[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [wg-c] Who should vote for new gTLDs



On Sun, Jul 25, 1999 at 10:33:14AM -0400, Cade,Marilyn S - LGA wrote:
> 
> Voting is an interesting idea... I've been following the discussion, and
> want to raise a question.  How will the DNSO take into account and consult
> with the other two SO's.  

I think, and as you indeed hint below, the consultation must be 
wider than that.

> Voting will bring in those who are involved, but not necessarily those who
> are affected, or knowledgeable... Maybe instead of voting (and this may be
> what you mean), a poll which can advise those who are representing various
> constituencies could be a first step...and then they could consult about the
> pros and cons, impact on various other stakeholders, etc.

I have been reviewing Eric Brunner and Michael Froomkin's suggestions
for new TLDs.  Neither is a gTLD, but the mere existence of such
proposals, I believe, provides important guidance for our work here:
there are other TLDs than gTLDs that must be considered for inclusion
in the root zone, but there is only one root zone, and there is only
one name space in it.  Every use of a name preempts some other use of
the name.  Every gTLD name preempts the name of a potential
"chartered" TLD. 

Some have suggested that the DNSO should set standards for
"accrediting registries" and then let the accredited registries pick
their TLD names (restricted by some rules).  I believe that approach
is exactly backward.  We should first have a process for the
selection of *names* in the TLD space (not just the gTLD space), and
then, perhaps, let accredited registries bid on the names. 

In particular, every TLD needs to be up for an extensive public
comment period, and there needs to be a process for evaluating those
comments.  It can't just be a vote of the GA, or the SOs.

The public comment period has to address not only the name, but the
intended use.  A "privacy-enhanced" TLD has policy implications, as
Prof Froomkin indicates, and those policy implications are, in that
case, more important than the name.  (In contrast, for a ".att" TLD,
the policy implications derive from the specific name, and the 
controversy surrounding the choice of name would dwarf any policy 
considerations, I would imagine. :-))

That is, the name and the intended use for a TLD must be considered
together.  The choice of registry is a secondary concern that may or 
may not be important.  [A while back a proposal was floated for a 
".eu" TLD as a pan-european pseudo ccTLD.  I don't remember the 
details, but presuming that it was proposed by by the European 
Commission, the choice of registry might well be constrained by the 
policy for the TLD.)

The mechanism I have proposed for incorporating this public comment
is the Working Group process.  It has been suggested that this would
slow down the addition of new TLDs -- many see that as a significant
advantage, in fact.  But it is important to realize that a WG could
propose a set of TLDs.  For example, someone might propose a set of
numeric TLDS, ".000" to ".999", to be managed as gTLDs.  There are a
number of things to consider about such a proposal -- are there
technical implications to numeric TLDs? should they all be run by one
registry?  Is it a good idea to add 1000 TLDs in a single block?  
What if some enterprising entrepreneur wants ".00000" to ".99999"?

So to return to Javier's original question -- I don't think TLDs
should be approved just by a vote.  *Every* TLD has a policy
associated with it, explicit or implicit, and that policy cannot, in
general, be disassociated from the name.  That polic/name combination
is precisely the first thing that must be approved, before the 
registry even enters into the picture.

-- 
Kent Crispin                               "Do good, and you'll be
kent@songbird.com                           lonesome." -- Mark Twain