Slashdot Mirror
Advances in Decentralized Peer Networks
PureFiction writes "Peer networks are gaining some
attention these
days given advances in much more
decentralized
search
architectures and
swarming
distribution
networks.
Research has indicated that these decentralized networks are resistant to
legal and
technological attacks.
The continued proliferation of
broadband
and wireless networking
will ensure
pervasive deployment
of distributed
peer networking infrastructure
that will drive significant innovations in
personal and
community
digital communications services."
that every word couldn't be a hyperlink.
Chicago, what?
But I wish he had posted a link or something.
Offtopic, but...
15 links on a story!! And it's not a Slashback... now I see why my stories never make it...
I guess it would've paid off to listen more in algorithms class. Next thing you know, someone's going to put one of these NP circuits in P, just so music trading can be made easier/more efficient.
That's what I thought when I saw all those links.
"I think all foreigners should stop interfering in the internal affairs of Iraq"
-- Paul Wolfowitz, 7/21/2003
"Research has indicated that these decentralized networks are resistant to legal and technological attacks. The continued proliferation of broadband and wireless networking will ensure pervasive deployment of distributed peer networking infrastructure that will drive significant innovations in personal and community digital communications services."
So what P2P networks are resistent to "last mile" control tactics? Port blocking. Speed limiting. Quotas. Remember control of P2P has one thing in common with security. It doesn't have to be perfect. It simply has to be innconvient enough that people give up and go away.
As a security person, I hate these peer to peer applications: there are so many worm strategies which can exploit these, creating fast and stealthy attacks.
As such, if I was in charge of corporate security administration, I'd ban them completely. At the universities, I'd packet-shape-them into a much lower priority.
Unfortunatly, they are growing considerably more stealthy as a result of these legal attacks and the effects of packet-shapers. This may be a good thing for those who want their 1337 WAR3Z, D00D!!, but is rather unfortunate for those who want to create secure systems.
Test your net with Netalyzr
I'm not trying to troll here. Since napster all i've ever seen P2P used for is piracy of music, applications, pretty much whatever you wanted.
:) I saw no less than 100 users sharing the file. Nobody is supposed to have it but I do. Thanks to the decentralized nature of P2P there's no accountability. Websites that were hosting the file got a nasty attorney letter though.
I am that pot calling the kettle black. I am your average joe user. I have kazaa and routinely use it for downloading music. Yes I am a criminal. As are %99.9999 of all other P2P users.
I understand the benifits of P2P, each client acting as a server and bonding the collective bandwidth of all the clients together. Yes I know it can be used for free speech, and I know for legitimate file distribution it can't be beat. That's just it though, it's never going to be legitimate without some type of DRM.
I downloaded doom3 alpha (Sorry Carmack, it kicks ass though
Here is my slashdottish geek comment. P2P creators need to start focusing on making their clients good for legitimate uses. For example, I think it would kick ass if the distro's started using P2P for their distro's, or a P2P based web server/browser. Anything to turn it from a black to a white sheep.
*Note to mods
Sorry I don't mean to be Mr. Obvious here, but I just feel any future P2P doesn't stand a chance if it doesn't have a legitimate foundation to stand on. The RIAA & MPAA has already proven what a great team of legal sharks they have and can overcome any technological advancements made in P2P
my 2 cents
As more people get broadband, you'll be able to get faster downloads from more sources on peer to peer networks. Thank god we have slashdot to point out the obvious.
Everyone is entitled to their own opinion. It's just that yours is stupid.
Then again, I have that problem anyways. Blimey!
Is it just me, or did that mess of links and words look strangely like things I read at work?
If I wanted to read horrible PC-edited, buzzword-laden, over-link tripe, I would look at the stack of paperwork on my left that is being shunned in favor of reading SlashDot.
------
Today's Top Deals
Why is it that for all these articles about P2P networks, I have never once seen Direct Connect mentioned?
There is more information (pirated software, movies, games, tv shows, etc) available on those networks than on any other network I have ever seen (except *maybe* napster). Everyone who uses it must share, and essentially must have a fast connection.
But there's never been a mention of it...
?
AND FOR THOSE OF YOU IN A HURRY... A- R- T- I- C- L-
E-. -- S-
U- M- M-
A- R-
Y-
Sure, but will it play on your Trusty Palladium PC?
Totally on-topic co-incidence; just a little while ago I ran into a pointer to the MNET project.
When the Mojo Nation P2P effort ran out of money they released the client code under the GPL (it is in Python). MNET is a stripped-down variant of the code (without the micropayment kruft) delivered as an API with some sample apps, one of which is a file sharing application.
So, if you know a little Python, you can grab MNET and whip up your own distributed file-sharing network. Cool!
- -
Are you an SF Fan? Are you a Tru-Fan?
Okay, I guess while a bunch of that is interesting, I'd like to point out that the whole point of the 'Internet' originally was to have a decentralized network of machines, so that in the event that any of them was destroyed or disabled, we'd still be able to keep the network running. Knowing that a decentralized network is more resistant to physical damage is OLD NEWS. We knew that TWENTY YEARS AGO.
:)
That other stuff is neat, though.
It is early days yet, but at least this demonstrates the type of things it can be useful for.
At a quick glance I swear that it looked like the headline read: "Advances in Decentralized Beer Networks"
I was thinking, "The geeks have really gone too far this time."
The global economy is a great thing until you feel it locally.
Using P2P technology (I'm thinking Freenet may be the best starting point) it should be possible to build a distributed file system with built-in redundancy and access permissions such that every PC shares files with every other PC, yet none are dependent on any one other workstation. Everyone can reach the files they need no matter where they are (ideally they don't even know which files are local and which are remote), nobody can reach files they shouldn't see (e.g., personell records), and every file is backed up somewhere else. If Joe's PC crashes, just plug in another one and Joe's back in business. All of Joe's data is backed up elsewhere on the network, and all the backups on Joe's PC lost in the crash are restored (re-backed up) onto the new PC.
Feasible? Perhaps, but unfortunately I don't have the time to persue it. Still, if it could work it's an example of legitimate Peer to Peer file sharing.
If all this should have a reason, we would be the last to know.
agreed, but that interface isnt going to convert a lot of people quickly.
members are seeing something, your seeing an ad
Wow, just imagine. Someday everyone will have a P2P network of their very own! Er, wait....
Why are P2P networks such a security nightmare? Because they can support very fast worms, very stealthy worms, and/or very easy to write worms.
1: Given a "client->server" vulnerability (one that can be exploited by the initial initiator of communication), you can write a very fast topological worm. Probably ~1 minute to infect all on-nodes of a monoculture peer to peer network.
2: Given a "server->client" vulnerability (one that can be exploited by the responder to information requests), you can write a probably fast contageon worm. Best guess is probably ~1 day to ~1 week to infect everything, but very, VERY stealthy.
3: No vulnerability at all, a "bait worm" which is simply a smarter version of Gnuman or similar such nasties.
Test your net with Netalyzr
Piracy would be buying a CD, making copies of it, and selling those copies to people for a profit. In this case, you entered into an agreement with the copyright holder of the CD, and are violating applicable laws.
When you download something off the Internet, you are under no obligation to be aware of its source, or the license agreements associated with it.
Even posting copyrighted material on the Internet is not illegal if you have not been explicitely notified that you are in violation of the copyright agreement. The notice on your CD case counts, I believe, if you own the CD.
So if you download something off the 'Net, you're well within the bounds of legality, and you can retransmit that until the copyright holder asks you to stop.
American laws are so strange...
Anybody with knowledge of writing software knows there's going to be now way of stopping peer to peer, at least not technically.
The only thing the powers that be can do IMHO is to make it illegal to share files and folders on a publicly (free) available network which contain copyrighted data.
When we get to the point where everyone has a phone/pda with 1 GB of storage and Wifi built this is going to be a real serious problem for the record companies.
Have you ever heard Epiphany Radio? It's a shoutcast station I used to listen to, until I ran into a 12 user limit imposed because the broadcasters couldn't afford the bandwidth to support many users. Thanks to peer to peer technology (the Gnutella protocol), I can once again listen to this station, via their peercast stream.
It does not concern me at all that P2P content is currently mostly illegal warez, music, movies and porn. It's an excellent testbed for the ability of the technology to withstand technological and legal attack.
It does however comfort me that P2P works and is widely deployed. The average person is familiar with it and knows how to use it. If corporations and government ever get too oppressive, P2P is an powerful tool that the people can use to preserve their rights. Current P2P networks could easily be put into this type of service at need.
I see the net just now starting to realize the ideals that is was founded under. Decentralization, free access to information for all, everyone can easily be both producer and consumer, etc. All trends (wireless, weblogs, P2P) point to this.
I've thought about this one a lot. Port blocking and other targeted attacks on P2P wont work (they can always be circumvented in a variety of ways), but it is still possible to attack the usage of bandwidth itself - and I believe this is what they will do.
The feds are both for sale and beholden to the news outlets; a double whammy. The major media guys own them. The FCC is dismantling all of the "competition" parts of the TA96, and allowing telecoms and media megamergers right and left. I'm already reading about bandwidth "shortages" that don't exist; the writing is on the wall. The broadband suppliers (especially cable, who are generally owned outright by the media giants, i.e. Time Warner) are publicly floating ideas to end broadband as we know it. In other words, no more unlimited bandwidth; instead, quotas, caps, and per-traffic charges.
May sound crazy, but it's already happened in Canada and Australia ("test markets"); it could start here in as soon as the next 12 months. The only wildcard are the telcos, who "need to come on board" with content control rather than salivate over the potential flood of customers to DSL if cable gets crippled and/or suffers a massive traffic-based price increase. Then again, as the telcos consolidate, it gets easier and easier to "play ball" with a smaller and smaller boardroom full of people, and this is the boardroom at its best: in engineering the switch to bit-based-pricing, we're talking about price-fixing on a massive, even global scale. And who doesn't want to make lots and lots of money?
There are complications, though, in that there are a number of legitimate uses for bandwidth on or near the scale that P2P uses it. Video games, for instance, and especially the new class of broadband-only games, would be destroyed by such a move - and there are billions at stake in that industry alone. So if the collateral damage to other wealthy players is high enough, we might have some hope.
Perhaps the discontent of millions of citizens would be a factor too, if this were still a democracy in more than name.
Want to Know How to Cheat the GPL? Read On!
Unfortunately, the legal work we did early on when designing the system only prevented people from suing us (the code creators) and it did so by pushing liability off on to the users. That was the closest that one could really get to safety given the structure of the DMCA. Contrary to the widely held fantasy among decentralized p2p systems, "willful blindness" is not a valid defense against DMCA attacks -- something that I think the upcoming Kazaa et al. trial is eventually going to reveal after all of the appeals and other legal wrangling is worked out.
BTW, the only app that mnet provides is a publish-retrieve shared data system identical to the old mojonation (sans distributed resource management), file sharing is not "one of" the apps for the API, it is the only app.
"Peer networks *DING!* are gaining some attention these days given advances in much more decentralized *DING!* search architectures *DING!* and swarming *DING!* distribution *DING!* networks. *DING!* Research has indicated *DING!* that these decentralized *DING!* networks *DING!* are resistant *DING!* to legal *DING!* and technological *DING!* attacks. *DING!* The continued proliferation *DING!* of broadband *DING!* and wireless *DING!* networking will ensure pervasive *DING!* deployment *DING!* of distributed *DING!* peer *DING!* networking *DING!* infrastructure *DING!* that will drive significant *DING!* innovations *DING!* in personal *DING!* and community *DING!* digital communications *DING!* services *DING!*
Cheers,
Bowie J. Poag
Users can backup and restore their own files; user self-help for cases of "pilot error" and random system crashes means that IT does not get the "hey, I accidentally deleted my presentation that I have to give to the board in 2 hours, can you help me out?" call that interrupts whatever they were doing. Doing daily/hourly snapshots to an online storage mesh also means that the backup tape monkey does not need to spend time trying to balance backup runs to fit within various backup windows, you can let the distributed system handle the snapshots of current data and use the tape for weekly/monthly offsite archive.
There are lots of cool and interesting uses for p2p outside of simple content distribution, you just need to look a little harder.
This submission has no less than 15 links in it! I'm still not sure where the real content is. Worse, what appears to be the main link points to another few line discussion that references yet another article.
Posters, and especially editors, Please don't post/allow submissions that are filled with nothing but links, and links to links. You may think you're providing more information, but most of those links are just noise. This seems to be an emerging trend on slashdot, and I think it's a very bad one.
Taking a slight tangent for just a moment, sadly, the biggest problem with slashdot is there's never any real discussion (mostly that would involve the editors) about slashdot itself. There's constantly re-posts of material, and poor quality submissions. This needs to be fixed somehow, but can't happen unless the slashdot editors start talking about this problems and not just remain silent.
AC REPLY ""sadly, the biggest problem with slashdot is there's never any real discussion (mostly that would involve the editors) about slashdot itself.""
the problem lies not just with the editors but with everyone on slashdot (especially whoever happens to be moderating). anytime people (like you) actually comment about slashdot (even when its not a troll and is a cogent point, like yours is), it is moderated down, offtopic, or just ignored (as your post was).
it's the built-in self-censorship that's the problem. not only does it limit the voices heard loudest on slashdot, but it prevents fixing of the very socially dysfunctional way slashdot works a lot of the time.
With the predominance of wireless open networks, I'm surprised no-one has suggested all hooking up all the wireless nodes together into one giant peer-to-peer network, not just for file transfer, but for Internet as well.
Get a wireless hub and a wireless card that can call up your neighbours hubs. Then when a request comes through, throw it to your neighbours (hot potato style, ala FIDOnet) until the request comes to a computer that can handle the request (perhaps a machine with a land-line connection).
No ISP. No last mile. *Everything* is the last mile.
Probably about 5 years ahead of its time. Need enough wireless saturation to make it work.
Here's yer link :)
Live for the present, learn from the past, and dream of the future!
sites like www.sharereactor.com and www.jigle.com are doing a lot to help this. A quick browse of those sites, and experiencing d/l a popular file from the network, and you'll realise they are actually quite popular, esp amongst people that actually share. Its design also means less leeches, so for every person on the network, equals about 100 or something like Kazaa.
I.O.U One Sig.
I think one of the most exciting things about P2P is that the costs are borne by the consumer, not the publisher. This holds true with Freenet, and holds true with Gnutella and Kazaa as well. If I have a popular, non-commercial web page, I the publisher have to pay to keep my pages up, and the more popular the pages are, the more I pay. With P2P however, the consumers act as distributors as well, so whether it's an audio file, video, web page or whatnot flowing over Freenet/Kazaa/Gnutella, the cost for me to publish is not there. I like this because it means popular, non-commercial media can spread by virtue of popularity, and the Internet can't be monopolized by people who can control the flow of information simply because they own the printing presses and distribution networks. I also think this is what makes P2P something disdained by the powers that be. The RIAA/MPAA's activities are just the short-term, tactical activities of the people who fund them, I care very little for their rationale and look for what the long-term effects would be if they were fully successful, and it doesn't look good - I don't really care about the supposed morality of their authority or whatnot, I'm only interested in the effects of their actions. Thousands of years ago, the concept of property in this economy of scarcity was created. Recently this concept has been extended to the spectrum, to bits of information flowing between me and a friend's computer with it's economy of non-scarcity, and even to species themselves. If we do not build a technological foundation that helps put power in the hands of the people (like Gutenberg, Wozniak, and Justin Frankel), accompanied by social movements that protect people from the powers-that-be using law and authority to dominate them, I think we are headed into a dire future.
If you think about it, technology frequently "gets it's feet wet in dirty water" so to speak.
Home vcrs allowed people to watch porn in their home as opposed to sleezy theaters.
The internet allowed users to view....well....porn. Since then, well you know.
Mp3's in the begining were nearly EXCLUSIVELY used for piracy. Although indie bands and average people were quick to adopt it, for a while the only mp3's on the internet were without permission. Now however there are perfectly legitimate sources of mp3's available.
If I was old enough to have the required insight, I'm sure this could be applied to lots of technology going back centuries, particularly entertainment technologies.
-Chris
--an unbreakable toy is useful for breaking other toys--
I second the poster above. May I suggest that the Slashdot editors do some editing for a change.
Each story should have a clear message, with the main link clear from the story. There should be no more than two supplementary links.
Gnutella servents can already get around port blocking - they can use random ports every time if the user enables the option (there's a field in gnet pong packets for the port number). As for getting past content screening, there's a simple solution: diffie-hellman key exchange before _any_ other network conversation, followed by data encrypted using the symmetric cypher du jour. ISPs _can't_ know what data is being transferred without executing a man in the middle attack. Since this has to be done at the beginning of the connection, the ISP has to modify the very first data sent by its customer, the ISP has two options:
1. Not know what data is being transferred.
2. Block or mutilate all traffic it can't identify.
If it decides it really doesn't want to know what's being said, great. It's none of their business. If it decides to block any traffic it can't identify, it will face customer backlash. Users may not be aware and active enough to fight monthly transfer caps, but they sure do complain when they can't connect. If it tries to execute man in the middle attacks, it will succeed if the traffic is p2p, but it can't know ahead of time whether or not that will happen. If it tries to snoop an arbitrary new connection that turns out to carry p2p data, the ISP will be able to see the data that's being transferred. If that connection ends up being something non-p2p, the first part of the connection will be garbled beyond recognition and the connection attempt will fail.
Of course, there's nothing you can do about transfer rate caps and monthly quotas, but if that becomes a problem it's really your own fault. If you not only continue to elect representatives that voted for deregulation and caused the gross lack of broadband competition we have today (or even allow them to be elected by not telling everyone you know how owned they are by special interests), but also stay with companies that implement monthly data tranfer caps, you really have nobody to blame but yourself.
You can make a difference. Being a vocal, informed advocate works. A polite but firm letter containing the word "constituent" works better. The same letter but with money works best. Enough money virtually guarantees you'll be heard and obeyed.
For the record, I turned 18 after the last election (but I put my voter registration card in the mail the day after my birthday), and I'm switching from Cox (max 2GB down per day or 30GB down per month, whichever is lower/max 1 GB up per day or 7.5 GB up per month, whichever is lower [Yes. They _just_ implemented this in my area.]) to DSL (no transfer limits, but mind-bogglingly incompetent tech support).
Oh MAN, it will take us HOURS to slashdot these servers down! We should get paid for this stuff.
Jouni
Jouni Mannonen | Game Designer, Consultant
Spyware database: http://www.surasoft.com/cgi-bin/spysearch/spydb.cg i?search=direct+connect&header=Software