[Mailman-Developers] suggested improvement for Mailman's bounce processing

[Ian Eiloart]

--On 9 August 2006 12:18:13 -0500 Brad Knowles <brad at stop.mail-abuse.org> 
wrote:

>
> Otherwise, RFC-1893 would have been sufficient to answer all possible
> questions about this feature, and all MTA authors and all mail systems
> administrators would have been able to perfectly follow those guidelines.
> We wouldn't have needed RFC 3463, or the updates from RFCs 3886, 4468,
> etc....
>


Neither RFC makes any significant changes here.

"Appendix B - Changes from RFC1893

   Changed Authors contact information.

   Updated required standards boilerplate.

   Edited the text to make it spell-checker and grammar checker
   compliant.

   Modified the text describing the persistent transient failure to more
   closely reflect current practice and understanding.

   Eliminated the restriction on the X.4.7 codes limiting them to
   persistent transient errors."

RFC 4468 adds two new codes.

> So, show me a parser that fully understands all possible correct
> interpretations of these RFCs, plus all possible incorrect but likely
> interpretations of these RFCs, and we might have something useful to talk
> about.
>


It's not necessary to understand all interpretations. There are a few codes 
that mean the remote address isn't available. When we see any other code, 
we should not count the bounce against the specific address, because the 
error isn't related to that address.

-- 
Ian Eiloart
IT Services, University of Sussex