[liberationtech] New protocol sacrifices bandwidth for metadata privacy

[Marc W. Abel]

On 08/12/2014 05:19 AM, Aymeric Vitte wrote:
> I don't know what is the state of the art regarding this but is there
> not a way while parsing the slots to determine that this one might have
> your key so you don't have to decrypt everything and just decrypt the
> slots that might have it? (I have a similar issue with Peersm to check
> the pieces)


Hi Aymeric,

By design, the answer is no.  I'd either be leaking key material 
(directly or indirectly), or leaking metadata.  Although to date, there 
have been no user reports of CPU constraints here.  Or any other user 
reports....

With that said, it would be great to have a Better Way.  I haven't found 
one yet.

Of course if you can work out offline what IP addresses the peers which 
may interest you are at, you can skip checking a lot of packets. 
Although the software has yet to implement a way to exploit this knowledge.

Marc