Proposal for the charter

John C Klensin klensin at
Wed Mar 12 23:30:10 CET 2008

--On Wednesday, 12 March, 2008 19:26 +0100 "Marcos Sanz/Denic"
<sanz at> wrote:

> All,
> Repeating here what I said at the BOF regarding the following
> sentence in  the charter:
> "Separate requirements for valid IDNs at registration time,
> vs. at  resolution time".
> My suggestion is to change the wording because it is too
> "registration  centric": consider that the act of registration
> generally only occurs at  the second or third level, towards a
> domain name registry. However, we  want a solution applicable
> to all levels of the DNS tree, where  delegations naturally
> occur without any registration. For the matter at  stake, I
> like the wording from RFC 3454, which makes a distinction
> between  "stored strings" (which encompasses registered
> domains, but then also  other kinds of storage like e.g.
> entries in a certificate) and "queries"  (volatile existences
> to be matched against stored identifiers). The  proposed
> substitution could be
> "To take into account the different nature of 'stored strings'
> vs  'queries' (cf RFC 3454) and leverage on them to pursue
> interoperability  with IDNA2003"


I don't really care about this change one way or the other,
i.e., it there is consensus to make this change, it is fine with
me.    However, please note that the term "registry" is
introduced in RFC 1591 as applying to any entity that maintains
a zone file.  Also, FWIW, the rationale document reinforces and
strengthens that definition and explicitly notes (as do some
base DNS documents) that "resolution" and "lookup" are
equivalent.  So I don't think this change is substantive.

Just my personal opinion, of course.


More information about the Idna-update mailing list