[ga-full] RE: [ga] Older registrations

[Simon Higgs <simon@higgs.net>]

At 11:23 PM 3/20/00 +0100, Roberto Gaetano wrote:

>Simon Higgs wrote:
> >In 1995/1996, IANA solicited new iTLD requests under the guidelines
> >laid out in RFC1591. IANA received many new iTLD applications and published
> >the results on the IAHC-discuss mailing list (archived at www.gtld-mou.org
> >for some inexplicable reason). At the same time, IANA started the process
> >to establish unquestionable authority to formally introduce new TLDs
> >(beginning with the "Postel drafts" and now under ICANN). The iTLD
> >applications received by IANA were printed out and placed in a file at
> >IANA (I've seen the file - it exists) pending the results of that process.
> >
> >And here we are.
>
>I was not there, but I would like to comment anyhow on your passages
>above.
>
>What strikes me is that, at the origin, we had a "simple" problem of
>enhancing the Domain Name space with the introduction of new gTLDs (or
>iTLDs, if you prefere).
>We went through different efforts, committees of any sort, papers of
>different colours, even the attempt to define a new model for Internet
>Governance, but, as you well put it "and here we are", years later, with
>  a lot of discussions, and maybe even results, but with the starting
>question about new gTLDs still unanswered.
>
>Why?

Roberto,

OK, here goes. It will almost certainly garner the ire of those who should 
know better (watch the dismissive responses to my messages to see who they 
are). I have included references where I could.

When the original discussions took place, there were a couple of dozen 
people, mostly communicating via mailing list (newdom). The original 
intention of IANA was to do three things:

1. Open up domain name registration competition to NSI by the creation of 
new registries and TLDs
Before the feasibility of shared-tlds, the Postel plan was to create a 
number of new registries that would each provide up to three new TLDs. This 
is what generated the majority of the TLD applications. Each registry would 
theoretically be equal to NSI's .COM/.NET/.ORG franchise and provide 
economic incentives for pricing competition.

2. Open up competition to NSI by sharing registration services
This was the creation of the registrar function and was originally 
discussed on newdom. I created the original shared-tld mailing list to see 
if a shared registry system was feasible. It was very straight forward on a 
technical level but fraught with policy & economic incentive problems which 
were not properly discussed - in fact the problems that ICANN is now having 
were foreseen but have been continually dismissed. Later versions of the 
Postel draft included shared-tlds (Jon copied the outline from my draft 
with my permission). My draft at the time[1] also proposed guidelines for 
defining the type of registry services a TLD would require - 
Shared/Specialized/Private - which I gather is now being explored by the 
various ICANN sub-sub-sub-committees. I also proposed the creation of a 
descriptive charter for each TLD which described both it's purpose and how 
it would be delegated. Kent Crispin has since championed the TLD charter.

3. Create a small series of ad hoc committees
These small ad hoc committees from the internet community that would 
evaluate applications for new registries. There would be a published set of 
criteria which would determine if an application was to be approved. This 
included the selection and maintenance of the new TLD(s) that the registry 
was created to support [remember who serves whom here - it's a fundamental 
law that everyone seems to ignore]. The committee would be informal and 
have a revolving membership to allow continuity, but would not allow 
unnecessary outside interests to interfere with the approval process.

Now we come back to reality - what actually happened is as follows:

IANA indulged in some experimentation with various people who had expressed 
an interest in new registries/TLDs. Participation in the AlterNIC root 
system was actually encouraged for a while for functional testing of the 
proposed new registries. AlterNIC failed for a number of reasons we don't 
need to go into here, but the testing which IANA had proposed continues to 
this day. By the time the IAHC had come into existence, several personality 
conflicts had emerged which resulted in outright lies being broadcast about 
various people's intentions. These include IANA-acknowledged registry 
applicants fulfilling IANA-proposed operational requirements being accused 
of TLD squatting.

Unfortunately, the IAHC committee that was eventually formed was the 
complete opposite of what was originally envisioned. It was a blue-ribbon 
interest-serving panel (ISOC, IANA, IAB, FNC, ITU, INTA, WIPO), all of whom 
had a claim to control the internet via the control of the domain name 
space. None of the committee members were prepared to work within the 
framework of what had gone on before (even the IANA representatives opposed 
Jon Postel's public views[2]). The IAHC accepted shared-tlds as the 
exclusive standard for introducing TLDs without attempting any due 
diligence or study of the real-world consequences. This had the effect of 
instantly disenfranchising the majority of the new-TLD supporters and all 
of the IANA-acknowledged applicants[3]. The committee then created it's own 
exclusionary community, with a [flawed] signatory document (the 
gtld-mou[4]) as the mandatory method of entry. The gtld-mou community 
(under the auspices of CORE[5]) then attempted to create 7 new shared-TLDs, 
two of which conflicted with IANA-acknowledged applications. This resulted 
in a lawsuit from Image Online Design (who was part of the pre-IAHC 
face-to-face discussions with IANA). The point isn't in the lawsuit (it was 
withdrawn leaving the findingst inconclusive), but the fact that TLDs were 
chosen deliberately to conflict and eradicate the validity of the previous 
IANA work.

Briefly, CORE started to implement a shared registrar system, and had their 
test registry system stolen in mysterious circumstances. Fortunately, CORE 
has now become .COM/.NET/.ORG registrars. However, along the way CORE used 
Jon Postel to attempt to split the root zone in order to get their proposed 
TLDs into the non-US Government controlled root servers. The US Department 
of Commerce/NTIA rightfully intervened[6] and has attempted to put the 
process back on track by establishing IANA's successor, ICANN[7].

History aside, the reason I believe we are no further along is because the 
fundamental laws of domain name delegation are not being recognized. Top 
level domains can be divided into three groups which determine how they are 
to be delegated:

1. Shared (non-exclusive) which are served by multiple competing registrars 
AND REGISTRIES
2. Specialized (exclusive) which are served by a single registry that has 
the necessary expertise to address specific industry issues (.INT)
3. Private (exclusive) which are required to serve single large 
organizations (.GOV, .MIL)

In the case of Shared TLDs the registry back-end should also be 
competitive. The only practical way to achieve this is to use the Postel 
draft to create the registries for new shared-TLDs and the ICANN registrars 
to provide the public interface to these TLDs. There's nothing wrong with 
combining two good ideas, since everyone will get what they want.

Another issue is in recognizing the correct ownership of the name space. 
RFC1591 states that the name space is recursive. What works at the 6LD 
works at the 5LD, and also works at the 4LD, 3LD, 2LD and ultimately the 
TLD. Obviously, exclusive use can be delegated at any level, and the rights 
associated with exclusive use are inconsistent with the idea of a public 
trust. The gtld-mou deliberately failed to recognize this, claiming the 
rights to the TLD layer under the guise of public resource[8]. I'm not 
saying that TLDs must be owned, but instead saying that famous marks may 
have enough rights for exclusive delegation.

Next you have to decide what new TLDs to introduce. Let's assume you have a 
list of proposed TLDs by various organizations. What is stopping these from 
being introduced? Think about this carefully, because the reasons you will 
end up with are the real stumbling blocks to the growth and stability of 
the internet. Is it because:

1. the organizations requesting TLDs cannot qualify as registries?
2. the trademark community can't decide what trade categories will minimize 
conflicts?
3. the lawyers can't keep up with the Intellectual Property work?
4. the [governmental body] won't be able to regulate/tax the name space?
5. the [company name] won't be able to profiteer from the name space?
6. the [insert reason/motive here]?

Roberto, there are a lot of reasons "why?". I doubt if I've mentioned them all.

[1] http://www.iahc.org/contrib/draft-iahc-higgs-tld-cat-03.txt
[2] http://www.gtld-mou.org/gtld-discuss/mail-archive/00527.html
[3] http://www.gtld-mou.org/gtld-discuss/mail-archive/00990.html
[4] http://www.itu.int/net-itu/gtld-mou/gTLD-MoU.htm
[5] http://www.gtld-mou.org/docs/core.html
[6] http://www.ntia.doc.gov/ntiahome/domainname/domainhome.htm
[7] http://www.ntia.doc.gov/ntiahome/domainname/proposals/icann/icann.html
[8] gtld-mou SECTION 2. - Principles
     The following principles are adopted:
     a.the Internet Top Level Domain (TLD) name space is a
     public resource and is subject to the public trust;


Best Regards,

Simon

--
DNS is not a sacred cow that cannot be replaced by something better.

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