[ga] IDN issues (was: On Elections)

[JFC Morfin <jefsey@xxxxxxxxxxxxxxxx>]

At 10:00 05/10/2007, Debbie Garside wrote:
> I agree.  But if you remember, my involvement started when I could not bear
> to see 90% of the posts on this list bashing ICANN over the head with a big
> stick in a completely destructive, certainly non-constructive way. :-)  The
> other 10% being made up of posts threatening litigation/slanging matches
> with only 1% being useful, informative and constructive.
I note that the response I am going to make here is authoritative.
This is therefore typically what Danny, Sotiris, and Debbie want for 
the GNSO/GA.
> >Do you have any idea how much expertise is
> > represented on this list?  I think it's pretty safe to say
> > that after 8 years of existence this ML community has matured
> > to a point where knowledge has an open arena... shall we not
> > make use of it in an ordered, measured, thoughtful way?
>
> That would be an excellent idea - wish I had thought of it before :-)
I hope it is presented and will be read and possibly debated in an 
ordred, measured, thoughtful way. This is why it will be a report as 
one may expect in a competent decision making conference. This is 
also why I will ask for advices from the members of BoD, 
Constituencies and Staff participating to this GA.
> >I want to
> > thank Andy Gardner for teaching me something about IDN and
> > ascii issues that I didn't know...
>
> This is a subject very close to me and at present I am unable to comment too
> much due to an ISO NWIP for Internationalized Country Codes that is
> currently out for vote.  I am the Editor of the said NWIP and at this
> present moment it looks like it might fail which is a damn shame because if
> it does it is my belief that there will be mayhem if IDN TLDs are introduced
> piecemeal in an unstructured way.  I may comment further after the 10th.

This comment from Debbie Garside is important. As Debbie and I are in 
fact at the core of this issue, which affects the international 
structure and naming of the Internet, I think I should document this 
in more detail.
NWIP - ISO New Work Item Proposal

An ISO NWIP is a proposition for a new standard to be initiated. It 
is proposed by a country Member (ISO is not a UN body, but rather an 
association of national standardization entities, such as ANSI in the 
USA, DIN in Germany, etc.). A Proposed Project Leader is associated.
The text of the NWIP that Debbie refers to was worked out in Paris on 
June 8, 2007 (at an ISO/BSI/AFNOR/ICANN/Jefsey conciliatory meeting) 
in opposition to my suggestions preventing it from becoming (as 
expected by the ISO Central Secretariat in calling for this 
conciliatory meeting) a consensual and technically sensible document.
It is noteworthy that Debbie continues to claim that she is the NWIP 
Editor. I understand that she is claiming this in order to be 
entirely responsible for the British BSI proposition about making the 
"internationalization" of the namespace an official standard, as a 
continuation of the Unix/IETF approach. She is actually listed as 
Project Leader.
Technically, internationalization with the localization of the ends 
equates to the globalization proposition (cf. Unicode). It is, in 
other words, limited support for "non-ASCII script" languages in 
using English as a core and pivotal language (cf. RFC 3935). This is 
considered adequate only by English speakers. Apart from the obvious 
US strategic and commercial interests therein, it seems straight out 
of the RFCs, and it is said to be the maximum that Internet 
technology can deliver.
ICANN Position

I want to note that ICANN was represented by Bart Bowswinkle and Jaap 
Akkerhuis, and that their position was completely neutral.
They explained that ICANN had an internal process underway to 
determine a position, did not request anything from ISO, in turn 
respecting what the international standardization, user practice, and 
IETF could come up with. Due to immediate and persistent odd rumors 
stating that I would have supported the NWIP text, I had to to let 
everyone know in the following hours to that my agreement with the 
ICANN position.
I still support it, even if I disagree with most of the ccNSO/GAC 
reports: their propositions are not the final ICANN positions (at 
least as of yet) that I feel are much more realistic and in tune with 
the IGF common thinking. I also want to point out that I confused the 
propositions made by Debbie in Lisbon and her her claim that she was 
a Member of the ccNSO/GAC demonstrated her being supported by ccNSO. 
I was corrected by the ccNSO Chair and Bart.
Internationalization vs. Multilingualization

The technical inadequacies and limitations of this 
"internationalization" is the reason for the delays in linguistic 
domain names, the Multilingual Internet, the local sustainable 
development of e-commerce, etc. The reason is that it does not work 
well or easily, and it is not very appealing to users: 
internationalization results in violations of the WTC TBT rules 
(technical barriers to trade), linguistic Human Rights, GAC and WSIS 
principles; it does not match UNESCO campaigns to support cyberspace 
linguistic diversity, universal demands for multilingualization, such 
as Andy's last one, millennium objectives, $100 computer program 
specifications.
This does not mean that Debbie was wrong in supporting 
"internationalization". She actually represented the IETF and Unicode 
line of thinking followed by IETF for a decade, under the leadership 
of two Chairs (Harald Alvestrand [now an ICANN BoD Member] and Brian 
Carpenter) that were also Unicode BoD Members. She did this after 
having gathered a brilliant panel of experts 
(<http://thewldc.org/>http://thewldc.org) that I already documented, 
which includes leaders from the IETF BCP47 current process and ISO TC37.
"Multilingualization" means to be able to support every language just 
as English is supported today, in which internationalization makes 
way for every language to be limitedly supported as a localized 
version of an internationalized English.
A new change is that we know now that we can make multilingualization 
work much more easily than internationalization (cf. infra).
My suggestion after the AFNOR decision

What is now under discussion (of which Debbie refers to) is that the 
negative result of the NWIP vote (which has already been documented 
online by ISO) will only be announced next Thursday [4 week delay]. 
French AFNOR wants this to occur in this way in order to permit 
additional comments. I have, therefore, proposed for it to also be 
accompanied by an agreement from all the involved parties to jointly 
work on a new consensual NWIP that would be voted on as soon as possible.
For the international network and community, this would transform a 
negative response into a positive and consensual result by way of four points:
1. we accept that it has demonstrated a consensus between normative 
bodies and civil society, the business world, governments, and 
international entities (IGF) [this is a de facto rejected appeal of 
Debbie against the prior ISO vote of the ISO 3166-1:2006 standard, 
which is in this way fully confirmed].
2. that "internationalization" is not an appropriate final solution - 
but it is widely supported, and should continue to improve. 
Therefore, the priority is interoperability with every new emerging 
proposition - along with the WSIS Tunis agreement (which is the 
reason why the IESG was able to accept [just a few hours later] the 
publication of RFC 4646, without any international legal objection of mine).
3. that metalingualization - as per the ISO 3166-1:2006 paradigm and 
the IETF proposed script/language split - provides a good, simple, 
and stable basis to develop multilingualization solutions atop.
4. that we convene, under the authority of ISO 3166/MA, a working 
group including all the concerned parties (starting with ICANN, GAC, 
ccTLDs, Debbie's corporation, WDLC she is the CEO, MINC, UNESCO, ITU, 
IGF, IAB, IETF, ISO/TC37 and ISO/TC46, MLTF, AILF, Unicode, etc.).
It should write the new consensual NWIP (of which Debbie could remain 
the Project Manager).
Proposed priorities

I proposed that we should then pay special attention to:

- interoperability during the transition development and emerging 
multiple new services of this effort and the resulting Internet 
architectural dynamism that it should foster;
- immediate support of IDNs along GAC and IGF principles, equally 
respecting every national sovereignty and local community 
governance's decisions (especially for small ccTLDs),
- urgent work on local/lingual TLDs and distributed namespaces (in 
relation with the work that the MLTF [Multi Lingual Technical Forum 
that I moderate] has engaged in on semantic addressing),
- technical specifications [along the ISO definition] for the way to 
approach the ISO 3166-1:2006 paradigm.
Technical considerations

This is a very complex and architecturally new issue and I was 
largely opposed by the wide majority of the IETF.
The whole issue is complex because of the way the IETF and ISO 
considered it. From a historical progress point of view (Harald 
Alvestand's successive RFCs and Unicode development). Therefore, they 
never referred to it as part of a complete communication theory 
extended model.
Recently, John Klensin proposed an important change on 
idna-update@xxxxxxxxxxxxx (in June but not even yet worked on) to 
split the script and language issues. Until then, IETF was engaged in 
a _major_ layer violation [not yet acknowledged by many] between what 
belongs to:
- on the one side, to the physical transport and digital presentation 
of the content - what Internet and applications usually do (with the 
additional difficulty that TCP/IP has no presentation layer)
- on the other side, to its semantic understanding (what the carried 
content does mean).
Debbie's NWIP result brings a normative experimental confirmation to 
what TC46 has already applied in ISO 3166-1:2006, which was voted on 
one year ago; and what the IETF has only started to discover.
We can certainly understand an initial bitterness. However, she did 
believe my positions were to be opposed and tested. She did it: 
others only continued with their ad-hominems (the technical arguments 
of those having none) and exclusion. This is the correct way 
standardization has to work and I am glad she did it (however, it 
could have been done differently)
The consequences

It was perfectly correct to hesitate about all this and about exactly 
when to consider the transition. Confusion came from at least two 
main separated places. Understanding this might have a major impact 
on the Internet, its architectural culture, paradigm, and naming business.
(1) the DNS belongs to physical and digital addressing (the current 
IETF major issue) and to semantic addressing (the meaning of the 
associated mnemonics). That is the source of the ICANN mission creep 
(Trade Marks, UDRPs, IDNs) and of the IETF main-layer violation.
Semantics is out of the IETF scope and IAB opposes the very use of 
the "multilingual" term in RFCs. When considering today's importance 
of metadata and semantic issues one can better measure the importance 
of this clarification. This led me to create the emerging MLTF in 
order to support the Multilingual and Semantic Internet architectural 
and technical issues.
(2) there is a need for at least three different definitions of what 
languages are, at:
- ISO TC37 (where Debbie's ISO 639 related work belongs), which is 
interested in cataloging and describing languages in search engines, 
libraries, etc. However, that first definition is not yet given.
- Unicode, IETF, and ICANN, which should be interested in relation 
spaces and routing upon external formated language information. 
However, that second definition is not given either
- ISO TC46, which is interested in the way they are used. A language 
is implied by the WTC TBT agreement to mean: it is a human relational 
system that can be imposed without being (and that must be supported 
in order not to be) a technical barrier to trade.
These are three orthogonal things that the IETF confused in the RFC 
4646 debate: my tweaking of their text was to permit the resulting 
RFC 4646 to remain interoperable with the three of them. I hope the 
ISO TC46 confirmed clarification will now help ICANN, Unicode, IETF 
and possibly TC37 to better analyze what a language represents at 
their own individual layer.
Questions posed to this GA

I also hope that our past technical opposition does not prevent us to 
concert. I already proposed this when the NWIP was started, in which 
I was rebuked on this very list. If this is the case once again, the 
MLTF would only continue its Open Standard 3166-4 initiative, along 
the ISO 3166 rules.

(1) throughout this process the Internet was represented by ICANN. 
The ISO 3166/MA invited ICANN/IANA as a Member. They never attended 
meetings or participated until the last yearly meeting. At the Paris 
meeting, last June, ICANN made it clear that they had to comply with 
the market/users expectations [they cannot support an XY ccTLD when 
everyone around the world uses ZX as a country code for the concerned country].
Question: since GAC is only a representation of the richest 
countries, who already are members of the IGF multistakeholders 
governance, I wonder if ICANN should continue to be represented by 
the IANA or if it should be represented by an ICANN GA, or not 
represented at all but rather through a technical liaison?

(2) this process leads IAB and IETF to consider liaisons with ISO TC46.

Question: should ccTLDs and Users not also have such a liaison. 
Should it be through this GA, the ccNSO, and all-ccTLD constituency 
or dynamic coalition, a national Internet community representation, a 
whole ICANN GA?

(3) I am engaged in a negotiation with the ISO Central Secretariat 
over the protection of the ISO users' interest and their 
representation. ISO Members are in most cases regulated by national 
laws. However, there is no one to represent the international users 
of international norms and standards as such.
The special case of ISO 3166 is central, as it standardizes the core 
of the regalian attributes (sovereignty, country name, country code, 
non-TBT languages and script, administrative units, and [through ISO 
11179 compatible metadata registries] will extend its influence to 
all data having a national and lingual aspect. As such, it is said to 
be the Constitution of the virtual world.
Question: how could user control and contributions be organized? 
Should it be specific to the Internet (an Internet-GA?), should this 
GA be involved? Should an ICANN-GA be involved? or should this be 
through an ISO User Interest Group to create? Could this be an IGF 
Dynamic Coalition?

(4) GAC and WSIS principles state that a country has no capacity to 
influence the namespaces of another country. RFC 1591 says the TLD 
Manager is the trustee of the local community. This seems to 
contradict some of the GAC, ccNSO, ICANN, IETF conceptions and 
practices. The Internet RFCs consider the global and local Internet 
communities. WSIS believes that the Internet Governance is carried on 
a multistakeholder basis.
Question: should National Internet Communities not form an ICANN 
constituency parallel to the ccNSO, GAC, and BC? Should they not be 
involved in the ccTLD delegations, creation of lingual TLDs, etc. 
Should they not be involved in the IANA language and country databases?

(5) MLTF engaged work on semantic addressing and routing (the 
location of concepts, information, etc. and their relative meaning, 
in every possible language, technology, and services) tends to show that
- the DNS is a very small part of it,

- the most appropriate systems to resolve these addresses are DDDS, 
an IETF defined system family, which includes the DNS.
Question: this creates a need to coordinate Semantic Addressing, 
Trade Marks, cataloging, etc. governance and business management. 
Where should such an emerging issue be addressed?

(6) IDNs, and further on LDNs (local/language Domain Names), raise a 
large number of issues, in which none of the ICANN numerous 
committees have discussed or engaged in proper dialog with the 
concerned experts.
Question: should not GNSO/GA Members form an IDN-WG, and be welcome 
in every ICANN IDN related WG or SIG on a reciprocal basis?

(7) There is a need for an open, serious, measured, diplomatic, 
professional, competent, and representative mailing list to discuss 
"3166-4" (how to use ISO 3166 in a multilingual context).
Question: could this GNSO/GA mailing list be a candidate for such a 
task? Should it be a new specialized list? Would you like to 
participate in such a list? How should such a list articulate along 
with the GNSO/GA mailing list, a possible ICANN-GA and/or an 
Internet-GA mailing list.

(8) IPv6 will probably not make it. This means that address shortage 
will start hitting large and ISP projects in different ways. The same 
IDN and Multilingual Internet delays affect national or service 
providers projects. It is likely that new and multitechnology IP and 
Semantic addressing needs and solutions will emerge.
The priority is to maintain the end to end nature of the Internet and 
to foster the development of the brain to brain semantic 
interintelligibility relations. This calls for the interoperability 
of the new practices and standards that are being developed and of 
the emerging operational modes.
Question: where should such an interoperability be documented, 
maintained, and debated? Would this GA as a users' GA be a proper 
bootstrap for an initiative in that area? Should it be in cooperation 
with some Open Standard efforts? How should multilingual 
contributions be supported?

I thank you for your time and cooperation in reading, and possibly 
commenting on, this note.
jfc
I thank you for your cooperation in reading and possibly commenting this note.
jfc