Re: Last Call: draft-klensin-rfc2821bis

	There was a call for me to explain this statement.
	
> Mark Andrews wrote:
> > 
> >     Also getting rid of implict MX records would "deal" with all
> >     those ISP's that insist that they need to re-write NXDOMAIN
> >     responses.

        Todays there are ISP's that replace NXDOMAIN with a A record
        pointing to a web server which has a search page.  This is
        often turned on for everyone with a opt-out which doesn't
        help the masses as they don't know that this is bad.

	The ISP usually claims it is being "helpful".

	Any MTA using that caching resolver, directly or indirectly,
	get A records rather than NXDOMAIN on typos, hosts that
	have gone away etc.

        The implict MX rule then causes all those typos to be valid
        email destinations.  Which either queue up for a week or
        are intercepted and hopefully bounced without being read.

        An explict MX rule would miminize the damage caused by doing
        this re-writing, as would publishing a "MX 0 ." record, if
        we can get that standardised.  Either change would allow a
        MUA to validate the RHS by performing just a MX lookup
        rather the MX and A.  NODATA is as useful as NXDOMAIN with
        a explict MX rule.

        Mark
-- 
Mark Andrews, ISC
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742                 INTERNET: Mark_Andrews@isc.org

-- 
Mark Andrews, ISC
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE:	+61 2 9871 4742		         INTERNET: Mark_Andrews@isc.org
_______________________________________________
IETF mailing list
IETF@ietf.org
https://www.ietf.org/mailman/listinfo/ietf