On 05/12/2014 12:11 AM, David Barrett wrote: > In a recent conversation about piracy and whether it could win, we > concluded the primary vulnerability of BitTorrent was its susceptibility > to copyright enforcers just connecting to everybody in a swarm, > downloading one block, and automatically sending out a takedown notice > to the owner of that IP. This in theory could be "scaled up" incredibly > easily to effectively monitor and police all the top torrents out there. > The only defense Bittorrent has is the "blocklist", which if I > understand correctly, is just a list of IPs to block entirely. So my > first question: > > 1) How are these blocklists created?
1) Pick a random number between 0 and 2^160. 2) publish it to your favorite DHT (azureus and mainline are popular) 3) wait for nodes to claim they are peers for your random sha160(torrent). They come rather quickly. Makes using said DHTs annoying for using said DHTs for p2p introductions. > 2) Would it be possible to create a "big data" automated blocklist service? Repeat the above. > Given this, I'm curious if anybody has built an automatic blocklist > generator that basically accumulates a list of every "leacher" -- Leacher seems impractical. After all every peer starts with 0 blocks. However claiming to be a peer for a non-existent torrent seems rather malicious. > Q: What's to prevent this service from being blocked? > A: This would be a service built into the tracker itself. Every swarm > can have many trackers, and of course there are many swarms. To fully > block this service would require blocking all trackers -- which if that > were possible, would already be done. Trusting 3rd parties on reputation is problematic. But if you publish 5 non-existent torrents and 5 real torrents you should have a good idea on which clients are faking it. _______________________________________________ p2p-hackers mailing list [email protected] http://lists.zooko.com/mailman/listinfo/p2p-hackers
