On Sat, Jan 17, 2015 at 3:26 AM, Tony Arcieri <[email protected]> wrote: > https://en.wikipedia.org/wiki/Dining_cryptographers_problem > > It seems like it's possible for modern anonymous messaging protocols to have > formally provable anonymity guarantees by virtue of secure multiparty > computation: > > http://dedis.cs.yale.edu/dissent/papers/analysis.pdf
Better overview: http://bford.info/pub/net/panopticon-cacm.pdf "DC-nets" are smallish groups where everyone broadcasts constantly to the other group members. Messages are XOR'd somehow so group members can read the sent messages, but can't tell who sent them. Maybe that's more analyzable and has some nicer properties than onion-routing or mix-nets. But it's also making strong and very limiting assumptions (group members only communicate with each other; small anonymity sets; entire group is online and constantly broadcasting). Trevor _______________________________________________ Messaging mailing list [email protected] https://moderncrypto.org/mailman/listinfo/messaging
