A position paper

[Daniel Kaplan <dkaplan@terra-nova.fr>]

Following is a draft paper we have begun discussing within the French 
Internet community (please read the "status" paragraph before you say it 
comes somewhat late).

We provide it to the list:
- because it took us some work to write it (in bad English) :-)
- because we'll use it as background to some of our remarks on IAHC's 
dec. 19 draft, which we consider an excellent proposal.

After sending this paper we'll keep silent for 1 or 2 days, trying to 
catch up with the hundreds of messages which have accumulated while we 
were celebrating Xmas...

Regards to all,

Daniel

-------------------
POSITION PAPER ON IAHC DISCUSSIONS

December 20, 1996

DRAFTED BY:
- Daniel Kaplan (dkaplan@terra-nova.fr)
- Olivier Iteanu (iteanu@calva.net)


1- STATUS OF THIS DOCUMENT:
- This document has been drafted at the request of the French Chapter of 
the 
  Internet Society and the French Telematics Association (AFTEL). It has 
yet to be 
  approved by both organizations and discussed between the members of the 
French ad 
  hoc mailing list (isoc.nommage@univ-rennes1.fr)
- While it arrives after IAHC's 12/19 draft, we still believe it to be a 
useful 
  contribution to ongoing IAHC discussions, both (i) as a basis for our 
future 
  comments regarding the draft and (ii) as food for thought in the 
ongoing process 
  which IAHC has only initiated.
- The authors are not technicians: Daniel is a consultant and Olivier is 
a lawyer.
  They are therefore willing to admit that some of the technical 
consequences of 
  their proposals may have been overlooked.

2- STATEMENT OF PRINCIPLES
2.1- The DNS is not a directory. However, its role is to provide mnemonic 
ways to reach hosts, and its databases will (must) be used by the 
providers of directories.
2.2- TLDs are not private properties to be managed at their owner's will, 
rather they are a common ownership of the Internet community. Only the 
leafs of the DNS tree may be considered as private properties.

3- IS THERE A NEED FOR NEW iTLDs?
Yes.
3.1- Because iTLDs already exist and removing them is not realistic;
3.2- Because we feel it necessary to promote "universality" 
("universality" not being the opposite of "diversity") on the Internet;
3.3- Because the present concentration of names within .com creates too 
many problems, the two main ones being:
     * name speculation
     * excessive advantage for first comers.

4- IS THERE A NEED FOR SHARED iTLDs?
Yes.
4.1- Because iTLDs are not a private property;
4.2- Because it provides for a better stability of the DNS, in case one 
registar disappears;
4.3- Because it promotes competition, and therefore better and cheaper 
service for registrants;
4.4- Because the present .com situation results in the fact that only US 
courts are able to solve name disputes.

5- MUST WE PREDEFINE WHICH iTLDs WILL BE CREATED ?
Maybe in the short run, not in the long run.
5.1- In the short run:
- We believe the expansion of iTLDs should be an experimental process, 
and undergo 
  continuous reviews of its consequences and possible drawbacks;
- The first new iTLDs to be created could refer to very generic 
categories, for 
  which we feel that the Internet community has a right to predefine a 
rather 
  stringent set of rules (a charter) to be enforced by registries.
5.2- In the long run:
- We will always fail to properly map the naming universe. It can only 
properly map
  itself by a continuous trial and error (or "sort-of-free-market") 
process;
- Prior mapping of the naming universe, eg by using WIPO trademark 
categories:
  * neglects the fact that not only trademarks are protected (there can 
be 
    protection for artistic works, for "famous" persons names, etc.);
  * neglects diversity in the way names are protected in the various 
countries
    of the world;
  * will probably neglect the fact that not everybody speaks English in 
the world.

6- HOW SHOULD iTLDs BE CREATED IN THE LONG RUN?
6.1- iTLDs should be created through an iterative process managed by 
IANA. This  process should be continuous.
6.2- IANA should approve or reject proposals based on the usefulness of 
the new iTLD to the Internet community. This usefulness will mostly be 
asserted through:
- the iTLD's name;
- its proposed charter (see below some provisions that should be included 
in the charter) and policies.
6.3- New iTLDs should be created with a "pilot" status, pending review of 
their achievements after 3 and 6 month. They will then reach an 
"operational" status.
6.4- In order to reach "operational" status, an iTLD must be operated by 
at least two independent registries based in two different countries.
6.5- Two kinds of new iTLDs may be proactively created by IANA, with 
stringent charters designed in order to ensure maximum efficiency for the 
Internet community:
- "Generic" iTLDs which designate a common resource of the Internet (eg, 
.WEB, 
  .WWW, .MAIL, .NEWS...)
- Some iTLDs reserved for non-profit organizations.
One of the provisions enclosed in the charter for those iTLDs should be a 
low price-cap for registrations, or a profit-cap for registries.

7- HOW SHOULD REGISTRIES BE APPROVED?
7.1- Registries receive a delegation from IANA, not an permanent right. 
This delegation is valid for a period of (3-5) years.
7.2- Registries will be approved on objective criteria such as:
- Proven technical ability to run the registry;
- Financial stability;
- Specific charters and detailed policies designed to enforce the 
charters 
  of the iTLDs they intend to run;
- A commitment to have a non-discriminatory attitude towards registrants 
as well as
  towards other registries sharing the same iTLDs.
7.3- In order to facilitate sharing and the creation of directories, 
registries should commit to a common "replication" method of their domain 
database. By "domain database", we mean only the part of the DNS tree 
that the registry organizes; the registrants' names, physical 
coordinates, etc., are not part of that replication. A specific RFC 
should define the replication method.
7.4- Should registries fail to enforce these commitments, IANA may remove 
their delegation. In case of a conflict regarding the interpretation of 
these commitments, both IANA and the registries should recognize an 
organization (eg, ISOC) as the appeals authority.

8- HOW DO WE DEAL WITH TRADEMARKS AND OTHER PROTECTED NAMES?
8.1- Two principles must first be asserted:
- registries can simply not check wether an application infringes on any 
protected
  name. Indeed, some *trademark* registries (such as INPI in France) 
don't do the 
  checking: they just assume the registrant has done its homework and 
knows the 
  risks it's taking.
- however, we must (and can) make sure that:
  * nameholders who feel that a domain name infringes *can* find a court 
in the 
    country where the registry is located;
  * there can be sufficient presumptions of good faith from the 
registrant to 
    reduce the number of disputes and/or clarify disputes.
8.2- Registries should be located in a country which has signed the Paris 
Union Convention on Industrial Property (132 countries) and (or?) the 
Bern Union Convention on the Protection of Litterary and Industrial 
Property (111 countries in 1995).
Why? Because if a registry is not located in one of those countries, 
chances are noone can prevent a registry from colluding with registrants 
and massively speculating with trademarks or other protected names.
8.3- iTLDs / registries charters should mandate that registrants provide 
some documentation which asserts that it is a legitimate holder of the 
name it applies for: first or last name, trademark, company name, title 
of an artistic work...
The registry does not need to check for the document's validity: it only 
needs to store a certified copy. In case of a dispute, the court will do 
the checking, but this will help it evaluate the registrant's good or bad 
faith, and right to the name.
8.4- Apart from these provisions, registries should commit to "take all 
appropriate measures in order to prevent or limit name speculation, and 
abide by IANA recommendations in that order". Future experiences may (or 
may not) bring IANA to issue recommendations in order to reduce name 
speculation, and the registries should follow these recommendations.

9- SHOULD WE GIVE PRIORITY TO ORGANIZATIONS WHO HAVE ALREADY BEEN WORKING 
ON NEW iTLDs ?
Not formally, at least.
We should definitely disallow all preemptive attitudes towards new iTLDs, 
which have used little-publicized openings by IANA to try and assert 
their rights over valuable iTLDs.
However, having worked on the topic for several months, these 
organizations stand very good chances to become some of the first 
approved registries.

We welcome everybody's comments.

Daniel and Olivier

-------------------------------------------
Daniel Kaplan         dkaplan@terra-nova.fr
Consultant - Media & Commerce Electroniques
           - Electronic  Media  &  Commerce
61 rue Monge   -   75005 Paris   -   France
Tel/Fax +33 (0)1 4217 0754 GSM 06 0981 0377
-------------------------------------------