Re: Inputting mixed SC/TC (Re: [idn] A question...)

["Chun-Hsin Wu" <wuch@gate.sinica.edu.tw>]

From "James Seng/Personal" <jseng@pobox.org.sg>
> 1. Lets do away with the personal attack on me. It does nothing to help
> your arguments.
Agree. I'm not interested in personal attack.
 
> 2. I acknowledge I am weaker in my daily chinese compared to people in
> China/Taiwan, this does not mean I do not know the chinese script and
> its orthography. And no, I understand you perfectly well. But I doubt
> you understand (or hear) me - See note 6.
I'd like to learn more experiences from you. :-)

> 3. Have you speak to your IRG and SC2 representives lately and ask them
> about this problem? What is their unambigitous response?

There are differences among name conversion, unicode conversion and
DNS identifier conversion.

> 4. I stand correct in my statement last mail: You can input either TC or
> SC by manually scrolling down the system. Your screen shot represent
> nothing. Different IME will give you different predictive result
> depending on the frequency of a personal usage of the characters.
> Therefore, a screen shot on your computer would be totally different
> from mine.

At least the IMEs used in my last examples do not depend on the
frequency of personal usage. If you do try these examples, you should
be able to see the same screent shot as mine.
This is one of the design points for IME. Other IMEs may depend on
personal usage, but the demonstrated IMEs don't.

> 5. You have only repeated problems that we already knows. You have not
> demonstrated any solution which is technical possible now. This is not a
> 3 months or 30yrs plan thing. It is about what is technically feasible
> and what is not.
Hope IDN friends do really understand the problems.

> 6. Maybe working for another 30yrs time, we might come up with an AI
> that can do TC/SC, and for all the other language issues. But it appears
> you and your colleague continue to ignore (something we been telling you
> again and again):
>     1. DNS is an identifier system. We deal with scripts, not languages.
>     2. Human language is complex. Expect DNS to deal with language
>             is technically impossible right now. (Proof me otherwise).

I think IDN is also complex. :P
Hopefully, we don't expect language-level conversion or complete
Unicode conversion, what we need is identifier-level conversion,
for identifying domain name rather than solve all the conversion issues
in Unicode or human-space.

> 7. . Since TWNIC is so insistent on the solution, I suggest TWNIC (or
> anyone else) put a total ban on all chinese characters in the
> registration system. That will also solve your so-called "2^n" problem
> in your system. Then continue to work with IRNSS (or some other areas)
> for a solution you can accept.
Is it a personal or organizationall attack?
I'm a researcher and I do not stand for TWNIC.
I expect a global unique IDN, not a fragmented DNS nor Internet world.

> 8. Some suggestions: If you want other to consider your issues
> favourably, then
>     1. present your arguments unbiasly.
>     2. consider a broader scope of the problem
>     3. understand other people issues and willingness to compromise
>     4. learn how the process works in IETF
> So far, I have seen all otherwise.

Personal attacks again?  :P

Chun-Hsin Wu