NNSquad - Network Neutrality Squad
NNSquad Home Page
[ NNSquad ] Re: Speculation, how AT&T can implement "copyright filtering" without wiretapping/dpi...
Executive Summary ? aside from the ?moral? issues, the real problem with all this meddling is that they lock us into naïve models of usage ? just as circuits limited us to very expensive isochronicity that made many of today?s standard applications too expensive be they YouTube or IM. Even today SMS continues to be priced at a level that prohibits many valuable uses because they are not valuable to the carriers. The delayed key approach makes a lot of sense for legitimate distribution and has a lot of precedent. In the current throttled net there is likely to be far more capacity available off-hours than during prime hours so it would make sense to distribute extra content at night. This works very well for satellite and broadcast distribution but also makes sense for taking advantage of available capacity ? you could download the entire TV lineup for the day or the week. Then, if you want to emulate the limitations of the TV grid, you can unlike a the 8PM shows at 8PM for viewing then with the option of charging if you want. There are obviously many variations of this scheme. Of course using P2P type multicast is one protocol one can use ? even if it?s not among users it still works for designated Akamai-like caching servers which can share the load and handle spot-congestion transparently. Once you understand this the encryption aspect is only an option ? it makes sense to distribute copyrighted material to the appropriate destinations while using the algorithm among your caching servers. How is an ATT going to know what the intent of the bits are ? after all they are copyrighted and marked and all that. This was an issue raised at NN2008 ? how do you determine what is a violation by looking at the bits in transit? One company that was pitched to me used P2P for legitimate distribution. I chose not to invest because they hadn?t characterized the problem correctly and thus P2P wasn?t the right solution but what?s important is that these protocols are gaining increasing acceptance among the very companies threatened by it! The other sleeping dog here is traffic shaping ? perhaps you can stretch the term to include Comcast?s actions. Another company I did invest in was distributing video over local campus networks ? but found that traffic shapers were being installed and thus their video distribution became problematic. What makes this especially frustrating is that managed gigabit routers are getting to $200 ? attempts to focus on scarcity instead are probably far more expensive and create scarcity. -----Original Message----- From: firstname.lastname@example.org [mailto:email@example.com] On Behalf Of Richard Clayton Sent: Tuesday, January 29, 2008 03:49 To: Nick Weaver Cc: firstname.lastname@example.org Subject: [ NNSquad ] Re: Speculation, how AT&T can implement "copyright filtering" without wiretapping/dpi... -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 In message <email@example.com> , Nick Weaver <firstname.lastname@example.org> writes >Thus all the schemes for "video DNA" and "audio DNA" tracking are >robust in the arms race, at least with a bit of effort. If the >transformations to avoid detection require inserting a huge amount of >noise and disruption, the MPAA and ISPs wins, because that shuts down >piracy just as sure as blocking the content. I agree up to a point -- and that point is an assumption about when the distributed files are decrypted. Suppose the files are distributed in an encrypted form, but the key to decrypt the files is not sent out until a week on Tuesday. Although the title of the file "latest film" may be suggestive of it being a copyright violation, it might just be called "r2d2.pt6" or "linux.v7.39" all of which are rather open to interpretation. Once the key is revealed (which is a very low bandwidth operation and impractical to block using any technology) then you will know if you have a copy of 'No Country for Old Men' or 'Star Wars VI' or a copy of the latest version of Linux. In fact, a little work with the encrypting engine (check out Waldman/Maziere's Tangler) and you can distribute all three at the same time (though it will be a bigger file) and later reveal the key to decrypt the one you want... ... although this delayed gratification may seem unrealistic for getting hold of movies on BitTorrent, the "release the key later" is seriously looked by researchers who are trying to find ways of preventing the Bad Guys from reverse engineering operating system patches faster than they can reach the end-users... ...and of course there's other "censorship resistant" publishing schemes in the literature, with various interesting properties in each case. So don't think as encryption as a trivial obfuscating technology, there's more to it than that. You're left with a large file being distributed and various nods and winks as to what it might contain... and indeed it might even be manipulable to contain more than one thing. In a world such as this, the lawyers will surely become ever richer as they argue as to whether the balance of the nodding and winking means that there is a prima facie case for blocking... >Accept it, identifying that a LARGE piece of content is a blatent >copyright violation is a straightforward technical problem, and robust >to the arms race with just a little effort. That's true under a certain set of assumptions, but not under others - -- Dr Richard Clayton <richard.clayton @ cl.cam.ac.uk> Computer Laboratory, University of Cambridge, CB3 0FD -----BEGIN PGP SIGNATURE----- Version: PGPsdk version 1.7.1 iQA/AwUBR57oipoAxkTY1oPiEQKMOwCeMzhTTqEk+YWr5QsmiW7G7ddm42YAoOa5 e/mvucnRVHvOc5ChofxImLPA =+mXX -----END PGP SIGNATURE-----