Lee,
I don't see any discussion in your draft about why rDNS is needed in
this space. IME there are typically 2 uses cases:
1. "Residential" users, or more specifically, those who will not
be/should not be running services on their addresses
2. "Commercial" users, who may be running things like mails servers
Obviously (and yes, I am being facetious since there is still
controversy on this point in some quarters) the latter need the ability
to run proper rDNS so that at minimum their mail servers can pass
industry-standard (and arguably best practice) forward/reverse
verification. Hopefully the solutions for that are obvious to the
participants on this list.
However, for users that are not running services the primary desire
(that I'm aware of) is to have an easy way to flag that those are ranges
that we would not expect mail to come from. The guidelines for flagging
"dynamic" addresses in the mailop community are well known for IPv4, but
I don't see any mention of that in your draft (although admittedly, I
only skimmed it).
So while on the one hand documenting that there are options for actually
providing valid rDNS for end users probably has value (the "How") I
don't see enough discussion about "Why?" to give me a good feeling about
this draft. Particularly I would like to see some discussion about why a
wildcard set at a high level for your "dynamic" range isn't a valid
solution.
Doug
_______________________________________________
DNSOP mailing list
DNSOP@ietf.org
https://www.ietf.org/mailman/listinfo/dnsop
