Re: des-cbc cipher

[RJ Atkinson <rja%inet.org@localhost>]

At 10:29 29/11/01, Jeffrey Altman wrote:
>The way this should be handled in the IETF is to publish the
>deprecated functionality as a separate Historical RFC.  

        The IETF has not traditionally required that.  It would
be one approach.  It would not appear to solve all issues (more
below) in any event.

        Instead, a common IETF approach is to document the deprecated 
capability in an Appendix of the regular documents, clearly marked as 
deprecated or informational (I don't object to the labelling either way), 
so that an implementer trying can easily locate it and then make an 
informed decision about what to do in her/his implementation.

        In any event, it appears (I hope I'm confused and this can
be clarified) that some folks want to make sure the protocol element 
changes -- so that the existing deployed implementations won't be 
interoperable or compliant with the SSHv2 standard (regardless of
how DES-CBC were documented or even not documented).

        Sigh.

Ran