Comments on Registry/Registrar split and Single Entity new gTLDs

["Paul Tattersfield" <gpmgroup@xxxxxxxxx>]

Dear Sirs,



Thank you for considering my comments.

The excellent CRAI report has highlighted significant points of discussion
for one small defined area which will be impacted by expanding the number of
new gTLDs in the root  namely Registrar/Registry separation.

I think it would be very helpful if consultants like CRAI could do similar
analysis on other areas of detail and for each area of detail examine the
impact of introducing new gTLDs. Whilst recognizing the current RFP is a
draft and "work in progress", it seems to introduce quite a few unknowns and
ambiguities which need to be clarified before anyone can make substantive
comments on the draft as whole. Hopefully there will be a further chance to
comment on the draft before it is submitted to the board.

It's easy to call for less regulation of markets and complain about
restraint of trade but markets are amoral not moral and as such need
regulating. As we have seen all too clearly in the recent and current
financial markets it is important not to be complacent in the face of
recklessness and greed.

Large registries do and will continue to push the boundaries wherever those
boundaries are drawn and it seems ICANN will need more tools rather than
less to ensure a level playing field. For example the warehousing of names
using shadow entities and non compliance of the RRA e.g. not permitting
registrants transfer domains to their preferred registrar.



(Using the wording of the RRA to delay the implementation of consensus
policies for as long as possible, amending user agreements and then
subsequently claiming registrants have agreed to breaking the ICANN
consensus polices of their own volition and using the notion that their own
agreements take precedence over those they have agreed with ICANN. But what
can the board actually do if they want to enforce this? It is highly
unlikely given the size of the registrar concerned that the board will or
even could terminate (which seems to be the only sanction available under
the current RRA) the registrar concerned to enforce this transgressions of
the RRA.)

At the moment there are only a handful of registries and it is likely any
attempts at capture would quickly be brought under the gaze of public
scrutiny. If there are to be hundreds of registries and some with small
numbers of registrants, attempts at capture may pass unnoticed if there are
not explicit rules against it. As is often the case rewinding wrong doings
is likely to be much more involved than preventing the problems in the first
place.

Watching major and on the whole responsible registrars twist and turn to
gain economic advantage even in good economic trading periods shows just how
important it is to have clear unambiguous rules to govern the market even
before the introduction of any new gTLDs.


I think the report begins to struggle when it comes to trying to predict the
future of new gTLDs and in doing so wrestles to introduce the concept of
"Single Entity" new gTLDs. To many Single Entity new gTLDs it seems is a new
notion and this is almost certainly because the full implications of
introducing new gTLDs as proposed has not been fully articulated in the
public arena.

New gTLDs as proposed are likely to introduce profound and historic change
to the nature of the DNS and I am certain that many do not fully understand
the implications of these changes and just how far reaching these changes
will be. I am sure many of the participants in the actual new gTLD process
know what they have in mind, but these thoughts and details should be
presented and discussed in public so that there can be a consensus developed
from the community on exactly how new gTLDs should move forward thereby
minimizing any adverse impacts on innocent third parties.

Adding a limited number of new gTLDs as and when needed allows for an almost
infinite expansion of the namespace, whereas moving "selected" single
entities, ones with huge marketing budgets for branding purposes eg. .ibm or
.google etc. will, if successful   fundamentally change the nature of the
internet.

One of the reasons for the phenomenal success of the current internet model
is a cost effective single worldwide level playing field.

If the creation of super brand single entity gTLDs is successful then by
their very success they will disadvantage all other brands, which will in
turn necessitate the other brands into obtaining their own brand gTLD. This
not only places huge additional costs and complexities into doing business
on the internet but also has the paradoxical effect of limiting choice.

(There can only be one .apple At the moment there are 260 or so apple.tld.
possible. All the different entities called "Apple" from around the world
can try and work out a way of coexisting without some registrants having to
choose a different or longer name. For many businesses and individuals this
process over the last 15 years has been resolved.
Admittedly .com is the most cherished TLD but introducing new single entity
gTLDs will not for the most part change the preference of .com over all
other existing gTLDs. Further it is unlikely any of the single entity gTLDs
will give up their domains in other existing extensions,  again the new
super brand gTLDs are providing little benefit or innovation to the DNS as a
whole.)

On a side note what happens when these brands merge? Do they have to go
through the whole process again and re-apply for a new gTLD?

One area the report doesn't touch upon are the implications from the
creation of pure generic gTLDs and how to guard against the creation of
monopoly positions. It is simple to make the statement for allowing open
competition and let the market decide, and on the surface many people will
support that notion. Of all the people who support the opening up of the DNS
to allow generic new gTLDs like .search for example perhaps run by Afilias
or Verisign etc. How many of those same people would show the same
enthusiasm if .search was secured by Microsoft?

In most systems, a trademark will be registrable if it is able to
distinguish the goods or services of a party, will not confuse consumers
about the relationship between one party and another, and will not otherwise
deceive consumers with respect to the qualities of the product.

On this point the proposed new gTLD process seems to be divergent with
current Trademark Law.

Yours faithfully,





Paul Tattersfield
Grange Project Management
http://www.gpmgroup.com – blog http://www.gpmgroup.info