I prefer option 2. I think it fulfills the implicit obligations inherent in 1) -which would be to fill the hole of uncertainty. Its succinct, and it covers the cases I think define the condition.
I would ask if 2) also needs to define ".. or cannot be resolved" because "[or not at all]" is a bit wide to encompass "that FQDN didn't actually resolve" for a listed NS. As it stands it seems to terminate in a belief you have something to ask. It's equally lame to be pointed to this.name.hasnt.been.delegated.alt. as an NS Irrespective I prefer 2). I don't like the dangling fate of 1) I think the lack of agreement is only conditionally true: I think we could reach consensus on 2) if we tried a bit harder. -G _______________________________________________ DNSOP mailing list DNSOP@ietf.org https://www.ietf.org/mailman/listinfo/dnsop