Slashdot Mirror
Decentralizing Bittorrent
An anonymous reader writes "Exeem is a new file-sharing application being developed by the folks at SuprNova.org. Exeem is a decentralized BitTorrent network that basically makes everyone a Tracker. Individuals will share Torrents, and seed shared files to the network. At this time, details and the full potential of this project are being kept very quiet. However it appears this P2P application will completely replace SuprNova.org; no more web mirrors, no more bottle necks and no more slow downs. Exeem will marry the best features of a decentralized network, the easy searchability of an indexing server and the swarming powers of the BitTorrent network into one program. Currently, the network is in beta testing and already has 5,000 users (the beta testing is closed.) Once this program goes public, its potential is enormous. "
If it's allowed to be reached anyhow.. I have a feeling it's going to be tied down if it's the "next" big thing..
I'm not the devil.. just his advocate.
Didn't realize these guys were hackers, too. Wonder how many RIAA/MPAA scum got in on the beta test?
I think the biggest win is the ease of finding files. Theoretically this would allow file information to propagate, and I tihnk the most interesting problem that will be faced is stability. How do you make effective searches that do not loop around the network?
This could be a really cool development, and there is a lot of research in the EE/CS community right now going in to studying these decentralized networks. They show great promise!
The only time I ever had a problem with torrents is when downloading something very timely and popular, and the tracker would get soaked. This happened with both Fedora Core 1 and 2. Take away this one tiny problem, and you have a perfect technology.
So BitTorrent took the whole "everybody's on the same network" and converted it into "one network per file".... and now this new system puts it *back* like that? How is this different from every other p2p filewhoring system?
How can you truly decentralize P2P? Don't we still need to hit up a server that has a list of all the people? How can you track the trackers if you don't have a list of who is sharing? The only way I can think is just crude port scanning across subnets...can anyone clarify this for me?
This is excellent news from a user standpoint. I use Bittorrent for just about everything - downloading Linux distributions, game betas and, uh... other commonly downloaded files. But I always seem to be a bit behind the tracker, and when I go to download there are hardly ever more than 5 peers at a time.
What I want to know is: basically, this is an indexing server that will allow torrents to be searchable. What happens with multiple versions of the same torrent? For instance, let's say there are 2 torrent distributions of Gentoo, identical files within the torrents. It would seem this server would ideally be able to recognize the similarities and kind of 'merge' the files - is this possible?
M
What would really matter these days is anonymity. It's a bit late to develop yet another non-anonymous network, when the real problem is the risk of lawsuits...
I realize that full anonymity is going to be a problem, but at least some degree of deniability and limited IP address propagation would be a boon. SuprNova might have the name recognition to really give something like that a good start.
Save your wrists today - switch to Dvorak
How is Freenet not mentioned in this context. It is decentralized and other than the dropped packets / routing needed for anonymity it is swarming dowloads since any node might have the data you need.
RudeDude
Perl/Linux/PHP hacker
My IP address is out there now. A copyright owner has to provide SOME evidence other then my IP that I am the one distributing their copyrighted material. I know they are providing very little evidence now in their quest but only providing an IP is stretching it a little bit.
I think they're ignoring the fact that to be the "next big thing" requires being more than just incrementally better than what it replaces. Bittorrent itself is exponentially better than a FTP or HTTP server when demand is high. And Suprnova works quite well as it is, so I think it will be interesting to see whether Suprnova holds tough if people don't switch to the new technology fast enough.
1)With all the lawsuit attempts and legislation in the works, we still haven't seen filesharing development dwindle as much as one would expect.
2)The RIAA and their comrads are lawsuit crazy, but you haven't seen any "cease and decist" orders issued out to projets like this. A bigger thing to note is the fact that everyone seems to be a target - except companies like LimeWire who actually sell the P2P service and make money off of it (they get paid for the ads in the free version as well).
3)How the heck can any judge take these cases seriously, when, like one of my fellow posters made notice of, companies like Sony pratice the business tatics that they do. Their electronics division sells the mp3 players, but the record companies that they own forbid you to transfer the songs to mp3 players.
Go Figure...
Well, on the other hand, the more networks there are, the less users lose when the {MP|RI}AA gets around to sabotaging/suing-to-death any individual network.
In the old Napster days, there was no reason to make the trade-off of less content for better resiliance. Now there is. No P2P network is totally safe from meddling, so it's a matter of whether the resiliance effects of proliferation outweigh the content-reduction effects.
Of course, I can't afford home internet access, so I don't care either way right now.
Until they release some info about the inner workings of this app, there's not much to talk about.
There are serious problems with decentralising BitTorrent. One of the reasons that people have such good transfers on BT is that there is central tracker supervising particular file and knowing all users serving bits and pieces of this file. This way in case of high demand/high popularity files I achieve speeds over 1MB/s (yes, that's megabyte).
Depending on design choices you can have couple of trackers with subset of users on each of them, or every user seeding file has his own tracker. In first case your client wouldn't be able to use all cloud, and in second tracker would disappear when original seeder turned off his computer.
You can of course design some communication between trackers, or elections or some other magic, but it's too early to tell at the moment. I'll wait for more information.
Whatever they do, I hope that there will be some console based client for this, because asymmetric connections at homes plainly suck at upload (hence on torrent at download too), and I'd rather keep running my torrents on the server plugged into the fast network.
Robert
Bastard Operator From 193.219.28.162
You haven't received a C&D before? I cannot speak for most ISP's, only my own.
When a friend a few months ago received a C&D, it included his IP, the time and date of the offence along with some info on the file... the ISP in question, just sends a warning letter on the first offense, unplugs you until you call in for the 2nd and unplugs you for 30 days on the 3rd, and this is only from the complaint.
You might have an open wap, or a trojan on your PC, or any number of other legitimate reasons of why your connection was used to DL unauthorized material in a way that was not authorized by you... it doesn't matter to this ISP, a complaint is a complaint, and as per the safe harbor provision of the DMCA, they act upon it.
Were you to get a 3rd C&D on this ISP and get unplugged for 30 days, you could always haul them into court and demand to get reinstated, but by the time you got your hearing, the 30 days would be over.
More so, it is the ISP's network, and by using it and paying for their services you agree to their rules, and if their rules say "we may suspend your access at our slightest whim should we receive information saying that you had allegedly infringed on someone's copyright", they can.
This is not the court system we are dealing with, this is free enterprise, and is little different than me refusing to personally associate with anyone under 5'0", simply because that is how I do business.
Help Brendan pay off his student loans
I was one of the beta testers and I can say it works
.torrents? Can the .torrents be used in a normal bittorrent client? Is there a console unix client? I'll still be able to use my btlaunchmany.py wrapper script right?
It's slashdotted, so I can't find these things out for myself. Is this just a p2p app for
Give me Classic Slashdot or give me death!
Decentralized Bittorrent? Wake me up when they have secured Bittorrent and then I'll listen.
My ISP, Mediacom, scans my network packets to determine if I'm grabbing a torrent of questionable nature. If they see it, they'll send me a nasty email. Hence, I'm on the edonkey networks now because BT is clearly not an option at the moment. I'm sure they'll scan those packets, too, at some point.
Unsecured BT is fast, sure, but if your ISP is snooping...well. And illegal or questionable content aside, it'd be handy for distributing other files to people in a more secure manner.
Or is this out there and I'm just missing something?
Blog,Twitter
How would encryption save you? Once the protocol is understood, one need only build a client (or modify an existing open source one) to download a particular file, all the while logging each and every IP it gets a bit of the file from. Once done, you verify the file is infringing on someone's copyright, and bam, you let the C&D's fly against all of those IP's you logged.
Unless you can completely remove the IP from the process (ie go through some anonymous proxy), you are still catchable once BayTSP and crew decide to track people on a network.
Help Brendan pay off his student loans
Any new big thing needs absolute anonimity. I already worry for all of the innocent civilians out there using bittorrent now to get their favorite shows and movies. I'm sure their transgressions are all being logged for future lawsuits.
And yes they are INNOCENT. Here's one good reason why. We first must ask, why did the founders of the US constitution feel it was important for accused criminals to be convicted only by a jury of peers?
I believe this is because they knew that honest citizens doing honest activities will often run afoul of the law, especially in a broken government where England (back then) or corporations today make all the laws. The jury of the peers is built into our criminal justice system in order to prevent just this kind of thing. I mean the hope is that a jury of bitorrent users will never convict a fellow bit torrent user. That's probably why we're only seeing civil lawsuits today by the RIAA and the like. I think I criminal jury trial for file sharing would be quite interesting.
"...being kept very quiet" (until now)
/.ed, so I can't speak directly to Exeem, but it sounds from the blurb like these features are a possibility. Hope it's free in all senses.
This should be good... BT is without question the fastest p2p app (in fact, the only thing that has ever topped out my 'net connection), but it needs two features to kill off the others in my book:
1. Search - it's no fun to rely on third party websites to find things. Hopefully now we'll be able to do this.
2. Anonymity - BT could use an option for a system like Freenet's for making it really hard to tell who's serving who. Combined with the distributed nature of BT, it would be difficult to prove anything at all about BT users.
The article is
Here's another thought: the current BT system is really good at dispersing new content, like distro ISOs and TV shows, through RSS feeds from central websites. It would be cool to be able to subscribe to network-wide custom feeds, to stay informed about new files that match certain criteria.
perl -e 'foreach(values %SIG){$_="IGNORE";}while(){}'
With the IP addresses still out there, wtf is the point?
.torrents off of suprnova.org and onto the users. This way we wont have to deal with tracker downtime due to overload/ddos.
The point is to move the bandwidth load of providing all those
Give me Classic Slashdot or give me death!
Yes, those damn pirates! For example, for last couple of years I engage only in illegal downloading of TV series.
I mean, it's despicable, how can people distribute and watch TV shows that are normally viewed for free!!! This must be incredible loss of revenue for rightsholders, mustn't it? Especially when they don't care enough to release those shows on DVD.
Frankly, I'd prefer them make those shows strait-to-dvd so I could buy them for 20-40 bucks a season. Maybe this way no power-hungry fatass exec would cancel shows like Farscape, Firefly, Jeremiah or Angel. No more fucking games with "ratings-scheduling feedback loop", just simple rules -- either it sells with a profit or it doesn't.
Robert
Bastard Operator From 193.219.28.162
Anonymity is probably the most important feature of modern P2P applications and crucial for their survival. From my understanding the first version of Exeem does not guarantee anonymity yet. How soon will it be implemented? A great distributed P2P application is Freenet: http://freenet.sourceforge.net
Already done with Ants p2p that uses a torrent type protocol for file trasfer it also encrypted end to end and point to point and uses Virtual IP's who needs exeeem.
n /h ome.jsp?page=custom&xmlName=ants
http://www.myjavaserver.com.nyud.net:8090/~gwre
The great thing about BitTorrent is that you are being pointed to a known file. You can judge for yourself who points you at a given file by what website is hosting the tracker.
It is possible to provide the same chain of trust in a decentralised network, just digitally sign the release notes and hash values. By checking the digital signature you can check that the file has been announced by the same group or person that announced previous files.
If content announcers issue digital certificates you could pass on the digital crtificate of release groups that you trust to your firends in the same way that you could reccommend them to download via suprnova.org
According to them no linux version for a long time. This probably means no open source either. Forget it.
/. will give eXeem a pretty big audience though."
Some other stuff:
"The main problem with kazaa is that it doesn't have hash system which means that if you make MP3 with same name and same size that's already on the network and someone downloads one part of this file from you the MP3 will be corrupted. (This is exactly what RIAA did to kazaa). And since people don't delete bad MP3's from their computer you have more and more of this files in the network. And here is where our client is different you wont be able to corrupt files in the network because they have hash.
One more difference from kazaa is that we wont have entire folders of files on the network only those that will be manually uploaded from users. Kazaa has so many viruses because users don't even know they have them on there computer. So I personally think that we will have a lot less fake files on our network and we also plan to implement rate system so that if people find fakes,
viruses, spyware in one of the files they will vote it as bad so hopfully not many people will download that file.
What we are trying to do is bring best of P2P world and best of bittorrent together."
About eXeem replacing suprnova:
"That's a reporters view on it. Remember, they probably know next to nothing about eXeem, and are doing what reporters do best: bullshit.
I'm an Exeem beta tester that's been trying to give it a fair shake. I'll probably get banned just for this post, but here's some general details about the new client.
First off, it's in beta testing, but it's not ready for beta. It has some serious isses at the moment. Torrents disappear off the network for no reason is just one of them.
Second, they don't have 5,000 beta testers. They sent out 5,000 serials, but my best guess by looking at the network is that there are less than 1,000 actually testing it and never more than 200 or 300 people running it at the same time. They actually sent out new serials to all the 5,000 beta testers because they didn't have enough people.
Third, it lacks the details. With most BT clients, such as BitTornado and G3 Torrent, you can see all kinds of details about the file you're trying to acquire, how many seeds, portions of seeds, how many complete copies are distributed amongst the peers if there are no seeds. Exeem lacks all of these details.
Fourth, it doesn't use bitTorrent. It's based on bitTorrent, and uses libTorrent, but it's not a torrent. It's their own unique format. Exeem will not be compatible with other BT clients. It's use their client or don't connect. It almost appears to be a Kazaa rip off with bitTorrent features.
Fifth. 'But it's open source? Why can't we just write our own clients?' From everything I can tell, they have no intention of making this an open source project. They're talking about the type of ads they want to put in it.
Sixth: Pr0n. A lot of people like Suprnova.org and other torrent sites because there is no pr0n. Exeem has an adult filter, but 'Adult' is one of the more popular categories for Exeem users at the moment.
Exeem will not replace bitTorrent. The problem I see is that Exeem is being developed by the same guys that run suprnova.org. Whether Exeem ever works or ever becomes popular, will they take down what appears to be the most popular torrent site on the web because of it?
There are more problems with Exeem, but these are the major ones that I see. I'm sure some of the coders of Exeem will be reading this post. Please feel free to tell me where I'm wrong and why.
Aero
Please stop hurting America -- Jon Stewart
keep bit torrent like it is, if you make it as easy to use as napster, kazaa, whatever, it'll take off mainstream & next thing you know **AA will be going after it.
god i hope noone makes an easy to use usenet searcher/downloader program.
Well, first, the existing leechers can failover to the new seeder/tracker, and that's still useful to let them all finish even if nothing else.
Second, whichever seeder is elected as tracker can advertise itself for indexing onto the "tracker search" network I proposed in my upthread post. So then new searches find the new tracker.
Third, the web pages or whatever that are linking the torrent can (manually?) re-link a generated new torrent for the new seed, which has meanwhile kept the torrent alive rather than letting it all fall to pieces.
What I am missing from the enormous array of file sharing tools is a simple one. Sharing a virtual LAN with your friends. There are many VPN server/clients out there, but they are all point to point. What I would like is some software that emulates a workgroup LAN, so you could use simple SMB or FTP filesharing over a trusted, encrypted, distributed network. The tricky part would probably the broadcast packages and the IP range.
The big deal is that if you are caught running a big mp3 server you are dead meat pretty much everywhere. In the U.S. that can mean an appearance in federal criminal court on a felony charge of copyright infringement.
Being lawful and being ethical are completely different things, and many times mutually exclusive. Try not to confuse them.
One is that the Bittorrent protocol is thusfar the best protocol for transferring large files. The clients are designed to transfer large files. The Edonkey/Kademlia protocol exists to transfer large files as well, but it is just not as good as Bittorrent. It is much slower.
p2p has to be looked at as a process. There is the search for information. There is the response to the search. Then there is the request to download a file. Then there is the download of the file. Each of these parts is separate and important. In Bittorrent, only the last part, the download is decentralized. The prior parts are not decentralized, are not p2p - even the request to download goes to a centralized torrent.
Despite this, Many people figure that Bittorrent's partial file sharing, protocol attributes and program attributes are what make the downloading good. Of course, having a good source of current holders of the file - partially or fully, is important, as is having a good hash of the file, or multiple hashes in the case of Bittorrent. But this can all probably be done via p2p as well.
As far as the comments on hashes and file integrity, this is not a problem at all. There are many ways to deal with this. If you want, you can still have a central torrent, but you could only check it once instead of many times. Or maybe there could be distibuted PGP signatures of the validity of certain hashes.
As far as other comments, I'm interested in this so I'm glad to know, although I agree its vaporware until release.
As far as Freenet, encryption, IP addresses and so forth - I think for technical innovation reasons, unencrypting, non-masking p2p technologies need to be developed for now. I'm also glad, alongside this, anonymous, IP masking, encryption-capable p2p networks like Freenet are being created. And once p2p becomes mature, I hope the technologies implement any encryption and anonymity that does not put in too much overhead. Turn it on by default, and let people manually turn it off if they want.
As far as copyrighted material and so forth, I really could give a damn. Big corporations hate the idea of sharing, and trying to kill something like Linux or a GPL open p2p protocol and client is instinctive to them, just like the enclosure of the commons was.
And as far as non-centralization being one of the benefits of Bittorrent, and decentralizing ruining it, I completely disagree. As I said before, file integrity and hashes are not a problem, you can do PGP signatures on hashes or something. Any problem can be dealt with. Bittorrent is good because it is the best protocol to deal with partial file sharing of large files. Any of its centralized features can be decentralized, some of them very easily, as I'm sure Freem is doing.
Sneak into a ballgame, and explain this to the security guard as he throws you on your rear on the pavement outside.
Copyright infringement isn't stealing, but its not legal either.