<html>
<body>
At 14:11 09/09/2010, McTim wrote:<br>
<blockquote type=cite class=cite cite="">On Thu, Sep 9, 2010 at 2:47 PM,
jefsey <jefsey@jefsey.com> wrote:<br>
Why would Google want this? or are you talking about the names n<br>
numbers? Why would Google want that? Far too much headache
for far<br>
too little ROI IMHO. They make Billions per quarter, DNS isn't<br>
"profitable" compared to search,etc that Google
does.</blockquote><br>
Tim,<br><br>
Sorry, but this calls for some explanations that you probably already
know, but that others may not. I will try to be as concise as I can.<br>
<br>
The current Internet is about English ASCII information dissemination
(passive content) with an evolution towards UTF8. The equilibrium that we
currently have will have to adapt to a full UTF8 support (naming and
content) of the linguistic diversity (which is also economic diversity).
Every information provider and naming administrator has a need,
therefore, to adapt to IDNA2008 and multilingualism. <br>
<br>
<br>
<b>1. ICANN<br>
<br>
</b>ICANN's job is to manage a small Excel table. They have always had to
be good enough to make believe that this depended on complex unnecessary
contracts to create and manage scarcities (TLDs, IPv4). They have
intelligently started first taking control of the IANA (focal point),
then in launching the TLD (linguistic groups of servers) saturation
proposition, and in pursuing its make believe "I am the center of
the world" permanent strategy with DNSSEC.<br>
<br>
At the same time, they made a mistake. To start Fast Track quickly
enough, they privileged IDNccTLDs over IDNgTLD projects. It is in this
way that they made a lot of money to bepotentially available, ready
to support any innovation that bypasses their blocking
"monopoly". They also made it impossible for local/cultural
non-profit organizations to plan a TLD. This wqy they opposed the planned
grassroots Multilinc and Projet.FRA tests.<br>
<br>
<br>
<b>2. GOOGLE<br>
<br>
</b>The Unicode consortium (mainly IBM, M$, Apple, Oracle, Yahoo!, and
Google - you could initially read this list in decreasing order, but now
in increasing order) has the internationalization strategy. <br>
<br>
<u>2.1. What is internationalization?<br>
<br>
</u>With the world's globalization, there is an accentuated dichotomy
between norms and standards. Until now, they have gone together as US,
French, European, etc. norms/standards, the norms being a concept for the
description of the local normality and the standard a concept of what we
want to make with this normality locally. Now, the norms tend to become
the global normality, and international standards infiltration is the way
to conquer foreign markets with the standards of national industries.
Internationalization (of national standards) is WW III. <br>
<br>
Until now, war was to plundering other countries to resolve sovereign
debt problems. Today, this is no longer possible when the wealth of a
country is its stable just-in-time production and services and debts are
interrelated. Therefore, what countries look for is a higher growth ratio
for their exports than for their debts so that they are then considered
as AAA++. The same is true for leading industries (e.g. Google). <br>
<br>
<a name="_msoanchor_1"></a>Standardization is a powerful and stable way
to obtain this by forcing other countries to run their just-in-time
processes with nationally compatible solutions and <a href="??.htm">not
overly innovative </a><a name="_msoanchor_1"></a>[MHL1] products. This is
why Mark Davis' (President, Unicode) "globalization" is a key
element in the Unicode consortium industrials growth as well as USA's
growth (FYI globalization = internationalization of the medium +
localization of the ends + filtering of the languages - on the Internet
and in the information centers). <br>
<br>
2.2.<u> Natural strategy<br>
<br>
</u>First, IMHOthere isn’t a finalized strategy offered by anyone.
However, there are natural trends and common sense attitudes. This is
either because people saw, see, or shall eventually see that things fit
better for them in one way more than another.<br>
<br>
The Internet is the internationalization's main medium: they must keep
some control of it. The IANA is its information core and RFCs are its
guidelines. However, the IANA's exposure makes it unadvisable to visibly
control it. Therefore, the Members of the Unicode consortium are/should
be happy:<br>
<br>
1) That ICANN is believed to be the one in charge (IETF actually is of
most)<br>
2) They have overloaded the IANA with the language tables (RFC 4645)<br>
3) Which may require an IANA computer response that only Google could
sustain (they openly suggested that they could provide the machines)
should these tables become used in the present state of Mark Davis' RFCs
(5646)<br>
4) i.e. permitting traffic and search service control on a cultural
(hence economic) basis - network neutrality is nothing vs. service
neutrality.<br>
5) They control the code of Mark Davis' ICU routines, which in turn
control that use: they are universally deployed in protocols and
languages to interoperate with Unicode information.<br>
<br>
De facto, through the IANA, the Internet can be DoS-ed by Mark Davis.
Just one line added to the ICU code, to update local user linguistic
tables on a regular basis, and the IANA would dump 90% of its content
into each user with a new ICU. This would represent a progressive (as
programs get updated) network overload. All the same, Google + Public DNS
can represent the largest resolution system. Yet it has to be protected
from networked alternative resolution habits, the DDDS possible evolution
(extension of the DNS type of database to other issues) and ISO 11179
metadata registry norms. <br>
<br>
Mark Davis, was with Apple, then IBM, and now Google, as is the case of
Harald Alvestrand, former IETF Chair, Member of the Unicode Board, Member
of the ICANN board and their Internet, VP, Vint Cerf, founder of ISOC and
Chair of the IETF WG/IDNABIS, which had to produce IDNA2008.<br>
<br>
<br>
<b>3. Internet Users<br>
<br>
</b>In all of this, the milked users have never been considered, except
through "the Internet for all" slogan, user's purse centric
doctrine, and democratic liberty campaigns to "free" them from
their national protections. The world information system is planned to be
centralized by Google. With the current IANA and DNS, this was in fact
possible.<br>
<br>
This has changed this year, however, with the adoption of RFC 5890's
suites documenting IDNA2008. For various reasons, IDNA was to be
separated from Unicode (including the changes in namespace due to new
Unicode releases). This was a strategic thread for Unicode leading
members. Maybe that is why the WG/IDNABIS mailing list belonged to Harald
Alvestrand, Mark Davis was the only one protesting at the end, and Vint
Cerf was the Chair. Everything was tightly locked.<br>
<br>
The reality is that they had bet on the wrong horse: Unicode cannot
fulfill the job. The reason why that is the case is that Unicode is a
typographic solution, for scripts. The need is for an orthotypographic
solution, for languages. The typical trouble is the French (Latin)
majuscules and their needed support by Projet.FRA. Because
"Etat.fra" means "State" and "état.fra"
means "status". Equivalent issues exist in every script.
French, and to a lesser extent German, Greek, Iranian, Tamil, etc.,
people attended the WG and a full Arabic mailing list watched
closely.<br>
<br>
Unicode has no way to support the "majuscule" or other
orthotypographic metadata information. Therefore, there would be no way
to support the Intersem (semiotics and semantics) and the semantic
addressing system (SAS) through the DNS, that is, if IDNA is controlled
from within the Internet by a unique RFC for all kind of uses by the
linguistic diversity. This is why there was a need for a multi-layer DNS
encapsulation that is specified on the user side. This is what I
initially proposed and that they had hoped that they could avoid.<br>
<br>
The result is what IAB/IETF wished for and what French/Latin language and
many other languages need: no character mapping within the Internet,
hence no interference by RFCs and IANA, and hence by paying ISOC
platinum/golden sponsors in languages and cultures. This means that
control of cultural expression in coherence with search engines (as
prepared by Mark Davis' RFCs) had to be abandoned by Vint Cerf. Google
may have to quote domain names that they do not know how to resolve
without also having to respect local, cultural, national naming standards
and UTF8 resolution processes. Men are not Google peripherals.<br>
<br>
A patch was found that permitted a consensus to be reached: the necessary
preparation (and mapping) of UTF8 entries will be made on the user side.
This was not a big practical issue for Unicode since their Members
already distribute users’ applications and libraries all over the place.
Moreover, Google Public DNS received publicity at that time: the size of
Google's own DNS service is such that they can become a de facto
operational reference. Example: If they introduce ".google" (or
any other TLD) they will go through; Vint Cerf has already played that
game when he adopted a new ".biz" against the existing one.
<br>
<br>
However, IDNA2008 cannot work (except in some default situations) in the
way most believe it can, i.e. as documented by the IDNA concepts.
IDNA2003 was specified on the Internet and on the user's side. There is a
major need to standardize IDNA2008 on the user's side as well, with a
non-IDNA (IDN in application) user side achietcture. IAB has documented a
few good reasons as to why that is in a soon to be published RFC. <br>
<br>
The only existing proposition is ML-DNS. As a precaution, I asked
everyone(WG, AD, IESG, and IAB) if they wanted to take over its
investigation and documentation. I am now left with the hot potatoes. IAB
will only further discuss the problem...<br>
<br>
<br>
<b>4. Present situation<br>
<br>
</b>ICANN owns the IANA and the root system. Google has the power to
influence the IANA and the root file and replace the root system in many
people's lives. IUsers do not need any of them, and the possibilities
open to them are very large. However, there is a need for experimentation
first (what we investigate with an Intertest project using the Internet
as its own test-bed (as proposed by ICANN in its ICP-3 document).<br>
<br>
Everyone wants stability + something contradictory with the others. ICANN
wants to retain control of naming; Google wants to retain control of
information processing. IUsers object to control except for in
identificative naming (vs. designative and appellative), but they would
like for it to be nationally controlled.<br><br>
jfc<br>
</body>
</html>