I see two cases: There are codes from the unified DNS error code space that are not visible at the top level of a DNS message. For example, the Error field of TSIG (RFC 2845) or TKEY (RFC 2930). For things like that I have no problem with Specification Required or Expert Review.
However, for top level RCODEs that appear in the DNS message header (possibly extended by an OPT RR), I'm not so sure. What is a resolver supposed to do when it gets such a top level RCODE it does not understand? You really don't want to deploy new top level RCODEs except in a context where you have a strong assurance that the receiver will understand it. I'm more comfortable with IETF review for this sort of new RCODE value. Thanks, Donald ============================= Donald E. Eastlake 3rd +1-508-333-2270 (cell) 155 Beaver Street, Milford, MA 01757 USA d3e...@gmail.com On Tue, Jul 14, 2015 at 9:26 PM, Mark Andrews <ma...@isc.org> wrote: > > At the moment there are "private use" and "IETF Review" > as the two levels. > > While doing dns-cookies it has became clear that "specification > required" and/or "Expert Review" may be more appropriate > than "IETF Review" in general. Leave "IETF Review" for the > additions to the first 16 which should only be used in cases > where EDNS is not available. > > Mark > -- > Mark Andrews, ISC > 1 Seymour St., Dundas Valley, NSW 2117, Australia > PHONE: +61 2 9871 4742 INTERNET: ma...@isc.org > > _______________________________________________ > DNSOP mailing list > DNSOP@ietf.org > https://www.ietf.org/mailman/listinfo/dnsop _______________________________________________ DNSOP mailing list DNSOP@ietf.org https://www.ietf.org/mailman/listinfo/dnsop
