[TLS] new version of draft-ietf-tls-pwd

["Dan Harkins" <[e-mail]>]

  Hello,

  I just uploaded draft-iet-tls-pwd-03 that incorporates resolution
to comments received from WGLC. These include:

  - test vectors of an example exchange (added to -02).
  - address the issue Bodo Moeller described that allowed a server
     to impose a weak group on the client.
  - inclusion of a technique to protect the username based on
     a comment by Trevor Perrin.

The scheme for username protection has a number of benefits:

   - it does not increase the number of messages
   - it does not require storage of any server secrets on clients
   - is resistant to passive monitoring attack and active attack
      by a malicious server (no I will not be providing a security
      proof to back up those claims)
   - protected username will be different with each run of the
      protocol
   - minimal overhead-- 32+16 = 48 bytes leaving enough
      for insanely long usernames

and the drawbacks are that it opens up a potential for DoS
attacks and it also does not have any forward secrecy-- server
compromise can result in the disclosure of the names of clients
from previous runs of the protocol.

  regards,

  Dan.