Public interest obligations in .net?

[Harold Feld]

Although .net is a generic TLD, it is useful to consider what, if any,
public interest obligations should attach to this transfer.
Because the NCUC is the one recognized place in the ICANN policy
structure for "official" non-commercial input, I hope we will be willing
to explore and champion the question of what (if any) explicit
obligations should apply to the .net TLD.

I cannot speak with any great expertise on international telecom
service.  I hope others more educated in international
telecommunications services will speak from their experiences.

I suggest that in conceptualizng the .net TLD, we think of this as a
resource similar in many ways to telecommunication numbers.  In the
United States, we have closely regulated telephone numbers for a number
of purposes.  First and foremost has been technical stability.  The
telephone system must, above all else, work.

Second, interoperability among systems must be preserved.  This includes
international interoperability (we want to be able to call other
countries) and interoperability between competing networks (wireless
phones and wireline phones must be able to call each other).

Third, universal service has been an important goal of the United States
telephone system since at least the Communications Act of 1934, which
established a broad regulatory regime for telephony services designed to
facilitate deployment of the telephone to all places in the Unites
States.  Milton can speak to this with far greater expertise.
Regulation of numbers to achieve this goal, rather than regulation of
the telephone system as a whole, was limited to the technical
coordination under items one and two above.  Nevertheless, given the
importance of the gTLDs to providing alternatives to ccTLDS in some
countries, I recommend consideration of the following aspects to promote
a broad concept of universal service for TLDs.
a) IDNs.  It is high time ICANN got over its Verisign fixation and got
into gear on IDNs.  The divestiture should therefore require/faciliate
more than the promise to obey ICANN decisions on the matter, as was done
with .org.  Rather, the entity bidding on .net should be required to
make a specific commitment of time and resources to working with other
TLDs to develop a working IDN system.
Arguably, this issue goes beyond .net and should be broadly addressed by
all TLDs.  Why the TLDs shgould wait for the IETF when a working group
among TLDs is what is needed is beyond me.  But we must start somewhere
and raising the profile on this issue is important.  While it is
presemptuous for an English-speaking American to champion IDNs as a
critical issue for universal deployment, I have heard a number of people
make this complaint.

b) Cost.  Marc is right that, since we seem stuck with centrally
controlled pricing, we should urge that the set wholesale price be
considered as part of the application.  But what about other issues of
purchasing domain names?  Are those in other countries at a disadvantage
because of lack of American currency?  Or does the near universality of
credit crads address this issue?  Is this something that can be
addressed at the TLD level?

c) Privacy.  In the United States, we spend a considerable amount of
regulatory effort on protecting the privacy of those who register phone
numbers.  What kind of commitments can we look for, especially in light
of the varioius WHOIS task forces?  If nothing else, a commitment to
privacy as a normative value (even if it is ultimately evicerated by the
intellectual property and law enforcement communities) is useful in and
of itself.

It is also important to look at the failures of regulation.   In the
past, the U.S. sought to encode useful user information as part of the
number content.  e.g., at one point, the three digit area area codes in
the United States were the only numbers with either 0 or 1 as the middle
number.  This was dropped many years ago because it simply did not
scale.  What concepts are currently embedded in registry agreements that
should be dropped? For example, now that Verisign has spun off its
registrar, can we please eliminate the restriction on communicating
directly with registrants?  Arguably, this made sense when a registry
controlled a registrar and could use its registry information in an
anticompetitive fashion. But now all this requirement does is get in the
way of potentially important communications.

Harold Feld