There’s this paper by this guy at KCL.
That he’s posted it on Dropbox is both relevant and ironic.
In it, and in his Twitter feed, he argues essentially that Tor is OK-ish, but promotes anonymity – which he sees as “bad” – and Tor Hidden Services are intolerable and should “no longer be developed” because they promote so many bad things.
There are a bunch of arguments one could have about morality, privacy, anonymity, etc; but that’s playing the game in the expected fashion, leading to much postmodern posing and wastage of breath; so I will try a different, more Turingesque machine-based approach.
It’s very simple:
Strategically there is no communications difference between Tor, and Tor Hidden Services; what do I mean by this? I mean that both are simply forms of communication, and all forms of communication are functionally interchangable. To explain:
Tor mirrors the Internet and provides a connected graph of nodes which can communicate peer-to-peer; Tor Hidden Services provide a client-server model akin to the Web which runs atop the Internet.
If we are talking about access to data at rest – then we can provide such access in both models; with peer-to-peer networks we use Content-Based Addressing (a-la “Magnet Links” on Bittorrent) and on client-server networks we use Resource-Based Addressing (a-la URLs on the Web)
If we are talking about access to data in motion – then we can also provide such access in both models; with peer-to-peer communications (Skype, Bittorrent, E-Mail, USENET) – which may be synchronous (VoIP) or not (store-and-forward); and on client-server networks we historically just emulate the endpoints of peer-to-peer communication (E-mail becomes IMAP).
If data is not at rest or in motion, what is it?
So: there are two sorts of data and two communications mechanisms which are equivalent, merely using alternate addressing strategies* to distinguish them; with this understanding there is no way to choose one over another, nor reject one as “bad” while the other is “ok” or “good”.
Therefore, when one is dismissing a communications mechanism as bad, one is not talking about the medium, because all communications media are technically equivalent.
Instead, one is talking about the message. Therefore one is talking about censorship.
Welcome to your new role, Clement. Censor. QED.
Also, Dropbox, really? That’s not a proper webserver at all. If anything, it’s a peer-to-peer network with hierarchical backing storage and distributed web-emulating frontends.
* Another example:
- Resource based addressing: “third shelf, fourth book along”
- Content based addressing: “says it’s authored by Dickens, begins with ‘It was the best of times, it was the worst of times…'”
Do you know what it is, yet?