Re: [xmpp] Self Introduction

[Jon Kristensen <[e-mail]>]

Hi Richard, and thank you for your question!

I may be wrong - I only skimmed through it, but the way I understand 
that specification, it's using public-key cryptography is used to 
encrypt and decrypt the messages. If this is the case, it would mean 
that the protocol does not offer repudiability; if Alice sends a message 
to Bob, Bob can prove that Alice (or, at least, someone with access to 
Alice's private key) has authored a message. In most off-the-record 
conversations, this would not be desirable. What OTR does is negotiates 
a shared secret (symmetric key), which is then used to encrypt and 
decrypt the messages. What this means is that Bob, having access to the 
encryption key, could have just as easily produced the message. Hence, 
Alice can deny that she authored the message in the first place.

Furthermore, the negotiation of a shared secret is continuously repeated 
in OTR in order to allow for forward secrecy. This prevents previous 
session keys from being compromised if the (long-term) private key 
happens to get compromised. I think this is another important security 
feature to keep in mind.

Best,
Jon Kristensen

On 02/27/2013 07:30 PM, Richard Barnes wrote:
> Any comments on draft-miller-xmpp-e2e?
> <http://tools.ietf.org/html/draft-miller-xmpp-e2e>