Re: [wg-c] Commission Working paper on the creation of .EU

[Eric Brunner <brunner@world.std.com>]

[Going over the daily limit, but stopping here for the day]

Joseph,

My query has to do with the authoritative source for the asserted policy
that two octet length TLDs in the IN class consisting of letters are
reserved, and to whom.

In the early versions of draft-ietf-dnsind-iana-dns-0?.txt we (Eastlake,
Manning and Brunner) wrote:

version -00

   All two octet length TLDs in the IN class consisting of letters are
   reserved for assignment to territories.  Those (1) allocated by [ISO
   3166] and (2) allocated by the Universal Postal Union [UPU] and
   reserved in [ISO 3166] even though not formally assigned by [ISO
   3166] (e.g., a few British Channel Islands), are assigned as so
   allocated by the generally recognized acting government of the area
   associated with the "country code" or on a first come first served
   basis to a designated registry if there is no such government or the
   government has not exercised control.
   ...
   Country codes consisting of a letter and a digit or two digits are
   not currently used by [ISO 3166] or the [UPU].  However, to permit
   possible expansion of the two octet country codes, they are reserved
   for future allocation as described in the previous paragraph.

version -01

   Two octet length ASCII label TLDs in the Internet CLASS consisting of
   letters are for assignment to geo-political territories.  Those (1)
   allocated by [ISO 3166] and (2) allocated by the Universal Postal
   Union [UPU] and reserved in [ISO 3166] even though not formally
   assigned by [ISO 3166], are assigned as so allocated.  Two letter
   codes reserved by [ISO 3166] for local use or the like are also
   reserved as TLDs as are two letter TLDs not yet allocated or reserved
   by [ISO 3166] or the [UPU].  A generally recognized acting government
   of the territory associated with a "country code" has priority to act
   as or designate the registrar for such TLDs. 

By version -03 we'd removed all reference to 3166. Now as the authors of
the proposed authoritative text on the Domain Name System (DNS) IANA
considerations, relative to the two-octet ASCII labels, I'm really very
curious where the assertion you made finds its authoritative reference.

Cheers,
Eric