[policy] When Tech Meets Policy...

Mark Andrews Mark_Andrews at isc.org
Wed Aug 15 01:58:52 UTC 2007

>This comment was added as a follow-on note.  Sorry for not being clear.
>Accepting messages from a domain lacking MX records might be risky  
>due to the high rate of domain turnovers.  Within a few weeks, more  
>than the number of existing domains will have been added and deleted  
>by then.  Spammers take advantage of this flux.  Unfortunately SMTP  
>server discovery via A records is permitted and should be  

	All it would require is a couple of large ISP's to adopt
	such a policy.  "MX 0 <self>" really is not hard and benefits
	the remote caches.

>Once MX records are adopted as an _acceptance_  
>requisite, domains not intended to receive or send email would be  
>clearly denoted by the absence of MX records.  SMTP policy published  
>adjacent to MX records also eliminates a need for email policy  
>"discovery" as well.  Another looming problem.

	Better yet us MX records to signal that you don't want to
	receive email e.g. "MX 0 .".  It has a additional benefits
	in that it is *much* smaller to cache than a negative
	response.  It's also smaller to cache than a A record.

	Since all valid email domains are required to have a working
	postmaster you can safely drop any email from such domains.

>Don't accept a message from a domain without MX records.  When there  
>is no policy record adjacent to the MX record, there is no policy,  
>and don't go looking.

More information about the NANOG mailing list