[wg-c-1] An approach to the gTLD question

[Ivan Pope <ivan@netnames.com>]

New gTLDs?

1. Should there be new generic Top Level Domains (gTLDs)? 

a) If yes: How many?

b) What should the new gTLDs, if any, be?

2. If there are to be new gTLDs, how should they be deployed? Should they be
introduced all at one time, one at a time, or in groups, and over what time
period. In what order should they be introduced?

3. Assuming that a limited number of new gTLDs is to be deployed over a
period of time in the near future, what should be the mechanism for
developing new gTLDs after the first new gTLDs are deployed.

4. If there are to be new gTLDs, should each one, or certain ones, have a
specific charter? In other words, should some gTLDs be limited to use by
certain entities, and if so should the limitation be mandatory? If the
limitation is to be mandatory, what should be the enforcement mechanism and
who should be the enforcement body? 

New gTLDs

I feel that the best way to come up with a gTLD solution that works is to
attempt to define a mechanism that would bring new gTLDs into the world in a
coherent and understood manner. Rather than ask specific questions as above,
I think we should assume that it is the job of this WG to come up with a
workable process that may or may not ultimately produce new gTLDs, but that
the WG should not concern itself with specific answers to questions such as
'how many' or 'for who'.

Firstly, we might like to set out some principles for new gTLDs (Remember:
these are suggestions only!):

Principles for gTLDs

New gTLDs should be shown to be in the general interest of the Internet

Management of a gTLD should be a privilege, not a right

Management of a gTLD can be, but does not have to be, a profit based
operation

New gTLDs should be inclusive rather than divisive.

Or

New gTLDs should be available to the person who shouts the loudest

New gTLDs should be given absolutely in perpetuity

New gTLDs exist only to create profit

New gTLDs should be run at the whim of the owner

(Probably somewhere in between!)

To expand on this further we might like to ask some different questions
about gTLDs:

1.	What is the essential nature of a gTLD
2.	What are the reasons for wanting to create a gTLD
3.	What are the barriers to creating a gTLD
4.	What are the benefits to creating a gTLD
5.	What are the disbenefits to creating a gTLD
6.	Does a gTLD have a value
7.	Who are the interested parties to the creation and running of a gTLD
8.	Who may be harmed by the creation of a gTLD
9.	Who may benefit from the creation of a gTLD

Once we answer some of these, we will see the beginning of a route forward.
We may find that we have shed some of our preconceptions and allowed
ourselves an approach that will work. 

Then we can set out to create an institution that will allocate and oversee
gTLDs within a framework set out by ICANN.

What is the essential nature of a gTLD?

I've never actually seen this discussed, but it seems to have some bearing.
What do we mean when we talk about 'new gTLDs'? We don't really mean the TLD
itself, but the authority to control entries into that TLD. So when we talk
about creating new gTLDs we are talking about an act that both creates a new
TLD and bestows some authority over it. The first thing that this brings to
mind is that in order to have some authority, there must be some contract.
So are we talking about creating a contracted relationship between one party
and another? If we are, we need to look at what we want that contract to
achieve. So we should look at this point at our Principles and see what we
are trying to achieve.

What are the reasons for wanting to create a gTLD?

We can be as frank as we like about this, but it would be useful to think
about it. Do we want to create new gTLDs in order to relieve some pressure
on existing gTLDs, or to create now Internet opportunities, or to create a
Directory of the Internet, or to weaken existing monopolies or what? The
question is as much an attempt to look at what reasons those who will ask
for new gTLDs will give. Why do we want new gTLDs? Is there one reason or
dozens? Can we accommodate them all or some or none? Reference to the
Principles will be useful at this stage.

What are the barriers to creating a gTLD?

There certainly are barriers, otherwise we would have new gTLDs by now. Are
those barriers legal, technical, practical? Can we learn a lesson about how
to create new TLDs by looking at what stops them just existing? I would
guess that one reason that they don't exist is because there are conflicting
interests that have different desires and understandings of what they are
for. Are we here to resolve those conflicts, to explain what gTLDs are for
or to create a new understanding? If we don't understand the barriers, we
risk running up against them again. Thus, our best laid plans will be
useless. 

I don't want to run through all the questions, and there are a lot more. I
want to set out a way forward that allows us to ask and answer questions and
not to rehash all the old arguments.

I would suggest that we accept that there will be new gTLDs.

That they will be run by a variety of organisations for a variety of
purposes

That there are better and worse TLDs available

That there is a commercial imperative involved that we cannot ignore

That there is also a non commercial imperative involved

I would suggest that we set up some criteria under which any body can apply
to be 'allocated' a TLD for management for a fixed yet extensible period of
time. The allocation would be made under contract with [ICANN]. The contract
would be standard and minimal. The Principles would be adhered to within the
contract. No extreme or unreasonable demands would be placed on applicants.
The Principles would 

Each TLD allocation would be made by a professional with no vested interest
in the outcome who would just interpret the rules.

Each application for a TLD would have to satisfy criteria:

*	Show existing commercial or IP need for such a combination of
letters

or

*	Demonstrate that there is popular demand for such a combination of
letters 

and

*	Show that there is no existing conflicting claim to the combination
or if there is a conflicting claim that accomodation has been reached
between the conflicted parties

What does this mean?

*	Show existing commercial or IP need for such a combination of
letters

	Applicant would have to show that they had a commercial or
intellectual property situation that gave them a need or desire for a
specific combination. (This is designed to allow the AOLs and the Chris
Amblers to demonstrate or not that there is a requirement for their chosen
combination.)

*	Demonstrate that there is popular demand for such a combination of
letters
*	We might mandate that a vote was held, or some populist exercise
undertaken. (This would be the 'startup' option.)
*	Show that there is is no existing conflicting claim on the
combination or if there is ...
*	Applicant would have to show that there was no known conflicting
claim or that if there was, that the conflicted parties had got together and
come to an agreement over the management. (This might be known as the
CORE-Ambler clause.)

Remember, a combination of the criteria would be needed. So an applicant who
claimed IP reasons would also have to show no conflict 

If we are smart and hold our breath, we should be able to create a situation
where we can get a few new gTLDs out into the world. Once we see that the
world does not end, we can probably go on to create a few thousand.

If we are smart, we should be able to satisfy the undoubted demand for new
gTLDs without breaking ourselves in the process.

If we are not smart we will never have anything. We should bear in mind that
we are all conflicted as hell in this WG. We should understand that we are
either all or none of us going to benefit from new gTLDs. So we might as
well let the party start and let the Internet breathe.

I look forward to further discussion on this subject.

Cheers,

Ivan