OSEC

Neohapsis is currently accepting applications for employment. For more information, please visit our website www.neohapsis.com or email hr@neohapsis.com
 
From: Alex Alten (Altenhome.com)
Date: Mon Jan 08 2001 - 02:27:10 CST

  • Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

    At 11:58 AM 1/6/2001 -0800, Eric Rescorla wrote:
    >> Eric Rescorla wrote:
    >>
    >> > Both browsers and servers ALREADY support anonymous connections.
    >>
    >> There is a huge difference between SUPPORT and working out
    >> of the box. In my world, httpd would work out of the box if
    >> the server started up and created a self signed certificate
    >> straight away and ran without any more intervention. Just
    >> like sshd.
    >As I said earlier, SSL operates under a different threat model
    >than SSH. SSH connections are by and large to servers you've
    >already contacted before. SSL needs to support spontaneous
    >connections to many hosts which you've never seen before and have
    >absolutely no prior relationship with.
    >

    Ian has a very valid point. The MitM threat model that SSL is trying
    to prevent was overkill for the vast majority of web users. A decent
    privacy-only design with say replay protection is all that's needed in
    general. Then you get rid of the lengthy handshake, the high cpu
    overhead of RSA and the cumbersome certificate management (for example,
    remember a year ago when many older browsers had their RSA certs expire,
    it was worse than Y2K).

    The only counter point I can think of is that you want to cover all
    types of transactions, including very high value ones. But, the
    default behavior of SSL (with server-only authentication) is no good
    anyway for those types of transactions.

    --
    

    Alex Alten

    AltenHome.Com