ProcessTree.com first paid distributed computing project.

[Train]

From the february newsletter:

<< This means that supplier Internet connections are used to do requests on websites run by customers of Envive. These requests are evaluated for response time and download speeds and that information is compiled according to geographic locations and ISPs used to improve the websites. Very little processing power is required for this job, but a fairly good Internet connection is a must to avoid a noticeable impact - something you know we try to avoid. >>

umm.... so they are gonna use my machine to ping servers? Ya right, I don't want that kinda traffic on my network. It seems processtree has a real problem getting sponsors, maybe TA should chip in an buy a few terrahertz for RC5

 

[Robor]

I think I'll pass as well. I'm sticking to SETI@Home!

Rob

 

[Mobster]

Why don't they first try to keep their stats-server up :|

 

[phule]

ProcessTree has seemed fishy from the start...this doesn't help them any in my mind.

 

[Train]

well i wouldnt say fishy,
but borderline unproffesional and mainly desperate, its obvious they are just in it for the money, which is why they took such a crappy project. No way in hell would i call pinging servers distributed computing, what a crock.

 

[cirrus1]

/me didn't bought a T-bird for pinging webservers :disgust:

 

[xtreme2k]

Actually $12.5 doesnt seem that bad per computer. Since it dont eat up your CPU power, you can still run RC5/Seti watever and still earn the $. And if you have a broadband connection, just pinging wont take up that much bandwidth. I better read the terms and conditions carefully though.

 

[ViRGE]

Humm, what I said initially, was in hindsight, not made with enough informeation to make an accurate judgement. While I still don't agree with what Dcypher's using it's tech for(I think they could do great things with it), I'm glad to see they're enhancing it.

Ugg, it allways pains me for someone going into the DC field for cash, then resorting to something like this for said cash. ProcessTree, you make me sick.

 

[Fardringle]

The good news is that now at least part of the ProcessTree team will be switching over to &quot;their&quot; paid projects, so we can take back the lead in Gamma Flux..

 

[networkman]

Yeah, what ViRGE said.. goes for me too! :disgust:

 

[vss1980]

Fardringle,

Only a small percentage of PT members are actually going to do the paid jobs. In this case I think there are only a couple of hundred people provisionally marked to do this project.

It certainly will not be available to everyone who signed up, and not because of CPU limitations.

 

[charrison]

I am quite displeased to see the negativity to the industry's FIRST commercial internet based distributed application. This is the first customer we landed, and if this is successful, Envive is going to want more nodes. We are also paying out more per node than is quoted here. As well as bonuses to users that are good suppliers. We are working on other contracts that will be wider based, but the market is new, so acceptance has been slow. Making full use of your computer does not end at the speed of your cpu and the requirements for this job is to keep it unobtrusive to users. We are going to be gathering REAL measurements for web site performmace. These results cannot be reproduced in a server farm somewhere.


Yes, we are in this for money, but we are going to share it. If you run with one of competitors, you MIGHT win a prize for doing work. You run with us, we WILL pay you. We are going to keep our free projects going, for the people that are in this for fun.

If you have any questions, feel free to ask. I will glady clear up any misconceptions that exist.


Chris Harrison
Distributed Science

 

[narzy]

I belive you are out of line if you critisize a company for Venturing into a new industry, this project may not be &quot;Distributed Computing&quot; by your terms, however it is distributed Computing, it is done over many computers, if you don't want to do the project that is fine, however you have no right to slander a company for choosing to do a project and take the risk of doing it. I am sure alot of hard work, time, and effort went into setting this project up by the ProcessTree team. give them the respect they give us.

oh and btw, dont bring up the issue when we were accused of haboring a hacker, those were misworded coments, and they did correct them.

You people who slander process tree for trying to make money like any other company on this planet are :disgust:

 

[mechBgon]

I think we all know that it's best to do the projects we enjoy. Speaking for myself, I found out by chatting with charrison that I know a lot less than I thought I did about the new project, and the complexity of its implementation. Let's be open to a little edification.

 

[vss1980]

I am gonna give a quote I said a while ago.

'It needs to be carried out on computers located across the world due to its nature, it cant be dumped on a super computer. Therefore it qualifies as a distributed computing effort. In some ways its more distributed than RC5 or others would ever be.'

I think many of you have overlooked what the point of the project is. It isn't a ping test. If they wanted to do that, they could set up a game server and advertise it over the world. Looking at the most basic level, it is a network logging process.

I'm not surprised at the reaction of some here. I mean, here we are in a position where we could try to find cures to AIDS or other diseases using spare CPU time, but would rather try to use a brute force attempt to crack a code.

No, I'm not on the Distributed sciences pay-roll, but I wouldn't have a problem taking money from them for just using my computer to 'ping a server'. Jesus thats a sweet deal.

As for being in it for the cash. Come on, half the companies in their respective fields are in it for the cash. I cant believe the managers at Red Hat Linux keep releasing new version because they feel hard done by Microsoft, no, they are making a profit (and a good one from what I hear), and while they are I am pretty sure Red Hat 7 will not be the last.

 

[mindless]

Im not sure how usefull this will actually be. In order to test web site performance the connections will need to have 0% packet loss. Just 2% packet loss will scew the results. Since most home internet connections aren't exactly reliable this makes it even worse. I think at this time website performance monitoring should be left to people such as Keynote who stick servers in datacenters world wide and measure them from there.

But I do like the idea of getting paid to use my CPU. I hope this project does go over well so future projects could be molded out of it.

 

[Train]

I wouldn't take it to hard, any negative comments in this thread torwards PT should be taken the way any smart business would take critisism: ways to improve. A good business needs to listen to its customers, and I believe you have found a great place here, we are a forum of Distributed Computing enthusiasts, some people here have been doing it for years, and many of us really understand the concepts behind it, and the potential to do great things. If commercial distributed computing is the new frontier, the people of Team AnandTech, DPC, Slashdot, HackZone, and other groups will be the first &quot;cowboys&quot; to explore this new world. Just because I say a few negative comments doesn't mean I'm not completely excited about a commercial distributed computing company, I am watching very closely and hope for the best. I complain about Microsoft products every day, but I still use them, and will probably buy the next ones.

Perhaps me being a very technical person, I'd like more detail about EXACTLY what the downloaded client would be doing, WHEN its gonna do it, for how long, etc, that would make it much more acceptable by the tech community, in my opinion.

 

[otlg24]

Mindless,

Re: Im not sure how usefull this will actually be. In order to test web site performance the connections will need to have 0% packet loss. Just 2% packet loss will scew the results. Since most home internet connections aren't exactly reliable this makes it even worse. I think at this time website performance monitoring should be left to people such as Keynote who stick servers in datacenters world wide and measure them from there.

Well you are forgetting most websites care about last mile performance. Envive has been testing from Data Centers for some time. However, we have seen that their customers want to know who content arrives in the last mile. PS. Envive isn't the only established company who is looking at using distributed computing to take these measurements. EVERY major player in the market place is looking at doing it.

Regards,

Steve Porter
CTO, Distributed Science

PS: For those that know me, I used to be otlg22, and then 23.. but I keep changing e-mail accounts and losing passwords

 

[otlg24]

Train:

Re: Taking it hard

It's not so much that we take it hard, but rather we are disappointed by the knee jerk reaction from people that Chris and Myself have personally corresponded with for a period of YEARS now. Guys like Virge especially hit hard, after I for one gave him 500K of my old d.net blocks. Now I don't hold anything against him, or expect any favors in return, but just a knee jerk reaction without ASKING us seems rather close-minded to me personally.

Regards,

Steve