Re: [idn] Re: Chinese Domain Name Consortium (CDNC) Declaration

[Kenneth Whistler <kenw@sybase.com>]

Janming said:

> Two comments:
> 
> 1. The TC/SC problem or variant problem comes from Unicode 

It does not.

"The TC/SC problem" derives from 50 years of political history in
China, where the PRC engaged in an orthographic revision of
the writing system that Taiwan (and Hong Kong) did not, during
the same time span that information technology was developed and
rolled out on both sides of that political chasm. The result was
the development of incompatible character encodings and
input method editors tailored either for the "simplified" Chinese
of the PRC or the "traditional" Chinese of Taiwan and Hong Kong.

This was not a "problem" to end users before the Internet, because
never the twain did meet, and computer technology generally
didn't have to support both, but only one choice at any one
site. Furthermore, traditional IME's didn't support entry of
simplified characters, even if the traditional character encodings
(like CNS) nominally had them in the list; and vice versa.

Now it *is* a problem to end users, because the Internet and
the web, and changed political conditions, and globalization
mean that everybody has to deal with all of the characters in
one context or another.

I would accept a reformulation that:

  A. The availability of Unicode-enabled systems, such as
     Windows 2000, with IME's that can support either traditional
     or simplified characters, and fonts that display both,
     has made the preexisting TC/SC problem acutely visible
     to end users on their computers.

  B. The move to Unicode implementations means that mingling
     of traditional and simplified orthographies is easier.
     In effect, users now have the rope to hang themselves,
     if they so desire. Whereas, before, the constraints of
     the deployment of IME's and fonts generally meant that
     you couldn't easily mix SC/TC, even when the code page 
     nominally supported it.

> and the
> decision
>    to adopt Unicode as the basis to develop IDN technology.

And the alternative is ... ?

What character encoding alternative at this point in history
would not have the exact same problem? Certainly no choice
of a current national character encoding out of China itself
would avoid the exact same problems you attribute to Unicode.

But you are going to have to pick *some* encoding for the
characters you represent in domain names. If not Unicode,
then what?

> If the
> working
>    group chooses to use Unicode as basis, then it becomes a partial
> solution 
>    to the "global" IDN problem because of this side-effect.
> 2. We do not want to delay your process either. Please refer to the
> draft
>    "Phased Implementation of Internationalized Domain Names in
> Applications" 
>    for more details.

I have looked at that draft in detail, and see no solution
offered -- but only denial and vain hope. In short, a proposal
to simply exclude *all* the Chinese characters from IDN, with
a vague promise that someday someone will figure it all out
and decide which of 20,902 characters are o.k. to let back in
and which are only variants of the "good" ones. That promise
will be dashed on the rocks of the reality of complexity in
Chinese character variation -- and no code point based solution,
Unicode or otherwise, is going to fix it.

And in the meantime, the Phased Implementation proposal merely 
exports the Chinese SC/TC "problem" in a way that also disables 
IDN for Japan. I don't think that is acceptable.

--Ken