protocol 2 performance gain?

[mouring at etoh.eviladmin.org]

On Tue, 6 Mar 2001, Higdon, David M - CNF wrote:

> has anyone noticed a performance gain with protocol 2?
> 

I assume your refering to the current snapshot. 

ChangeLog:

[..]
   - deraadt at cvs.openbsd.org 2001/03/05 15:56:16
     [myproposal.h ssh.1]
     switch to aes128-cbc/hmac-md5 by default in SSH2 -- faster;
     provos & markus ok
[..]
   - markus at cvs.openbsd.org 2001/03/05 17:17:21
     [kex.c kex.h sshconnect2.c sshd.c]
     generate a 2*need size (~300 instead of 1024/2048) random private
     exponent during the DH key agreement. according to Niels (the great
     german advisor) this is safe since /etc/primes contains strong
     primes only.

     References:
             P. C. van Oorschot and M. J. Wiener, On Diffie-Hellman key
             agreement with short exponents, In Advances in Cryptology
             - EUROCRYPT'96, LNCS 1070, Springer-Verlag, 1996, pp.332-343.


- Ben