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

[=?big5?B?dHNlbmdsbUCtcLr0pKSk3y6kpKRqLnR3?= <tsenglm@cc.ncu.edu.tw>]

----- Original Message -----
From: "Edmon" <edmon@neteka.com>
To: "ben" <ben@cc-www.com>; "Eric Brunner-Williams in Portland Maine"
<brunner@nic-naa.net>
Cc: <idn@ops.ietf.org>
Sent: Wednesday, August 08, 2001 5:24 PM
Subject: [idn] opting out of SC/TC equivalence


> From: "Eric Brunner-Williams in Portland Maine" <brunner@nic-naa.net>
> > What is the key value for users "opting out" of SC/TC equivalences?
>
> There are two main scenarios that I could think of that I would like to
see
> SC/TC opted out.
> Say I wanted to have different online identities in the PRC, HongKong and
> Taiwan with two Chinese charcters.  And I want to use:
> <SC><SC> in PRC
> <SC><TC> in HK and
> <TC><TC> in Taiwan
>
> I want them to resolve to different places. If SC/TC is folded, I will not
> be able to do so.  Which means if I have an employee John in both the PRC
> and Taiwan, I could not use John@<SC><SC> and John@<TC><TC> at the same
time
> and use it for different person.
>
> Scenario two:
> If I have a franchisee/partnership operations in the above three regions
and
> the ownership of each is very different, but we want to maintain the name.
> We wont be able to use the SC/TC/mix to distinguish our domains.
         In  HongKong, people use BIG5 code that is major in traditional
chinese characters,
After 1997,  Hong-Kong Goverment announce an extension set including fonts
of TC/SC.
MicroSoft provided them with GBK code but display the message in TC font.
GBK is the extension of GB, and  China use them in MicroSoft IE now.  GBK
include the font of some traditional chinese character.  The TC/SC
translation  majorly focus on BIG5 <-> UNICODE <-> GBK , especially for BIG5
<--> GB/GBK.
         If TC/SC  one to one  font/script mapping is fiexed, it shoud be
processed in normalization/unification phase .  But  one to many font/script
mapping can not be easily solved by this way,  if the mixing type
<SC><TC><SC><TC>  are all one one mapping in each character then they will
be reduced to one canonical  representation.  If the mixing type including
one to many characters  one string are assigned by the registrant in
registration time.
         That draft is just to solve your 2  problems.  Hope these can help
understanding.
>
> Edmon


L.M.Tseng