Mark Crispin wrote:

> On Wed, 29 May 2002, Lawrence Greenfield wrote:
> > Our local site policy doesn't offer DIGEST-MD5---but
> > that isn't what we're talking about.
>
> The point seems to be interoperability between compliant implementations.
> A client which only implements DIGEST-MD5 is not able to talk to your
> server.
>
> I think that we can require SSL/TLS + plaintext and for the most part
> reflect the world that exists today.
>
> I don't think that we can require DIGEST-MD5 and reflect the world that
> exists today; nor do I think that an implementor would be well-served by a
> document that implies that a viable product can be produced that only
> implements DIGEST-MD5.
>
> This argument would actually suggest against removing the requirement for
> CRAM-MD5 in my proposal and going only with SSL/TLS + plaintext.  Perhaps
> that's the best thing to do.

I would rather not do that, as it is not always trivial to add TLS support to
a client/server. I like CRAM-MD5 as MUST (for interoperability), DIGEST-MD5 as
SHOULD and cleartext + TLS as MUST.
So basically this is your original proposal without Kerberos and Port 993.

Alexey
__________________________________________
R & D, ACI Worldwide/MessagingDirect
Richmond, Surrey, UK
Phone: +44 20 8332 4508
Home Page: http://orthanc.ab.ca/mel

I speak for myself only, not for my employer.
__________________________________________


Reply via email to