I thought bloom filters + cryptography == group crypto.
(not a cryptographer, deep trawling memory of articles read many years ago)
On Thu, Jan 14, 2016 at 4:23 PM, Phillip Hallam-Baker
<phill(_at_)hallambaker(_dot_)com
wrote:
On Wed, Jan 13, 2016 at 9:43 PM, John Levine <johnl(_at_)taugh(_dot_)com>
wrote:
Well the question I responded to was wondering if there was a way to
have the list software re-encrypt to the members without being able to
decrypt the message first. If I understand correctly, you're saying that
S/MIME can't do that either?
Not unless there is something hidden inside all of the X.509 glop that
I'm unaware of. Message bodies are encrypted with a session key
which is in turn encrypted with the public keys of the message
recipients. Nothing magic there.
You end up having to modify the clients that decrypt the messages.
The sender merely needs to use an encryption scheme that is based on
an algorithm that there is a recryption scheme for. I have only seen
Diffie Hellman based schemes. But since the receiver code is likely to
be for the new CFRG ECDH algorithms, you are backed up on the sender
and receiver side.
There are probably ways you could shoehorn this into S/MIME and
OpenPGP but why not do the job right and extend wherever necessary?
the client code has to change anyway.