« This can only mean one Thing! Injection! | Main | Problems with Peerguardian »
Perfect Deniability
By erdgeist | February 12, 2007
I stumbled across this post the other day. Copy right enforcers nowadays seem to have trouble catching people ignoring their copy rights red handed. Black lists seem to spread fast enough to lock spies out. So all they’ve got is IP lists bittorrent trackers supply them for a given torrent.
As it seems, this really is enough for them to send out threatening letters to ISPs. While we certainly do not encourage anyone to share files they’re not supposed to, we do not feel well being missused as evidence distributors. Spoofing an IP address is not that hard and knowing that some trackers even parse IP addresses from query strings presented to them is not helpful either.
Whats worse: since bittorrents tracker protocol is based upon http, a webdot like
<img src="http://denis.stalker.h3q.com/announce?info_hash=01234567890123456789"/> on a frequently visited web site can bring an unanware internet user to announcing themself to our tracker.
So we decided to insert truely random IP addresses from known-to-be-used sub nets into all our answers. We do know that this will degrade overall performance and will cost extra traffic and connections. But we are sure that this kind of deniability, when adopted by other trackers as well, will force copy right spies to acquire hard evidence against file sharers. Spread the word.
Topics: abuse, free speech | 26 Comments »
February 12th, 2007 at 4:07 pm
well, this could as well backfire: you will generate real used IP adresses, and random people will get threatened based on this — as the mafia can’t tell, if it’s real or fake. And as we can see every day, they don’t go for quality but quantity, so I don’t think they’ll care (as long as the real-to-fake ratio is not too extreme and they will get problems for threatening only truely random users. But you really can’t insert 1000 random adresses for every real one…)
And: the mafia has no problem with setting up their own *z-servers, so maybe they will just download like mad and thus find the real adresses no matter how many random ones you give them…
February 12th, 2007 at 4:34 pm
But that is exactly the point we want to make. You MUST NOT trust whatever a tracker tells you. There are many ways that may lead to incorrect IPs being reported. We just make sure there is at least one incorrect IP per reply so you can always prove that an answer from a bittorrent tracker not even is an indication let alone an evidence.
And: bittorrent trackers became evidence only, when black listing approaches were smart enough to stop the most naive implementations of detector clients. However, we are not keen on protecting file sharers at all costs. We simply do not like the idea of being missused in any way.
February 12th, 2007 at 6:20 pm
Is it possible to accept a special URL which assigns the client’s IP to a random torrent? This increases OpenTracker’s complexity but eleminates the need to supply an info hash.
February 12th, 2007 at 7:05 pm
Why would we want to do that? This would waste your tracker’s ressources on otherwise unused torrents and brings us nothing in terms of deniability.
February 12th, 2007 at 8:51 pm
[...] besonders toll wird das ganze dadurch, dass einige Trackerbetreiber daher jetzt anfangen, beliebige zufällige IPs mit in die IPs, die an die Clients gegeben [...]
February 13th, 2007 at 11:37 pm
This is so great. I am glad I found this site and the tracker! Ace!
March 16th, 2007 at 8:10 am
Hey, whats the deal with mininova banning denis.stalker.h3q.com tracker?
March 22nd, 2007 at 11:05 pm
Isn’t it easy for the “Copy right enforcers” to ping the p2p IPs? So i can’t imagine how that could give extra security to the participants.
March 23rd, 2007 at 4:13 am
That depends on how you define “easy” for them.
It would also be “easy” for them to actually do a real bittorrent client connection to the IP-addr. to verify their accusations against the user of the IP-addr. But as “easy” for the companies who send those threatening letters to ISPs mostly is defined by “making money” and not “getting real actual proof” even sending an ICMP packet may be too much.
We don’t say we give extra security to anybody sharing files they are not allowed to share. We just give the power of deniability to users of our bittorrent tracker to allow them to say they didn’t share those CC licensed pr0n-movies, the fact that our tracker says they are is no proof at all.
This power in the case of “Copyright Enforcement” brings the accusant in the position to collect real evidence for their allegations which stands in the contrary of making money the “easy” way.
March 23rd, 2007 at 4:14 am
To kyle,
mininova is not blocking our tracker.
March 27th, 2007 at 3:25 am
[...] . Some trackers are now adding random valid IPs among the trackers list of peer IPs. This may sound like a good idea at first blush, but if my ISP is getting spammed with DMCA notices [...]
April 4th, 2007 at 4:33 pm
How about giving out IP addresses from some of the known “bad guys” lists, so the RIAA & co get “evidence” for their own machines?
September 25th, 2007 at 7:39 pm
[quote]mininova is not blocking our tracker.[/quote]
I was under the impression that if you uploaded a torrent tracked by this tracker, mininova would ban your IP. At least that is what happened to me, but I think I used sumotracker and the pirate bay’s opentracker too, and I don’t think mininova bans them. I might have used another tracker too though…
Oh, and good idea with the perfect deniability thing
September 25th, 2007 at 8:31 pm
@13:
Why should they do that? We are even listed (twice!) on their “Upload torrent” page: http://www.mininova.org/upload
January 23rd, 2008 at 3:14 am
It seems to me that the BEST method for throwing these “Copy right enforcers” off the scent…. would be to…
Promote legal BitTorrent downloading of Public Domain material AS MUCH AS POSSIBLE.
This way, BitTorrent would become associated more with “legitimate” (legal in the USA) use… IP addresses of people using BitTorrent for uploading public domain material would DILUTE the lists of all IP addresses.
Is this logic flawed?
March 2nd, 2008 at 7:50 am
denis.stalker.h3q.com Got banned as on my site he is a well known BAD BAD person and possibly the net police
October 20th, 2008 at 10:01 pm
[...] Polluting the evidence works like this. When a client asks for a list of peers who are downloading the same torrent, the tracker software automatically inserts several “random IP addresses” that are not in the swarm. They are based on existing sub-nets, but might be from people who may not even be aware that BitTorrent exists. This means that the evidence that’s being gathered by anti-piracy companies includes IPs that belong to people that were not downloading the movie or album they are accused of. Perfect deniability, as the people who coded the tracker software explain. [...]
October 21st, 2008 at 4:41 am
[...] Polluting the evidence works like this. When a client asks for a list of peers who are downloading the same torrent, the tracker software automatically inserts several “random IP addresses” that are not in the swarm. They are based on existing sub-nets, but might be from people who may not even be aware that BitTorrent exists. This means that the evidence that’s being gathered by anti-piracy companies includes IPs that belong to people that were not downloading the movie or album they are accused of. Perfect deniability, as the people who coded the tracker software explain. [...]
October 21st, 2008 at 6:52 am
[...] Polluting the evidence works like this. When a client asks for a list of peers who are downloading the same torrent, the tracker software automatically inserts several “random IP addresses” that are not in the swarm. They are based on existing sub-nets, but might be from people who may not even be aware that BitTorrent exists. This means that the evidence that’s being gathered by anti-piracy companies includes IPs that belong to people that were not downloading the movie or album they are accused of. Perfect deniability, as the people who coded the tracker software explain. [...]
October 21st, 2008 at 9:36 am
[...] Polluting the evidence works like this. When a client asks for a list of peers who are downloading the same torrent, the tracker software automatically inserts several “random IP addresses” that are not in the swarm. They are based on existing sub-nets, but might be from people who may not even be aware that BitTorrent exists. This means that the evidence that’s being gathered by anti-piracy companies includes IPs that belong to people that were not downloading the movie or album they are accused of. Perfect deniability, as the people who coded the tracker software explain. [...]
October 21st, 2008 at 4:08 pm
[...] The process goes something like this: When a client asks for a list of peers who are downloading the same torrent, the tracker software inserts several ‘random IP addresses’ automatically. These are based on existing sub-nets, that might be from people who may not even be aware that services like BitTorrent exist, meaning that evidence being gathered by anti-piracy companies include IPs, belonging to people who aren’t downloading that particular movie or album they are being accused of. That is an example of perfect deniability. [...]
October 22nd, 2008 at 3:47 pm
[...] Polluting the evidence works like this. When a client asks for a list of peers who are downloading the same torrent, the tracker software automatically inserts several “random IP addresses” that are not in the swarm. They are based on existing sub-nets, but might be from people who may not even be aware that BitTorrent exists. This means that the evidence that’s being gathered by anti-piracy companies includes IPs that belong to people that were not downloading the movie or album they are accused of. Perfect deniability, as the people who coded the tracker software explain. [...]
October 23rd, 2008 at 10:20 pm
Is this feature just implemented on your personnel or in all instances of opentracker? If so is there anyway to enable/disable that feature?
December 3rd, 2008 at 12:28 am
[...] Polluting the evidence works like this. When a client asks for a list of peers who are downloading the same torrent, the tracker software automatically inserts several “random IP addresses” that are not in the swarm. They are based on existing sub-nets, but might be from people who may not even be aware that BitTorrent exists. This means that the evidence that’s being gathered by anti-piracy companies includes IPs that belong to people that were not downloading the movie or album they are accused of. Perfect deniability, as the people who coded the tracker software explain. [...]
March 6th, 2009 at 8:18 pm
[...] Polluting the evidence works like this. When a client asks for a list of peers who are downloading the same torrent, the tracker software automatically inserts several “random IP addresses” that are not in the swarm. They are based on existing sub-nets, but might be from people who may not even be aware that BitTorrent exists. This means that the evidence that’s being gathered by anti-piracy companies includes IPs that belong to people that were not downloading the movie or album they are accused of. Perfect deniability, as the people who coded the tracker software explain. [...]
July 1st, 2009 at 8:37 pm
I ran a tracker for a while and did a similar but different thing.
I used the Bluetack lists in a custom built filter and anyone connecting from an IP in those lists was placed into a separate swarm. It basically meant that they were all talking to themselves.
It probably wasn’t Big And Clever but it was kind of fun to do and did at least make me grin