Re: [ga] gTLD Constituency Memo

[DPF <david@farrar.com>]

On Wed, 25 Jul 2001 11:19:21 -0400, "Gomes, Chuck"
<cgomes@verisign.com> wrote:

>The attached document is a memo from the gTLD Constituency that contains a
>list of questions that we believe need to be asked regarding the possible
>establishment of an ICANN ccTLD Supporting Organization.

Thanks for starting some discussion off.  My 2c on your questions are
below.  Please note that while I hold office within a ccTLD I do not
speak for them.

>I.  Contractual questions
>
>A.	Will it be a condition of creation of a new SO that the ccTLDs
>be required to enter into a contract with ICANN? 

I think this would put the cart before the horse.  Having a ccSO would
allow individuals ccTLDs to then enter into contracts with ICANN but
without this none seem too keen at present.

I imagine the ccSO will have a contract with ICANN but that will be
different to individual ccTLDs having one.

>Will that contract
>require compliance with consensus policies?  If not, especially with
>regard to open ccTLDs, would that constitute inequitable treatment in
>violation of ByLaw and contract requirements?

Any attempt to impose policies across 250 ccTLDs which operate under
250 sets of laws will IMO be doomed to fail.  Unless such policies
were of totally bland generality (The ccTLD shall have a dispute
resolution policy) it just won't work.

Splitting ccTLDs into "open"and "closed"is problematic also.
Certainly some ccTLDs such as dot tv compete to a fair degree with
gTLDs are marketed as such.  

However others like dot nz do not promote themselves at all, do not
pretend dot nz stands for anything other than New Zealand, yet do
allow anyone in the world to register purely as an efficiency measure.
It just costs too much to manually vet each registration - we want
quick, cheap and efficient.

dot nz would IMO be very resistant to being classified as an open TLD
subject to ICANN consensus policies rather than its own local
consensus policies.

>B.	What should be the relationship between a ccTLD-ICANN contract
>on re-delegation to a general ICANN policy on presumptive right to
>renewal for uTLDs (unsponsored TLDs) or sTLDs (sponsored TLDs)?

I do not believe limited terms for ccTLDs is useful.  Under RFC1591 I
submit that as long as a ccTLD is keeping the local internet community
(which includes the local Govt) happy there is no need to have a fixed
term contract and hence presumptive rights of renewals.

Many ccTLDs are non profit, which is different from some gTLD
registries who are commercial businesses seeking to make money (not
saying this is bad) from their registry.

>II.  Reasons to have SOs
>
>A.	If the rationale for creation of a ccTLD SO is (at least in
>part) based on ensuring the payment of fees to ICANN, would that
>rationale also support creation of other SOs more directly reflecting
>the interests and views of others who also pay a high portion of ICANN
>fees?

I do not believe the rationale is purely paying of fees.  It is also
to do with the nature of the relationship with ICANN Board and staff,
with fulfilling ICANN's mission from DOC to have wide international
support, and also the growing realisation that gTLD policy making has
little in common with needs of ccTLDs.

If one creates SOs purely on who has the money, then ICANN is not
representative of the Internet Community.  

>B.	If the rationale for a ccTLD SO involves assuring adequate
>representation on the ICANN Board of those whose interests may diverge
>and who are bound by contracts with ICANN, would that support creation
>of numerous additional SOs to reflect the distinct interests of gTLDs,
>sTLDs, registrars, registrants, etc.?=20

The ccTLDs are a very well defined community whose sphere of interests
have IMO little in common with those involved with gTLDs.

gTLD registries, Registrars, registrants all have inter-linked
interests in the DNSO.

Finally it should be remembered that some/many ccTLDs are not just
registries but include representation from registrants and registrars
from their local internet community.  Each ccTLD to some degree could
be viewed as a mini DNSO and the ccSO is the collection of all of them
coming together.

>C.	Are sTLDs sufficiently differently situated from uTLDs, with
>respect to their contractual relationships with ICANN and the role of
>Sponsoring Organizations in the development of policies, that they
>should have their own SO?=20

Possibly yes.  This is an issue worth considering.

>III.  Creation of policies
>
>A.	If ccTLDs had their own SO, would domain name policy issues be
>required to be referred, routinely, to two or more SOs? Will that
>increase or diminish the difficulty of developing and documenting
>consensus policies?

Not going to happen often.  Few policies would take in both gTLDs and
ccTLDs.  a gTLD policy would go through the DNSO only while a ccTLD
policy (such as best practices for example) would go through ccSO only
as not applicable to gTLDs. 

>B.	Would a ccTLD SO have multiple constituencies, allowing groups
>affected by ccTLD policies to participate in ccTLD policy creation?=20

At first I thought this might be a good idea but realised that at
least in the case of dot nz those groups already participate through
our open membership system at a local level which is where many people
prefer to participate.

I would rather promote ccTLDs having open processes at the local level
rather than try to have say a global group of registrants who would
have little in common because each ccTLD is so different.

>C.	Would gTLD registries have input into ccTLD SO deliberations
>or would ccTLDs have input into DNSO deliberations? If not, how would
>disagreements be resolved?  Would the Board resolve such disputes? How
>would consensus among affected parties be developed and documented?

No IMO.

>IV.  Effect on the DNSO
>
>A.	If a ccTLD SO were created, what structural changes in any
>remaining DNSO would be necessary?=20

Well the three spots on the NC they take up could be taken up by the
Individual's Constituency.  If they leave before this happens NC drops
to 18 members only which is no hassle.

>B.	Would current contractual documents be required to be revised
>if the Names Council no longer existed or if it was not a definitive
>source of judgment on consensus policies impacting gTLD registries?
>Should gTLD registries agree to any such changes?

No change is needed.  The changes of structures does not invalidate
decisions reached under previous structures.

>V.  Board membership
>
>A.	If a ccTLD SO were allowed to elect a specified number of
>members of the ICANN Board, would that diminish the role of the DNSO
>in selecting Board members? Would it increase the number of Board
>members familiar with registry operations?

It may decrease the numbers yes.  There seems to be a principle that
all SOs get equal number of Board members so if ccSO set up I would
suggest 2 reps per SO, which means DNSO loses one.

Personally I don't think all SOs are equal and would suggest DNSO has
3, ASO has 2, ccSO has 2 and PSO has 2.  Possibly even PSO has one.

Either way likely to have more Board members familiar with registry
operations IMO.

And goes without saying at large board members should equal the number
of SO Board members or even better be one greater.

>B.	Would creation of an sTLD SO increase the representation of
>the "registry voice" on the Board? If sTLDs declined to continue to
>participate in the DNSO, would that force the creation of a separate
>SO for each category of registry?

I am not aware sTLDs have clamoured for a separate SO and believe we
can get too hypothetical.  If they can make a good case for it, let's
hear the case.

>C.	If the main job of the Board is to recognize documented
>consensus, should Board membership be reallocated to assure that every
>group that could prevent a consensus (or unmask a false claim of
>consensus) on important policy issues has an appropriate voice?

Might need a Board of 100 to do that.

>VI.  Structural proposals
>
>A.	What structural changes will make sense in light of ICANN's
>purpose?

Have suggested Board allocation above.

>B.	Should restructuring provide an occasion for reallocating fees
>and costs across a wider array of groups?

These can be examined at any time and need not be linked.  We have an
annual budgetary process for this.

>C.	Should a restructuring lead to substantial innovations in
>meeting structure and use of online deliberation tools?

I don't think the issues are linked.

>D.	Would a realignment of interests around SOs, combined with
>separate at large elections of a few Board members and the creation of
>an open forum for public participation in each SO, help to resolve
>disputes concerning representation of registrants and individuals,
>diminish complaints regarding capture of the DNSO by particular
>constituencies and overlaps between DNSO constituencies, fix the
>dysfunctional performance of the General Assembly and Names Council,
>and remedy other problems?

I think this is taking the proposal for *one* extra SO too far.  This
almost seems to be suggesting all constituencies become an SO which
would not fix the DNSO problems, merely move them up a level.

DPF
--
david@farrar.com
ICQ 29964527
--
This message was passed to you via the ga@dnso.org list.
Send mail to majordomo@dnso.org to unsubscribe
("unsubscribe ga" in the body of the message).
Archives at http://www.dnso.org/archives.html