[LINK] Dot Asia a good idea?

[Kim Davies]

Quoting David Lochrin on Tuesday October 09, 2007:
| 
| > This is why there is a desire to introduce non-Latin based 
| > top-level domains. .测试 and .測試 amongst others will be 
| > introduced into the DNS root zone in just a few hours 
| > as the first step toward this.
| 
|    Don't leave us in suspense - what are the .测试 and .測試 domains?

They are the terms 'test' translated in Chinese (represented in
Simplified Chinese and Traditional Chinese respectively). There are
actually eleven such domains:

    .إختبار  (Arabic)
    .آزمایشی (Persian)
    .测试 (Simplified Chinese)
    .測試 (Traditional Chinese)
    .испытание (Russian)
    .परीक्षा (Hindi)
    .δοκιμή (Greek)
    .테스트 (Korean)
    .טעסט (Yiddish)
    .テスト (Japanese)
    .பரிட்சை (Tamil)

|    Why cannot there be semantic equivalents of registered names, such that translations within a given language domain use one and translations outside that domain use the Latin alphabet as a primary & default language?  (I think domain names could probably not be in mixed languages.)  And what do Arabic- or CJK- speaking regions do now?

I'm not sure exactly what you are asking, but here are some of the
fundamental issues pertaining to having ccTLD equivelants in non-Latin
scripts.

Firstly, ccTLDs today are largely non-semantic. "AU" means Australia
and "KP" means North Korea only by association, not because it is inate.
However, the concept of these abstract codings in other languages is not
existent. Some languages have no concept of abbreviations and can not
deal with having a .AU, therefore by necessity it would need to be
the translated equivelant of .AUSTRALIA etc.

Once the codes are semantic it introduces a whole new set of issues.
There is no internationally recognised list of translations of country
names. There are, for example, two different geopolitically entities
that claim the right to be called "China". Similar issues exist in other
parts of the world.

Another issue is that ISO 3166-1 allows domains to be guarded against
collision due to its finite set of possible codes (always two letters,
never more, never less.) There is no risk there will be a "COM" code
in the standard, and this allows future country revisions to be
accommodated and for countries to be guaranteed a right to their domain.
It is hard to work out a universal standard that can apply to scripts
that aren't expressed with letters, and without such a system you will
have a real problem as new countries are created in the future.

There are a lot of languages, and quite a few scripts, and a universal
list of codes is unfortunately nowhere near realistic at this stage.

|    This sounds to me like ad-hoc power-based rule making.  If the structure of domain naming is not properly architected the whole thing will eventually self destruct.

Well, many argue that the taxonomy is not really that important, and
that people don't intuit domains as they would have 10 years ago. I
think my key concern is that if there is an explosion at the root zone,
it centralises alot of things that are decentralised today, and is
inherently less scalable.

kim