg) do exactly that if motivated... As well as Windows and BSD and Solarus and well anyone really.
The number of people who worry about closed source binarys is significant but most will set such conserns asside... given the choice most would pick open over closed but not allways...
3. No Linux client yet
The IT sector dosn't think of Linux first.. they just automaticly think Windows first... so the Windows client was done first...
The Linux client is "on the way" but they couldn't wait for it...
Also likely the Windows client was a quick and dirty... to be replaced later with quality code.. Linux people are far more picky about this.. they can (and do) recognise shotty code eating cycles...
I don't actually exist.
When you look carefully at the requirements the issue is simply this: Pay someone else to spam the net for 'monitoring'. It's probably much cheaper for them to have you spam vs their bandwidth requirements. And you do all the damage, not them. No thanks. Run math problems, find by me, run spam searching, go elsewhere.
Since the project seems to have nothing to do with using spare computer resources and instead seems to use bandwidth, I wonder how the project factors quality of bandwidth into the reported statistics? If you are using a distributed environment to crunch numbers, everything is great, everyone crunches their numbers, the amount of CPU being simultaneously used by other processes on the machine is irrelevant because the speed at which the results are being computed is not relevant.
However, in order to monitor QOS on web sites, a clean pipe is necessary, because the amount of time that the "calculation" takes to complete *is* the actual desired information. Now, if I am downloading gallons of pron and warez over my T1, and this client software is trying to ping yahoo.com to determine site response time, how exactly does that give a realistic representation of yahoo's availability from my part of the net? It doesn't. I do not see any way that the client software could realistically determine any preexisting bandwidth latency prior to performing its QoS tests, so it seems like the results are going to pretty much be random.
Maru
I just had a look at my ProcessTree account and it seems they fucked up the accounts.. They had me down as having a lot of Mac G4/MacOS clients.. when in fact those numbers belonged in the Linux category (I have a Sun Sparc Classic that I put in the 'other' category and it somehow registered as a Mac G4 - I wish). I know it was set up right before, so it seems to have been corrupted. If you signed up for it, you might want to redo your account so they have the right information.
--
Delphis
I agree, it's called representative democracy but when they don't actually have to represent you it sounds a bit hollow...
I've often thought that the party system should be outlawed - they promise to represent you but then they get into power and vote for the party line, something opposite of what their constituents want.
There was a case of it in BC 5-8 years ago (I don't remember) where someone polled most of a district and found that 85% of people who responded wanted one thing - their MP still voted the other way.
I think they should be sued, for breach of contract maybe, treason at the outside... (it's basically subversion of the system.)
What's the sheckle to USD conversion?
Fight Spammers!
This just doesn't make sense from a "I want to make money with my machine's extra cycles" view. Their machine requirements aren't bad, but aren't the extra P166 I have lying around.
Their location, bandwith and other requirements dictate a decent amount of cash be spend by you, the potential money maker. And then, they haven't said what they'll pay. It can't be great though, because why else wouldn't they be leasing a real server from a commercial colo establishment?
Renting the sort of machine they reqiure and getting the typical 24/7 uptime a commercial venture can provide can't cost that much. It seems as if they just want to be shafting you, the cycle provider.
Okay, I know this is slightly off topic, but hear me out...
Yeah, it's cool that you can get paid for the use of your computer, but how about Getting Paid to drive your car!
www.autowraps.com and www.freecar.com are both websites that pay you up to $400.00 a month to drive arround with an advertisement on your car.
Sounds kinda silly to me, but I'd actually consider it.
That sounds a lot like reselling CPU time and bandwidth. I don't think ISPs will like this. They may declare that if you wanna do this you need a bussiness account. I know that Road Runner charges a lot more to businesses than consumers and I bet they aren't the only ones. Be careful...
I haven't read all the docs I could have, but it seems obvious. If the goal is website availabilty, then they might be testing for anthing that will knock the majority of users to the arms of Dr. Watson and his ilk: javascript errors from developers who use Macs and do not test on windows, download times based on certain IE and NS versions, etc...
I recent got the email about the paying client, and of course was like, finally! So I went and downloaded it and installed it, and initially realized my client was running Gamma Flux and not CSC. So I went to the Dcypher page and realized CSC is closed, yes the free one, Gamma Flux is still open. Then this morning I realized neither one of these are the pay client, and its this web checking biz. So my point is, that warning they emailed is a complete sham to get more people beta testing their other client software. So do yourself a favor and give your cycles to SETI, or *anything* else.
Just use VMWare and run the code in a protected virtual machine. That also solves the Windows problem: you can run a real Windows session on a Linux box.
Does the whole scheme make economic sense? Given the rapid depreciation of hardware, is it not cheaper to contract out to a bunch of datacentres which may have uneven loads (note that studies show web traffic peaks around 9pm). SETI@home worked for individual machines but that was a popular project that captured people's imagination and who were willing to "donate" cycles. If the cost of monitoring an individual machine + electricity cost + network costs + loss opportunity cost + risk, does it make economic sense? The risk factor of having unknown programs potentially loose within a coporate network is enough to give sysadmins heartburn, especially if someone puts in a trojan horse for data stealing.
It'll be interesting to look at their cash-flow distribution plan, not much point if you have to pay out 120% of income to offset marketing + cycle compensation time.
LL