P2P BitTorrent Tool Could Replace Pirate Bay
With the US and other G8 countries trying to outlaw The Pirate Bay and its ilk, an anonymous reader suggests that a solution may have emerged out of Cornell University. A new open-source project called Cubit is an Azureus plugin that provides decentralized approximate keyword search of torrents in the network.
They haven't even passed their unconstitutional law. And here you are already defeating it. You're supposed to give them a few minutes of satisfaction.
If you can read this, I forgot to post anonymously.
And with Guitar Hero replacing actual music, soon there won't be anything left to steal! Now *that's* innovation.
As I contemplated when AT&T started saying they want to fight piracy on the wire, the most effective way is for the ISP to cooperate with the MPAA, where the MPAA gives a graph of "These people are exchaning a large copyrighted file, block it".
If ISPs move in that direction, this defense won't help, and thats probably the bigger threat for blocking P2P piracy, as there are always countries of convienece to set up piratebay like operations.
Test your net with Netalyzr
And when can we expect *that* to get shut down?
Just a couple months after everyone has stopped using it and is using something else.
Seven puppies were harmed during the making of this post.
To my knowledge, Kademlia uses exact keyword searching, not approximate searching. While distributed hash tables are a fairly effective decentralized searching mechanism, it's tough to move them from exact-match searching to more general searching.
Other DHT systems are also used to list peers for trackerless torrents and to find peers for particular files on networks like eMule (by searching by hash).
The article from yesterday about Verizon and Comcast's pledge to support Bittorrent also includes information about Cubit.
Use of this will significantly increase the number of fake files uploaded.
At least TPB allows file comments which allows fakes to be spotted pretty fast.
Also, do not forget about the amount of traffic private torrent sites get - which this is not a real alternative to.
à_à
* The network is much more efficient.
* All this network is sharing is torrent metadata (.torrent files), while a BitTorrent client is doing the real transfer.
* Their keyword searching system, while allowing for finding the k-nearest keywords, is not fully general like searches on a Gnutella-like system could be.
"The Internet interprets censorship as damage and routes around it." -Joe Gilmore
Ubiquitously - A Ubiquity Developer Community
I've read the GP's post and I've been pulling out the Old Constitution trying to figure out where he's coming from.
We, the US, are governed by the rule of law. And sometimes, the rule of law is very unfair for a few of us. BUT, it will correct itself eventually and to be honest, I prefer "eventually" to a bloody revolution. I mean "bloody" in the "folks are dieing in the streets" bloody - not the British version.
I think there is already a mostly-unused torrent-tella-like system. It's really a very good solution, since Gnutella provides very powerful searching and BitTorrent provides high-bandwidth data transfer. This is actually more like using eMule's Kad to share .torrents.
As I mentioned somewhere else, though, people won't move from the index site + centralized trackers + a BitTorrent client until enough indexes and trackers get shut down that they need a new solution.
It seems to work the other way. It'll get shut down about a year after a better solution is developed and about a month before everyone starts using that better solution.
No. The decentralized-tracker problem is a ton easier than this problem, and there are already multiple decentralized-tracker solutions. Decentralized trackers are just done with simple distributed hash tables. What they've done is make a fancier DHT system for finding "near matches".
Encryption doesn't help. You can participate as clients of a swarm to get the identity of the members of the swarm, which is the information the ISPs need to block the swarm.
Test your net with Netalyzr
AEIOU: open-source anonymous internet currency
Gnutella (LimeWire et. al) has more than one way of searching. Through Ultrapeers, Ultrapeers and OOB-replies (e.g. not routed back through Ultrapeers) and Mojito (DHT).
Using Gnutella to search/index .torrents is already a long time feature of G2 (Gnutella 2, though it is NOT the successor of Gnutella), with Shareaza being the main client for the G2 network (along with very basic support for Gnutella, BitTorrent and eDonkey2000).
DHT-networks can be more efficient, but they are also vulnerable to attacks and pollution and are somewhat lossy.
It only takes one man to change the Wisdom of the Crowd to Tyranny of the Masses.
I think that the hardest part of adding search to any p2p system is that it is too easy for malicious users (*IAA thugs) to poison search results, and I don't see anything on their page that deals with that.
To design a reliable search system, you need to have a good rating system, and a solid trust model. At the same time, you need to avoid making the trust model so tight that new users cannot get any search results (freenet).
Also, I think it should be noted that a lot of bittorrent usage is moving towards the subscription model, so people should be able to search for channels as well, not just single files.
I am interested in seeing where this project leads, but I don't think people will be completely abandoning the well organized, well moderated torrent sites any time soon, but it will be nice to be able to search quickly for files without needing to open a browser.
There was. 1776 to 1783.
--Won't that be grand? Computers and the programs will start thinking and the people will stop. - Dr. Walter Gibbs
Tough luck. Time for a career change.
AEIOU: open-source anonymous internet currency