> I see Thomas' argument for tolerating occasional use of AAAA entries in the > global DNS for ULAs - but it seems that it leads to too many complications > to be recommended. Since I'm sure the IETF isn't ready yet to endorse the > reality of split DNS deployment, wouldn't it be best to say that ULA-Cs > SHOULD NOT be included in the global DNS? (And that is a significant > difference in scope and intent compared with PI.)
Note: a key question is whether the RIRs (as they hand out ULA-C blocks) allow PTRs to be inserted in the DNS. They either do or they don't. Saying folk "SHOULD NOT" doesn't answer a key question definitively. Note: we have no say/control over whether AAAA records contain ULAs. End sites can (and will) do what they want here; we have no way of preventing that (though we can recommend they don't and explain why it may not be a good idea). Thomas -------------------------------------------------------------------- IETF IPv6 working group mailing list [email protected] Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
