Viktor Dukhovni:
> On Tue, Sep 02, 2014 at 01:21:47PM -0400, Wietse Venema wrote:
>
> > Viktor Dukhovni:
> > >
> > > In the RFC editor queue as of 2014-08-29:
> > >
> > > https://datatracker.ietf.org/doc/draft-ietf-appsawg-nullmx/
> >
> > Hmm. This text says:
> >
> > When a submission or SMTP relay server rejects an envelope recipient
> > due to a domain's null MX record, it SHOULD use a 521 reply
> > code [...]
> >
> > Both 221 and 421 are sent at the end of an SMTP session. Postfix
> > uses 521 for the same purpose: is is like 421 except that the client
> > should not retry.
> >
> > Maybe the draft can be modified.
>
> It is rather late in the process, with the draft in the RFC Editor's
> queue the only chance to modify it is I think AUTH48, and the nature
> of the allowed changes is limited.
>
> Yes, the recipient and sender basic response codes are backwards,
> this is a bug. It should/could be 521 for rejecting a nullmx sender
> domain, but definitely not for a nullmx recipient domain.
Reply codes 221 and 421 terminate an SMTP session. For consistency,
reply code 521 ought to terminate a session as well.
Also, there is mangled text in section 4.1:
[...] the address was mistranscribed or misunderstood, for
example, to [...] or al...@examp1e.com rather than al...@example.com.
This is draft -08. Is that followed by AUTH48?
Wietse
