[ncdnhc-discuss] Re: [Implementation of Evolution and Reform] Another Exploration?

[J-F C. (Jefsey) Morfin]

Dear YJ Park,
This is a long response to keep responding on this complex issue. I hope it 
helps. Obviously a face to face meeting would be better.

I am very sorry to have to refer to my "happy days" as an old papyweb. But 
I think that most of the issues we are confronted to today come from a very 
simple confusion: between what the things have been designed for, and the 
way we would like to use them today to address our today problems.

There is the Internet and there is the real world. Initially the Internet 
was a "private" network and there was a public network system in the USA, 
then France, Germany, UK, Belgium, Italy, Japan .. and intereconnecting up 
to 55 countries (I talk of what I know, that is until 1986: what is OK as 
the internet structure we live with was set-up in 1983/85).

The Internet people and the ICANN people tought by them mostly see the 
world through the Internet development. Without geting real and refering to 
the real world you miss a lof of explanations, solutions and understandings 
of what is happening and goind to happen. You are like a kid believing that 
the world is his family.

1. the international public (ITU Members) network develops from 1977. 
Leading users communities are the USA and (roughly same traffic amount) 
France, UK, Germany with progressively large Packet Switch services 
Transpac, PSS, DATEX-P, then Minitel and millions of users). Then came 
Belgium, Japan, Italy ...

2. the namespace is then managed by Tymnet under FCC license, through the 
US IRC interconnect agreements and direct relations with the monopoly PTT 
or licensed Operators named after ISO 3166 3 letters codes and operators 
code. Some large private networks are connected through these monopolies 
with their own namesapce Bull, Philips, ESA, TRW, BofA

3. the development of the X.75 connexions and the vote of the X.121 numerci 
adressing standard   lead Tymnet to switch all the name oriented services 
towards X121 compliant numeric names (1982, completed in 1986).

4. in 1983/85 Internet wants to use the public network to supports its 
foreign relations, connect remote US hosts and provide local access to 
users. This leads to some naming organization and contractual connection 
agreements to avoid naming conflicts in names and numbers and to support 
local Internet communities  - the DNS design (RFC 882, 883 etc), 
implementation (RFC 920/921) and ISO 3166 2 letter code usage for ccTLDs 
take care of all this. The results is the defacto ".arpa" sub-namespace 
delegation (none at the time used such grandiose wording). It is extermely 
well respected until now. The only changes since 1985 are the eventual 
support of E.164 (ENUM) and the permitted additions of 6 TLDs and the 
".biz" conflict (I do not judge).

5. in that arrangement the Local Internet Community are the people granted 
a PPP access or an Host connection on ARPANET through their local public 
service. The way they connect is either by phone to the USA on ARPANET or 
through a Tymsat access with a local PTT Tymnet username. In some countries 
dial-up PAD is provided to the X.25 public service - but the X.25 "segment" 
rate made the connexion very expensive or slow, leading France, Germany, 
Switzerland etc. to keep the Tymnet access. The connexion to a foreign 
ARPANET host was complex as the ARPANET system was not allowed to carry 
international traffic and you had to access a foreing ARPANET gateway (US, 
UK, France).

So you see that States were Operators. And ccTLDs were local Internet 
@large communities supports. A big big difference.

Then came deregulation permitting new possibilities. But ccTLDs were still 
local Internet @large community support.

Then came the web, with a blunt increase in usage and a progressive 
important local use of Internet. This made the Internet users community to 
merge with the regulated operators users community while new non regilated 
operators (ISPs) appeared.

Here is the problem. States have rights on telecoms and duties to 
operators. States have no right on the @large communities and Internet 
registration processes. But @large communities are becoming the whole 
population and the operators are the main ISPs, merging the private 
Internet system with the public telecom system.

This problem is increased by the technology development which makes 
characters to win on digits. The real command of the world is going to be 
URL and no more the dialer. The brainware (the way the user uses the 
system) is sophisticating and the namespace becomes a key commodity.

This creates major changes. In that situation ICANN has two options :

1. to mess the things in wanting to "coordinate" (which means control) - 
this will lead to conflicts and network political splits.

2. to respect the RFC 920 in :
     - being the ".arpa" manager controling its gTLDs, some sTLDs,
     - relating  with
       - the ccTLDs,
       - the moTLDs,
       - the other complex namespaces (US, EU),
       - the ITU namespace (E.164)
       - and large nomenclatures
       through a G8 like committee for parallel development under rotating 
Chair

       Possibly offering it:
       - a secretariat service
       - the ITU-IANA service under International status and protecting the 
net from ITU/T

Another problem is that the ICANN "products" (DNS and IP addressing) are 
not stable and footing the need. IPv6 is now a political problem making Gov 
unhappy with this service to their economies. The DNS is to be reshaped and 
adapted to what it does in a real world (DNS.2) and extended to what people 
would expect it to do (DNS+)

To measure the real world problem. I will only take the 1986 French example 
as a maximum case, hence easier to understand the ballances we come from 
(and the roots of the situation):

- a public data network system (Transpac) under state monopoly with a tens 
of thousand of hosts.
- an international gateway (NTI) supporting both Tymnet protocol (to 
progessively close) and X.75 with relations with 55 countries using 5 US IRCs
- the Minitel service with 12.5 terminals and 40.000.000 of users.
- the ".fr" NIC with (then a few 10s of DNs, today 145.000 Domain Names)
- the Internet community (then a few thousands, today 6.000.000 users and 
20.000.000 one shot users)
- the @large community (then a few thousands and 1400 people at the 2000 
election)
- the france at large association (only one structured as @large) and the 
leading city of the icann.meetup ... 7 peoples.



At 06:35 31/07/02, YJ Park wrote:
>Dear Jefsey,
>Thank you for sharing your thought.
>
>Giving my full support to your comment that "I want it dominated by the
>@large, animated by the ccTLD under the protection of their Govs.",
>let me share the concerns I have before we deal with your proposal
>with details.

As you see the real problem is the "commercialisation" of the ccTLDs under 
the influence of the NSI philosophy and the DNs sales. Now let suppose that 
in Korea the first 2 DNs would be free, life long, allocated by the 
Internet @large user association and that to use the Internet you are to 
pay a yearly fee of $5 through your "internet presence" (web site and 
e-mail) provider.

> > >Q1. Can you explain what could be and should be the roles for the
> > >governments in the area of ccTLDs and gTLDs under this scheme.
> >
> > The Governments have no specific role: They have legitimacy and law over
> > what is national. ITU has legitimacy from Govs over what is international.
> > ICANN has legitimacy and delegation over what comes from the USG in the
> > Internet (.arpa today still leading sub-namespace) .
>
>Yes, I agree. The Governments do have a superior status to any party
>in their territory which has been established for centuries and centuries.
>
>This kind of "superior status" for most governments has not been quite
>applied to cyber space territory which can be interestingly compared with
>US. US does still have such status by initiating ICANN process under
>its oversight.

The US have the management of the ".arpa" namespace which related with 
ccTLDs' namespace.

But the ARPANET was a structured club of Hosts. Abraod it was a club of 
users. Hosts came from ISPs - merchants. So the ccTLDs are often now an ISP 
association instead of being a Host association. Would they be an host 
association they would be today an @large association. The real problem is 
business in non-profit structures. if it becomes a telecommunications 
business, it should fall under the deregulation rules.

But at the same times everyone understands that this is brainware/extended 
services. So it is special. We lack the modelisation and the understanding 
of it. We really have a cultral problem as we just do not know what we are 
talking about : telecoms, computers, information?

I proposed a model in 1985 which is still qui accurate. There may be 
others. But it shows the problem we face. Page 12 in 
http://utel.net/net_soc.pdf (beware in French)

>The challenges from "Internet community" against government has been
>so far "the governments except "US Gov" have not contributed to
>create and even sustain Internet therefore the governments in general
>do not comprehend Internet and Internet will be endangered if the gov'ts
>try to step in to this kind of coordination.

French Gov did sustained the Internet :-)  Ask Louis Pouzin (he fathered 
datagram and IP zones and used the ARPANET team as consultants at a given 
time to get them some budget).

But actually the entire ITU world sustained the Internet as I started in 
explaining. They gave it access. As I said the problem is to uderstand what 
falls under telecoms and what is not. What is information and what is not. 
What is IP and what is not. Where govs fit in and dont.

>My understanding is ICANN has tried to represent the Internet community
>all of sudden from 1998 under DoC's blessing by organizing several SOs,
>constituencies, committees, task forces, etc which has been still poorly
>particiapted in from many parts of the Internet-penetrated countries asking
>the governments in the GAC to observe ICANN's performace under DoC's
>oversight.

I was not here. But I feel they tried to coordinate instead of cooperate 
because for two reasons:

1. the lackof an us ccTLD to make them behave as others
2. the lack of understanding of their community representative role by the 
ccTLDs.

Let imagine that in 1998 the KRNIC had welcome an Internet meeting with 
30.000 people attending and said they wanted the Korean Internet to be 
Korean and integrated with the other national and International virtual 
newtorks ... a few things might have been different.

>Despite the governments' exclusive authority in offline space, at that time
>the assumption presented by the self-claimed "global" Internet community,
>ICANN, was the governements should not and could not exercise its
>power over Internet even local Internet community since Internet is global
>resource which is connected to all therefore cannot be controlled by
>one gov't.

This is the bluff. The ".arpa" is a sub-namespace. The ICANN has no global 
right and even the whole Internet has no global right (look at ENUM) but 
claiming it may be with woth $ billions (look at VRSN and ENUM).

This ".arpa" sub-namespace looked unique by accident. This is not a stable 
thing.
You cannot rule the world as if was a part of Kentucky.

>As a result of it, ICANN, self-claimed "global" Internet community could
>exercise their power as both technical coordinator as they promised
>and even policy coordinator which initially disavowed and later insisted
>on full recognition from the unidentified Internet community so far.

yes. But let be candid the problem was not easy. Joe Sims had the problem 
of not understanding (probably not even reading) RFC 920. He did not 
support the Internet but attepted to build an Intercontract.

For the network to work the ICANN is to administer the ".arpa" and relate 
with the other spaces. They used more the word "net" than the word "inter".

>and we all of sudden start to discuss the roles for the rejected party,
>more exactly intentionally gagged one such as the governments
>watching global at-large members thrown out by ICANN who used to
>be recognized as tools to give legitimacy to the self-claimed global
>Internet community, ICANN.
>
>I am now wondering whether the whole arguments we went through
>to make NCDNHC and Internet users voices heard in the ICANN
>process was merely used to fortify each others' positions during a
>political tug-of-war between the governments and Internet community.
>Ironically, NCDNHC and Internet users seem not be fully accepted or
>recognized as serious counterpart or part of their decision-making
>structure by neither party.

The whole ICANN (Joe Sims) system is to make the legitimate "owners" (the 
users, @large) to discuss while the operators (ATT/IBM/SAIC/MS/WOLDCOM) 
keep running. This is not necessarily stupid as at least it makes the 
Internet to operate.

Now, we have to ask ourselves is "if we say that the ccTLD and the users 
are part of the network, should we not demonstrate it by some @large based 
network projects. We should be part of the IETF, part of the ccTLDs, 
candidates to the BoD, etc;"

jfc