Forgot your password?
typodupeerror
Security The Internet

Researchers Outline Targeted Content Poisoning For P2P Data 201

Posted by timothy
from the subscribe-to-your-shackles dept.
Diomidis Spinellis writes "Two USC researchers published a paper in the prestigious IEEE Transactions on Computers that describes a technique for p2p content poisoning targeted exclusively at detected copyright violators. Using identity-based signatures and time-stamped tokens they report a 99.9 percent prevention rate in Gnutella, KaZaA, and Freenet and a 85-98 percent prevention rate on eMule, eDonkey, and Morpheus. Poison-resilient networks based on the BitTorrent protocol are not affected. Also the system can't protect small files, like a single-song MP3. Although the authors don't say so explicitly, my understanding is that the scheme is only useful on commercial p2p distribution systems that adopt the proposed protocol."
This discussion has been archived. No new comments can be posted.

Researchers Outline Targeted Content Poisoning For P2P Data

Comments Filter:
  • by Darkness404 (1287218) on Thursday July 23, 2009 @10:48PM (#28803091)
    I'm not exactly sure "researcher" is the right word here. From the paper

    Abstract: Today's peer-to-peer (P2P) networks are grossly abused by Illegal distributions of music, games, video streams, and popular software. These abuses have resulted in heavy financial loss in media and content industry. Collusive piracy is the main source of intellectual property violations within the boundary of P2P networks. This problem is resulted from paid clients (colluders) illegally sharing copyrighted content files with unpaid clients (pirates). Such an on-line piracy has hindered the use of open P2P networks for commercial content delivery. We propose a proactive poisoning scheme to stop colluders and pirates from working together in alleged copyright infringements in P2P file sharing. The basic idea is to detect pirates with identity- based signatures and time-stamped tokens. Then we stop collusive piracy without hurting legitimate P2P clients. We developed a new peer authorization protocol (PAP) to distinguish pirates from legitimate clients. Detected pirates will receive poisoned chunks in repeated attempts. A reputation-based mechanism is developed to detect colluders. The system does not slow down legal download from paid clients. The pirates are severely penalized with no chance to download successfully in finite time. Based on simulation results, we find 99.9% success rate in preventing piracy on file-level hashing networks like Gnutella, KaZaA,Area, LimeWire, etc. Our protection scheme achieved 85-98% prevention rate on part-level hashing networks like eMuel, Shareaz, eDonkey, Morpheus, etc. Our new scheme enables P2P technology for building a new generation of content delivery networks (CDNs). These P2P-based CDNs provide faster delivery speed, higher content availability, and cost-effectiveness than using conventional CDNs built with huge network of surrogate servers.

    This isn't unbiased in the least. Sure, arguably it is "research" but calling them researchers from an university makes them seem neutral at best.

    • by s-whs (959229) on Thursday July 23, 2009 @10:56PM (#28803147)
      ] Researcher is the wrong word.

      I was thinking the same thing. But not necessarily based on them being biased, but for this: Why would anyone want to 'research' this? I can understand making a protocol resilient to poisoning (same as making a computer resilient to virus attacks, there will always be a-holes trying to mess things up wether legal or illegal), or making it faster, adding some nifty features perhaps. But poisoning to prevent illegal sharing with the pathetic argument that this hinders commercial distribution? What kind of a researcher is that? A RIAA paid one I'd guess. Possibly as valuable as those 'researchers' for tobacco companies who said there was no health problem with smoking.
      • by Darkness404 (1287218) on Thursday July 23, 2009 @10:58PM (#28803161)
        Exactly, I was reading into the article thinking it would be presented as a vulnerability or proof of concept that could be exploited by the RIAA, not that the entire thing seemed to be written especially for the RIAA.
        • by siloko (1133863) on Friday July 24, 2009 @01:56AM (#28803903)
          Researchers find a topic that interests them and follow through on some hunch. When they have found out something potentially publishable (the meat and potatoes of a researchers career) they big it up. This abstract reads exactly like that - "we did some work and this is why it's the most important work in the world" - the fact that the spiel coincides with the RIAA party line is probably coincidence.
      • by sukotto (122876)
        People who develop new weapons are researchers. I don't have a problem with calling them researchers.
    • by Kuroji (990107) <kuroji@gmail.com> on Thursday July 23, 2009 @11:30PM (#28803329)

      Well, here's the thing: by having this information out in the open, people can look at how it's done and look at the protocols they use, and find out whether such vulnerabilities could exist. Sure, it might not help anyone right now if they're vulnerable, but it does mean that the protocols that people use in the future are a lot less likely to have such weaknesses that allow for data corruption.

      Copyright or not, when you have the ability to corrupt data on a whim, the network is quickly rendered useless.

      • by Darkness404 (1287218) on Thursday July 23, 2009 @11:35PM (#28803351)
        But it wasn't presented like that though. It would be one thing if it was "Hey, your network can be exploited if you do this, this and this" but instead its "Your network can be exploited by this, this and this, because of this you can do -insert illegal stuff- to get revenge on those evil filesharers". I mean, seriously the stuff you read in 2600 about exploiting things to make a profit seem to have less bias than this. At least a bunch of those articles say "please only use this for information".
        • Re: (Score:3, Insightful)

          by ZosX (517789)

          Aren't there laws against DOS attacks? If you jammed the RIAA's network you would surely go to jail if caught. They should leave the law enforcement to the police. Its too bad nobody can seem to get them on racketeering. They extort millions (heh, literally apparently) from the american public and at the same time have not paid millions of dollars owed to the artists that they supposedly represent.

          • Re: (Score:3, Insightful)

            by Darkness404 (1287218)
            Yes, and the RIAA doesn't seem to care. Just look at how they used MediaDefender ( http://en.wikipedia.org/wiki/MediaDefender [wikipedia.org] ).
          • by paganizer (566360)

            I see one good thing about this; they positively could not have poisoned Freenet, (unless they are talking Freenet 0.7, maybe) but every additional Freenet 0.5 user makes the network faster and more anonymous; by trying to screw with it, they made it a little better.

            plug: gotthefire.net

    • by cortesoft (1150075) on Friday July 24, 2009 @12:04AM (#28803497)

      of COURSE they aren't real researchers. The summary writer mistakenly thought the study authors were from UCLA, which would mean they would have been some of the smartest, unbiased, amazing people in the world. However, they were actually from USC, meaning they were spoiled, unprofessional, RIAA lapdogs who also smell.

      And yes I happened to go to UCLA, but that is besides the point.

    • by Jurily (900488)

      We developed a new peer authorization protocol (PAP) to distinguish pirates from legitimate clients. Detected pirates will receive poisoned chunks in repeated attempts. A reputation-based mechanism is developed to detect colluders. The system does not slow down legal download from paid clients. The pirates are severely penalized with no chance to download successfully in finite time.

      Oh, this cracks me up. Did anyone notice notice how this doesn't mention bittorrent, which AFAIK makes up 90% of the possibly infringing content? Of course, anyone who's seen a torrent client in action knows that clients sending bad data are banned fast.

      Now that I think about it, this "researcher" should rank high on the "Best ways to make money and improve your karma" list. He's obviously a better way to drain RIAA money than lawsuits :)

    • Re: (Score:2, Insightful)

      by Odinlake (1057938)

      ... with no chance to download successfully in finite time.

      That is mathematically speaking a pretty silly statement (as there obviously is some non-zero chance of obtaining each piece), moreover so considering the next sentence which says they had a 0.1% failure rate.

    • by hairyfeet (841228) <bassbeast1968@gma i l . com> on Friday July 24, 2009 @03:41AM (#28804431) Journal

      So in other words they just want to steal the P2P networks from those that actually built up the things and turn it into an iTunes store, only one where the cheap bastards won't even have to pay for bandwidth. Nice. Just when I thought they couldn't be even more piggish than they already are. It just goes to prove that just when you think they've scrapped the bottom of the barrel and can't actually go any lower, if you lift up the bottom of the barrel and continue digging, you can get even lower. Nice.

      Meanwhile they rob from us and our kids by eliminating the public domain thanks to eternal copyrights, and screw you out of first sale with crap like DMCA and DRM, which they pay to have rammed up our butts with treasonous bribes. Very nice. These bunches are the only ones that can make CEOs at tobacco companies and South American drug lords not seem so scummy.

      And for all the countries getting USA eternal copyrights forced down their throats? I'd like to say as an American I'm sorry, we didn't actually want that crap either, but we only have a two party system and both sides have sold out because all our politicians are whores to big business. Maybe you'll have better luck dealing with the multinational cartels than we did.

    • by Alsee (515537) on Friday July 24, 2009 @04:06AM (#28804541) Homepage

      I'm part way through the research paper, the article summary is just plain wrong.

      There is no vulnerability here. They CANNOT poison Gnutella, KaZaA, and Freenet, eMule, eDonkey, Morpheus, or any other existing network with this technique. To quote the paper: Presently none of these P2P networks has built with satisfactory support for copyright protection.

      The "problem" they want to "solve" is that existing networks to not possess adequate support for poisoning attacks. This paper proposes creating a NEW additional P2P network. They propose deliberately building in special support to ENABLE poisoning attacks.

      While I'm sure the RIAA will eagerly read it over while dreaming of world conquest by releasing their own deliberately crippled "legal P2P network" where they get paid for each authorized client-to-client transfer. As far as most readers here are concerned, this is a completely non-newsworthy story, the contents of this paper are completely irrelevant and harmless. There is absolutely nothing new or surprising about the fact that you can deliberately make your software insecure and you can deliberately leave it vulnerable to poisoning. Yes, a P2P new network could be built Defective By Design.

      -

      • Re: (Score:3, Insightful)

        by Alsee (515537)

        I'm not sure if I missed the last line of the summary in my haste to read to the PDF file, or if the summary was updated, but the last line of the summary is correct and it pretty well refutes the rest of the summary-as-written. The earlier statements in the summary about success rates in blocking particular existing networks are wrong. Those blocking percentages are modeled results *if* those sorts of networks were to become paid access networks implemented this deliberate poisoning capability into their d

    • by MrMr (219533)
      I'm surprised there wasn't a big disclaimer at the bottom of the paper about RIAA funding; like the ones you see nowadays on many medical studies...
      Could they be sneakier than big pharma?
      (Btw. I only read the last page of the FA, honestly)
    • Re: (Score:3, Insightful)

      by mcgrew (92797)

      These abuses have resulted in heavy financial loss in media and content industry.

      Bullshit. It's been shown that music pirates spend more money on music than non-pirates, and the same is probably true of movie pirates and software pirates, too. They've declared war on their best customers.

  • Actually (Score:4, Interesting)

    by Renraku (518261) on Thursday July 23, 2009 @10:50PM (#28803101) Homepage

    Actually, poisoning P2P networks as a commercial venture could be prosecuted as theft-by-deception.

    Stealing bandwidth is a crime. Downloading songs isn't, if you aren't profiting form it.

    • Re:Actually (Score:5, Insightful)

      by Darkness404 (1287218) on Thursday July 23, 2009 @10:51PM (#28803107)
      And so is DDoS attacks, but that sure didn't stop the RIAA from using MediaDefender ( http://en.wikipedia.org/wiki/MediaDefender [wikipedia.org] )
    • by timmarhy (659436)
      so those big warnings on every dvd i've ever rented that state copying this dvd is a federal office are lieing? fyi, i know slashdotters never RTFA but you take the cake for not even reading the summary - this doesn't work on small files like songs..
    • by VShael (62735)

      "Downloading songs isn't, if you aren't profiting form it."

      Depends on your locale, doesn't it? Not all laws are the same in every location.

  • by wigaloo (897600) on Thursday July 23, 2009 @10:53PM (#28803127)

    Two UCLA researchers published a paper in the prestigious IEEE Transactions on Computers that describes a technique for p2p content poisoning targeted exclusively at detected copyright violators.

    What's to prevent poisoning legal p2p? There are plenty of examples of copyrights being inappropriately asserted. The technology itself doesn't discriminate.

    • Re: (Score:3, Funny)

      by Darkness404 (1287218)
      Exactly, just go to YouTube and you will see the DMCA abused left and right. (Well, and if you read the comments page you will find the rules of spelling, rules of actually saying something along with the rules of grammar and common sense to be abused too....)
    • Re: (Score:3, Informative)

      by TechForensics (944258)

      What's to prevent poisoning legal p2p? There are plenty of examples of copyrights being inappropriately asserted. The technology itself doesn't discriminate.

      The article says the method works only on P2P networks that have adopted the authors' proprietary PAP protocol. That's not likely to be many of them.

  • by mewsenews (251487) on Thursday July 23, 2009 @10:55PM (#28803139) Homepage

    Humans had discovered methods to speedily and automatically transmit mountainous volumes of data. It was a new frontier, a utopia where information was shared peacefully between the people who wanted to see it. And what was its downfall? Not the anarchists, or the communists, or the Islamic fundamentalists, but the so called leaders of the free world.

    "We had to do it," they said, "there is such a thing as too much freedom."

    • Re: (Score:3, Insightful)

      by basementman (1475159)

      Get over yourself, the method doesn't do shit to bittorrent, the most popular p2p format so it's basically useless. If anything this will just get idiots off limewire into onto a decent network.

      • Although I agree (about LimeWire, KaZaA, etc)... the only reason this isn't happening to BitTorrent, is because they haven't figured out how yet, not because they think it's some infallible, untouchable, system nor that they think everyone should be using it instead of the others.

      • by Yogiz (1123127)

        One problem with bittorrent is that it has a centralized tracker. You see what is happening to The Pirate Bay. If legal issues are of concern, I'd say that it's the bittorrent guys that should start moving onto a more decent network. And if that is to happen, we need to eliminate problems like content poisoning.

        • Semi-centralized, really. There is no clean way for me to connect to a "different Kazaa network", yet mostly anybody can host .torrent files and a tracker.

        • One problem with bittorrent is that it has a centralized tracker. You see what is happening to The Pirate Bay.

          This may be of interest. [reddit.com]

          For those who can't be arsed to follow the link:

          TPB has been owned by a company for the last years since the raid so nothing there will really change except the names of the owners. The talk about TPB are going to be a pay site is wrong, the CEO that said that does not know what he is talking about. Now, the BIG change is that the tracker is going to be outsourced to a new formed company that wont know what they track, just that they connect peers, and the torrent listings will

  • Wow (Score:3, Funny)

    by taucross (1330311) on Thursday July 23, 2009 @11:06PM (#28803209)
    Poisoning the well. What an insightful revelation. Surely it's never been done before, maybe they should throw a patent on it.
  • They sound more like wannabe whores to me. How is this blatant soul-selling behavior legal and prostitution is not?

  • Freenet is gnutella? (Score:3, Interesting)

    by Anonymous Coward on Thursday July 23, 2009 @11:12PM (#28803247)

    I was curious as to how they were poisoning Freenet, which should be robust against this with its Forward Error Correcting.

    According to the paper, Freenet falls under the category of the "Gnutella family" (p.2). The Freenet Project that I know is in no way related to Gnutella.

    Are they referring to a different file sharing program by the name of Freenet, or is this statement of theirs just plain inaccurate?

  • Freenet (Score:5, Interesting)

    by evanbd (210358) on Thursday July 23, 2009 @11:12PM (#28803249)

    The paper won't download here, so I'm asking without RTFA, but how can this work against Freenet [freenetproject.org]? Do they discuss Freenet in the paper at all? Freenet does chunk-level hashing, and the network enforces that the data matches the hash at all steps. Nodes returning invalid data will rapidly get dropped by their peers. Attacks like this are something that Freenet is explicitly designed to prevent. Also, the anonymity guarantees that Freenet makes would make it hard (potentially very hard) for them to identify a single user, let alone "collusion".

    I'm forced to wonder whether the researchers mention Freenet at all, or if the poster is simply lumping Freenet in with other p2p apps that it has very little in common with. (Bittorrent and Freenet should be similar in some ways to their resistance against this attack, but Freenet's strong anonymity guarantees should make it more resistant. The fact that a node engaged in widespread poisoning will have trouble even staying connected makes Freenet even more resistant.)

    • Re:Freenet (Score:4, Insightful)

      by Anonymous Coward on Thursday July 23, 2009 @11:20PM (#28803283)

      They lump Freenet into the category of "Gnutella-like networks", and say that their attack against gnutella should also work against Freenet since it is Gnutella-like (p.2 and p.12).

      In other words, it is as you said, they are lumping it together with other networks.

      It makes me question the quality of their research if they think that Freenet is so similar to Gnutella that the same class of attacks would work against both.

      • Re:Freenet (Score:5, Interesting)

        by calmofthestorm (1344385) on Friday July 24, 2009 @12:11AM (#28803517)

        This is utterly absurd. The verification on freenet is based on asymmetric crypto. If they haven't broken that, the most they can do is flood the network with corrupt chunks, in which case the software will just start dropping peers who send too many corrupt packets at too high a rate. Translation: you need # of bad guys >> # of good guys to have much of an impact on network quality. And of course it's complete trash against a darknet, but I doubt these guys know what that is.

        Given the subject matter, weasel words, and shoddy methodology, I'm about as worried about this as I am about the zombie communist terrorist invasion predicted for 2012.

        • by evanbd (210358)
          Most of the data verification on Freenet is based on SHA256. There is a well supported mechanism for signed keys (SSKs), but those are almost always used to simply redirect to the hash-based keys that use SHA256. Signatures in Freenet are based on DSA (slightly different lengths than the standard specifies, but the math is identical).
        • by n30na (1525807)

          Given the subject matter, weasel words, and shoddy methodology, I'm about as worried about this as I am about the zombie communist terrorist invasion predicted for 2012.

          I believe you mean zombie communist alien vampire terrorist invasion

      • It's entirely possible that the authors do fundamentally believe in the rights of the copyright industry, but that doesn't mean they might not be frightfully ignorant of any number of closely related technologies.

        In fact my experience has shown me that fundamentalists tend to be the most narrowly focused people I meet (whatever their beliefs).
      • by Kjella (173770)

        They lump Freenet into the category of "Gnutella-like networks", and say that their attack against gnutella should also work against Freenet since it is Gnutella-like (p.2 and p.12).

        Except it won't because freenet isn't p2p it's p2swarm.... a client can request data with the right "magic code" but all the nodes inbetween would cache it and all the pirates get it from one of the non-authenticating nodes. Note that this is really all a stupid authentication system, the sending peer could simply ask the master server "is this an authenticated client too?" and send poison data if not. This is basicly already done and better with private torrents - with the rights flags set the tracker will

    • the network enforces that the data matches the hash at all steps.

      But what enforces that the hash matches the title, as opposed to a cuckoo egg [hand-2-mouth.com]?

      • by evanbd (210358)
        Nothing -- that's a key distribution problem. There are various people working on the general spam problem for Freenet through web of trust type solutions. Those would extend to cuckoo egg type spam as easily as any other spam. Get your keys and your torrents from someone trustworthy. Right now, that's done by message board apps, and people could easily post complaints about or verification of a specific file.
    • Re:Freenet (Score:4, Insightful)

      by MikShapi (681808) on Friday July 24, 2009 @01:03AM (#28803711) Journal

      Freenet is a hard target. Arguably, the hardest of them all today. It's also the least popular.

      The studios are playing a money game. Bang for buck. They want maximal deterrence for minimal spend.

      Much like virus-writers aim viruses at the highest targets on the "adoption-by-the-masses"/"soft-bellyness" index, RIAA go-getem's do the same thing.

      FastTrack - high adoption, soft belly.
      Torrent - high adoption, not-so-soft... and segregated into lots of independent share-specific networks.
      Freenet - low adoption, practically impossible to break.

      It's a no-brainer. They've got no reason to go for the last. They may be greedy scum, but they're not that stupid with their money. Freenet would need to be adopted by the masses and get a ridiculous amount of media exposure to even pop up on their radar. Their goal is not to technically "stop filesharing" altogether, they realize that's a waste of money and effort. Their goal is to mitigate it by taking pot-shots at just the targets that are easy to break, and leave the harder ones alone (for now).

      Being an informed geek, that actually makes me really happy. In a nutshell, It means we won.

      • by evanbd (210358)
        Yep. Freenet and TOR are both quite good at what they do (though they solve very different problems). Unfortunately, Freenet has a small userbase (current estimates ~ 10k). I think it needs more applications that work on top of Freenet before it will see more than very slow growth. It would be very interesting to see enough Freenet adoption that people took notice. There's plenty of reason to think it's reasonably secure, but you just don't know until someone actually tries to attack it.
      • Re: (Score:3, Interesting)

        by Deliveranc3 (629997)
        Freenet is a hard target. Arguably, the hardest of them all today. It's also the least popular. High overhead will do that to you, ISPs make the overhead look bigger by shunting encrypted data into low speed transmission. Ugly ugly stuff, arguably illegal and the government is too scared to step in. The studios are playing a money game. Bang for buck. They want maximal deterrence for minimal spend. I'm concerned that's the happy fun version, really they're looking to transfer us all over onto their p2p net
  • by cortesoft (1150075) on Friday July 24, 2009 @12:00AM (#28803471)

    These guys are from USC, not UCLA. As a UCLA graduate, I am extremely upset that anyone would make this mistake. USC students and professors are smelly, unclean, spoiled children who work for the RIAA. UCLA students and professors are the opposite.

    Never, EVER, confuse us again.

  • ...given the absolute rot most people are downloading on the networks. I mean honestly. What could be more poisonous than a Britney Spears song? I'd say let the downloaders have the content. Can't think of anything more poisonous.

  • I read the summary as them finding a way to create a p2p network of 'customers' (clients who pay to be in your p2p network where you deliver paid content) and protecting yourself from the 'customers' who 'collude' (e.g. hacked client s/w?) with non paying client s/w to allow non paying customers to get the content. I don't think it's about subverting an existing network, it's about protecting a network from subversion. If so then the techniques could presumably be used for other purposes, poisoning surveill

  • Paper summary (Score:5, Informative)

    by creidieki (110659) on Friday July 24, 2009 @01:33AM (#28803827) Journal

    As a comp sci grad student, here's what I got from a quick reading of this paper:

    Imagine that you're a content provider, with paying users. You've decided to distribute content to your users by running a Gnutella-style network. How do we make sure that only paying users can get our content? After all, it's an open network.

    We start by sending some sort of magic timestamp-thing to all of the paying users. I didn't read this part in much detail. Anyway, the paying users can all identify each other somehow. They mention that it maintains privacy.

    Some of your paying users (the "Clients") are good, virtuous folk, and they're running the Happy Authorized Gnutella software you gave them. Others (the "Colluders") are running Evil Hacked software. No matter what you do, the Colluders are going to send chunks of your precious data to the "Pirates" (anyone who hasn't paid you).

    Normally, we'd expect our Clients to ignore requests from our Pirates. This paper instead suggests: let's obligate the Clients to send poison data to the Pirates! The Pirates won't know which chunks are bad; they'll only find out that the file is corrupt once it's finished downloading. The Pirates won't be able to get a good copy, and they'll give up and go away.

    And there's one other great thing: we can set up *fake* Pirates, and check which users aren't giving out the poison they're supposed to! So we've served data to all of the Clients; we've identified all of the Colluders; and we've defeated all of the Pirates.

    (Bittorrent has data integrity checks for every chunk, instead of every file; that's why it's not vulnerable to this attack...I mean business model).

    In summary: This paper describes a way that a company can charge for distributing their own content on a peer-to-peer network. It only works if they control a centralized "transaction server" thThat's why no one has ever at organizes the entire network, and if they control the software of all the "honest" people. They can't destroy our existing networks with it, and it doesn't prevent anyone from turning around and posting the file to BitTorrent once it's downloaded.

    The tone of the paper is definitely not as neutral as I feel it should be. What they're trying to say is "there's no obvious way to charge people for running a Gnutella server, because pirates will eat your lunch. But we think we have a way." But it definitely feels like they're putting moral force behind what's really a network algorithms result.

    • As a note though, from a fellow CompSci grad student (though I didn't bother reading the paper), it IS possible to "poison" BitTorrent. I've read a handful of papers on the subject in the past year or two, but most of them focus on things like DHT entry poisoning or other similar techniques, and not on compromising the data itself.
    • Very well put. I didn't have space to explain this in the submission's summary, but this is the gist of the paper.
  • I once accidently did a minor DoS attack, when I was starting to write my own P2P client for the Kad network used by eMule, etc. it kept returning the same IP in response to every directory lookup.

    Sorry to whoever had 127.0.0.1 back then, if your connection went down it was my fault.

    (I don't remember the actual IP)

  • *knock on Pirate Bay's office door*

    "What the hell is that?"

    *Hannigan the traveling salesman enters*

    "Good evening, little girl, is your mommy home?"

    "Dude, this is the pirate bay office."

    "No worries precious, I'm sure your birthday party can wait a few moments longer until the dreadful pointlessness of existance crushes your youthful hopes and dreams like mine have been two decades ago, leaving me a hollow broken shell of a man seeking solace in cheap whores and nickel whiskey shots on hungarian hobos."

    "Who

  • I only use the eDonkey network for small files (music, images, books), and BitTorrent for the big ones, so that thing won't even affect that.

    The only bad thing is, that now rare bigger files (like lossless music, very specific software, etc) will be hard to get.

    But I really do wonder. Because as far as I know, no network out there works without checksums. So poisoning will be detected, and then circumvented (e.g, manually).

  • by Cajun Hell (725246) on Friday July 24, 2009 @11:08AM (#28807349) Homepage Journal

    is that you don't know who your peers are. They might not even be "peers" in the everyday commonly-understood sense.

    Solution: remove anonymity, or at least replace it with pseudo-anonymity. I don't know who the guy that signs his chunks with keyid 0xDEADBEEF is, but I know he's never sent me garbage in the past. The owner of keyid 0xF00C1000 sends me chunks that don't match up with the rest of the content. My computer has a hard disk. It can remember things like this.

    Gnutella blacklists mediasentry IPs. IPs are ephemeral. What they ought to do is use a signed protocol, and blacklist bad signing keys. Or better yet, greylist everyone by default and whitelist the ones who show a history of integrity. No wait, program the client to do all that, and don't distribute any lists at all.

Theory is gray, but the golden tree of life is green. -- Goethe

Working...