[NNTP] draft-ietf-nntpext-authinfo-05 & draft-ietf-nntpext-tls-nntp-03

Russ Allbery rra at stanford.edu
Tue Oct 12 15:20:29 PDT 2004


Ken Murchison <ken at oceana.com> writes:
> Clive D.W. Feather wrote:

>> In authinfo-05, section 2.2, para 5 (of 7) I don't see the point of the
>> last sentence and in particular the SHOULD. See the similar comment on
>> the TLS document for details.

> This dovetails with the fact that the capabilties might change after
> auth/tls and the client should check for this.  Russ, care to comment? 
> I'm not married to this at this point, so I'll let the chair decide.

As I understand it, any time that the capabilities change, the client
needs to query them again in order for the extensions mechanism to work
the way that it works for other protocols.  Therefore, the only time that
a client should continue without sending LIST EXTENSIONS is if they intend
to not use any extensions whatsoever, and if we're advertising the NNTP
protocol version in LIST EXTENSIONS as well, I think it would be better to
just have clients always send it.  So the SHOULD seems right to me.

-- 
Russ Allbery (rra at stanford.edu)             <http://www.eyrie.org/~eagle/>



More information about the ietf-nntp mailing list