On Wed, 29 May 2002 [EMAIL PROTECTED] wrote:
> You're missing one from the list: SSL and LOGIN (or PLAIN). This likely would
> be acceptable. I believe the IESG could be convinced this is the way to go.
> Perhaps this would be more acceptable as a mandatory to implement: It certainly
> avoids all the authentication source issues.
>
> Still another option is to go with more than one mandatory to implement. You
> could say that all clients must do SSL+LOGIN and DIGEST-MD5 while servers can
> choose between the two. (I don't know if this helps, but I thought I should
> mention it.)

OK, this is helpful and may be the breakthrough that was needed.

How about the following:

A "plaintext authentication mechanism" is one of the following:
 . the PLAIN SASL mechanism
 . the legacy LOGIN command

All clients MUST implement all the following:
 . plaintext authentication mechanism after STARTTLS
 . plaintext authentication mechanism in SSL IMAP (port 993)
 . CRAM-MD5

Clients SHOULD implement the following:
 . DIGEST-MD5
 . Kerberos

All servers MUST implement at least one of the following:
 . plaintext authentication mechanism after STARTTLS
 . CRAM-MD5

Servers SHOULD implement the following:
 . plaintext authentication mechanism in SSL IMAP (port 993)
 . DIGEST-MD5
 . Kerberos

This matches current reality.

-- Mark --

http://staff.washington.edu/mrc
Science does not emerge from voting, party politics, or public debate.

Reply via email to