Re: [idn] opting out of SC/TC equivalence

["James Seng/Personal" <James@Seng.cc>]

Hi Prof Tseng,

I think this TC/SC issues have been discussed again and again. The only
contention is of course how to solve the the problem. I hope we can
agree that so long this TC/SC is been handled, the exact mechanism is
open for discussion.

I spend several hours discussing with eric chen, wenhui, deng xiang and
others on this in London and I hope they have brought back my
suggestions to you.

My stand on this is as I stated during the meeting. Nameprep must be
uniform. There should be no optional folding which would be done in some
but not in others. Not every application of domain names have the luxury
of user interface or asking the user for information (note: email
servers, digital certificate authorization etc).

Therefore, if TC/SC is to be done in Nameprep, then I would prefer it to
be done across the board. But of course, this would means languages
which uses Chinese script (like kanji and Hanja) would also do the
same folding.

This is why I suggested that TC/SC to be done on zone file level and not
Nameprep. For those who knows Chinese, TC/SC is not really a 2^n problem
so lets not continue this misconception (It is closer to a n*2 to n*4
opertionally). And the memory problem could be simplified by ultizing
some algorithm (since it is fairly structured) to reduce the memory
usage. It can be done and it have been done. (We done it for 2^n btw
with reasonable memory use).

Of course, ultimately, the decision would still lies on the document
authors to come to an agreement on the approach and consensus in the wg.
But lets not forget why we are doing IDN in the first place so lets not
get too carried away in our discussion.

In the meantime, I would like to recommend the authors for tsconv to
complete the draft. The tables mapping is not complete. I would suggest
submitting the draft to Unicode Consortium or some other code point
expert group for review if possible.

Thank you.

ps: No, I did not attend Eric Brunner's UTF-8 dinner. Both Marc and
myself are tied in meeting prepration. We are interested in the
discussion tho so if anyone keep notes, please post.

-James Seng

----- Original Message -----
From: "tsenglm@計網中心.中大.tw" <tsenglm@cc.ncu.edu.tw>
To: "Edmon" <edmon@neteka.com>; "ben" <ben@cc-www.com>;
<idn@ops.ietf.org>
Sent: Friday, August 17, 2001 12:15 AM
Subject: Re: [idn] opting out of SC/TC equivalence


> Edmon:
>             I will try to give some explaination althought my english
is not
> good. I can told to you straightly that if many Chinese people
(including
> Taiwan, HongKong & Mainland China)see your example and wishness, they
will
> be very angry to your description. Why ?
>             In Twaiwan,  HongKong we use BIG5 code and many simplified
> scripts can not be key-in and displayed in their system. Many people
they
> know traditional chinese characters but they don't know its
corresponding
> simplified characters.  Reversely, in mainland china people only know
> simplified characters. The corresponding TC/SC has the same
pronunciation
> and meanning.  Todday people want to communicate each other, they
regist a
> domain name for some trade mark reasons , but the ML.COM system can
not
> provide the same chance to protect their rights. People in Taiwan
regists a
> TC name but he can not key-in to regist the corresponding SC name
because he
> even don't know how to do it . Then one day , some people in mainland
china
> they regist the corresponding SC name in SC.COM , it is the same
meanning
> and prouncing in TC.COM . What will be happen ?  You know that PRC and
> Taiwan , even HongKong are different goverenment ,  the argument in
> intellegent right is a big trouble.  The case is very like that you
allow
> one people to regist ABC.com and also allowed another people to regist
> abc.com and told them  that is an option and chance . Are you like
english
> are treated in this way ? Why case folding is needed ? Only the
history
> reason ? Why full case alphabet must be mapped to ASCII ?
>              Are you like to make trouble to us ?  That is why
department of
> China  information industry against chinese.COM so deeply.
>              I don't care the mixing of TC/SC , but I care why we are
forced
> .....
>
> L.M.Tseng
>
> ----- Original Message -----
> From: "Edmon" <edmon@neteka.com>
> To: "ben" <ben@cc-www.com>; <idn@ops.ietf.org>
> Sent: Thursday, August 16, 2001 10:29 PM
> Subject: Re: [idn] opting out of SC/TC equivalence
>
>
> > Hi Ben
> >
> > From: "ben" <ben@cc-www.com>
> > >
> > > "ColorCentre.com" and "ColourCenter.com" is ok in English.
However,
> > > "<tc><tc><sc><sc><tc><sc>.com" is a recipe for disaster.
> >
> > So, can you come up with one good reason why BE+AE (British
English -
> > American English) is OK and TC+SC is not?
> >
> > I have seen quite a number of brand names now beginning to have
TC-SC mix
> in
> > HK.  My point is that TC and SC are no longer simply a straight
mapping of
> > each other, or in technical terms a compatible character or simply a
> > different script.  Each character in the Chinese repertoire is
becoming a
> > character of its own right, the line between TC-SC is starting to
blur.
> > They are all "Chinese Characters".
> >
> > >
> > > From: "Edmon" <edmon@neteka.com>
> > > > Hi Ben,
> > > > Presently, I can have:
> > > >
> > > >    ColorCentre.com
> > > >
> > > > and someone else can have
> > > >
> > > >    ColourCenter.com
> > > >
> > > > I am mixing British-English and American-English.
> > > >
> > > > Are you saying that these two domain names should be mapped to
> > > >
> > > >    ColourCentre.com
> > > >    ColorCenter.com
> > > >
> > > > and all should be the same domain?
> > > >
> > > > From my perspective, this is the issue we are facing with TC/SC
> > > > mapping.
> > > > I respect that some registries and domain operators wish to have
> > > > TC=SC, but
> > > > I think, as a consumer, I wish to have the option to treat them
> > > > differently.
> > > >
> > > > Edmon
> > > >
> >
> >
> >
>
>