>It therefore seems to me to be not a bad idea to have an RFC or IANA >registry for the "reserved names", in ICANN parlance. It would also >be good if some operational rules about what "reserved names" means >were in an RFC somewhere (for instance, are there different classes >of "reserved" names? Why? Yes, that's what I'm trying to get at. As far as I can tell, there are at least three different kinds of reserved names. Names like .INVALID and .TEST are reserved forever and will never be delegated. Names like .IETF and .RIPE are names of identified Internet infrastructure organizations who could in principle have the name delegated to them. Names like .QQ are reserved until some third party (ISO 3166 in this case) says something about them. There are probably also names that are implicitly reserved like .123 or .XX--ABC (where the two letters in XX-- are anything other than xn) because of likely technical problems if they were used. And there appear to be names like single characters which are reserved until ICANN decides what to do about them. There's also one special purpose TLD, .ARPA, which is more or less delegated to the IAB although managed by IANA. For second-level domains, there are the three EXAMPLE names reserved by RFC2606, a fair number of names reserved by ICANN policy (single letters in gTLDs other than the few pre-ICANN ones), or by terms in gTLD contracts. Other than the EXAMPLEs I don't think any of those reservations are expected to be permanent. This is all reasonably well known to people familiar with ICANN folklore, but it sure would be nice to have it written down in one place so we can have an idea of what they are. >I'm actually indifferent to the publication means for this, though -- >an RFC, an ICANN document that stands on its own, whatever -- but >since we have IANA registries already that seems to me to be the >obvious locus for yet another list. IANA seems the most reasonable place to me, too. The trick is getting ICANN to agree to maintain its contents or at least contribute their entries to it. There's the somewhat separate issue of _foo tags used in SRV and other non-host records. That seems politically and technically straightforward, give or take liason with the group enumerating service and protocol names for SRV. R's, John _______________________________________________ Ietf mailing list Ietf@xxxxxxxx https://www.ietf.org/mailman/listinfo/ietf