On 26.01.2018 19:31, carlo von lynX wrote: > On Fri, Jan 26, 2018 at 06:46:14PM +0100, t3sserakt wrote: >> By the way, is it a stupid idea to let the endpoint of >> a cadet path be not the endpoint of user communication, >> to protect meta data? Maybe this is easier to accomplish >> than onion routing, or an additional protection together >> with OR. > That is actually the simple way to implement OR, > just make CADET connections between the relays > and therefore obfuscate the final endpoints. > The reason we don't do that yet is because we > don't have a strategy to decide which relays are > trustworthy - or maybe, if rps is functional, we > now have one. What is the difference - regarding trustworthiness - between nodes on the path between endpoints of user communication, and the endpoints of a cadet path? "Only" the number of nodes being involved is getting higher, and therefore the number of nodes "knowing" about the path. The chance of having an adversary node being part of the path rises, but the knowledge which node is the actual receiver of the cadet message is harder to get for an adversary node in the path. I think the answer to the question if this is making the meta protection better is: it depends. :-)
Maybe someone knowing cadet better than us can enlighten us. ;-)
signature.asc
Description: OpenPGP digital signature
_______________________________________________ GNUnet-developers mailing list [email protected] https://lists.gnu.org/mailman/listinfo/gnunet-developers
