Re: [idn] Traditional-simplified, yet again

["Tim Langdell" <tim@xtns.net>]

As many of your already know, we here at XTNS are proponents of a layer that
sits on top of DNS because we feel there are many issues and potential
future features of IDN resolution that are better handled (at this time,
anyway) via such a layer than seeking to devise and integrate solutions into
the root. And indeed, our XTNS System does just that, enabling IDNs to be
"resolved" within the Address Bar of Internet Explorer, and then being
translated -- if desired -- to ACE to be routed to the root, or simply at
this time to be resolved directly to specific URLs or IP addresses.

It seems to us that this issue of TC<->SC mapping is  a perfect example of a
problem that is perhaps best dealt with at this time in the layer that we
provide sitting on top of DNS, rather than perhaps seeking an elusive
'perfect' solution in order to integrate such mapping into the DNS itself at
this time. The power of dealing with the TC<->SC issues in the layer that we
provide is that it can be dynamic (allowing for evolution of TC<->SC mapping
in the future) and can enable the on-going development of a solution that
can eventually be integrated into the root system for all IDNs. Our layer is
perfect for dealing with issues that are not simple one-to-one mappings, and
where context dictionary based problems are to be dealt with.

Moreover, as our system also handles Japanese.Japanese, Korean.Korean, etc,
we can equally coordinate other mapping issues as and when they arise such
as the example given by Mark:

>If there were a recognized 1-1 subset mapping between SC characters and TC
>characters, I don't think unification would be an issue in matching. For
>example, it would matter little to Japanese users that an SC character
could
>be typed in instead of one of their TC characters, so long as two different
>TC characters used in Japan were *not* matched. This could be done as long
>as the mapping never identifies two different TC characters with the same
SC
>character.

Were there a situation where two different TC characters really "ought" to
be mapped to the same SC character (and I do not know if there would be)
then here would be a perfect example of an issue that can be easily handled
via the XTNS System sitting on top of DNS, but is complex to even consider
resolving within the core IDN system itself. One to one mapping could be
eventually integrated into the root, leaving the exceptions and complex
translation issues to remain to be dealt with in the layer sitting on top of
DNS.

Tim

Dr. Tim Langdell
CEO
XTNS

----- Original Message -----
From: "xiaodong lee" <lee@whale.cnnic.net.cn>
To: "Patrik Fältström" <paf@cisco.com>; "ben" <ben@cc-www.com>;
<idn@ops.ietf.org>; "Paul Hoffman / IMC" <phoffman@imc.org>
Cc: <jet-member@nic.ad.jp>
Sent: Tuesday, October 30, 2001 1:02 AM
Subject: Re: [idn] Traditional-simplified, yet again


> Dear Patrik:
>
> Everyone is tired in IDN, you and all. We do not want to make this
> in the truth. We brought up such problem, but it is a difficult problem
> to be solved based on so-called principles said by someone of IDN.
> So we must prompt new solution to solve the big problem for Chinese
> people again and again, which is not our aspiration. Who want to do that?
> not you, and not us too.
>
> You said some authorities on Chinese scripts think it is a complex
problem,
> I think so, CNNIC people and all Chinese people think it is not easy to
solve
> too.
> To do TS/SC in context is a difficult problem, it should not be solved in
DNS.
> but any difficult problem could be reduced in some case, and based on some
> priciples. DNS is identifier, so we solve part of TS/SC problem in DNS,
which
> will solve most of the TS/SC problem.
> I don't know how you explain such problem for those authorities, I still
think there
> are some misunderstanding in such problem, we have explain the special
> problem of TS/SC for our chinese scripts expert which is author of the RFC
about
> Chinese character, he agree with us too. we must differ this conversion
problem
> from other TS/SC conversion, which I have explain in my draft, and will be
explain
> again in our new draft that will be submitted soon after.
>
> So I think you might read our draft carefully and understand our question.
> maybe I cannot explain it in my draft for you for my poor English, but I
do want
> to discuss this with you  in detail.
>
> let's be more patient on such problem and more progress for IDN too. OK?
>
> Xiaodong Lee
>
>
>
> ----- Original Message -----
> From: "Patrik Fältström" <paf@cisco.com>
> To: "ben" <ben@cc-www.com>; <idn@ops.ietf.org>; "Paul Hoffman / IMC"
<phoffman@imc.org>
> Cc: <jet-member@nic.ad.jp>
> Sent: Tuesday, October 30, 2001 2:01 PM
> Subject: Re: [idn] Traditional-simplified, yet again
>
>
> > --On 01-10-29 21.56 -0500 ben <ben@cc-www.com> wrote:
> >
> > > How is it
> > > possible for people to delay the progress of the IETF by simply
> > > posting up messages and writing drafts?
> >
> > By
> >
> > (a) Talking about an important issue
> > (b) People tell the author what part of that issue can be solved
> >     in the DNS, and how, and point to other solutions for
> >     the other part(s) of the issue
> > (c) Talking about the important issue again, with no changes
> > (d) People tell the author again that things have to be changed
> >     because the issue can not be solved in DNS
> > (e) Go to (c)
> >
> > Myself, I have with the CNNIC people now passed point (e) four(!) times,
> > and I start to get tired on this. The proposals from CNNIC have not
changed
> > at all since (a), and conversations have occurred on the IDN mailing
list,
> > in private conversations (one is going on at the moment) and in face to
> > face meetings.
> >
> > Further, at second pass of (e), myself and others talked with the
> > authorities on Chinese scripts we know to verify that we were not stupid
or
> > misunderstood something. _ALL_ of these said that it was the CNNIC
people
> > which had not understood the complexity of the problem. Further, the
> > Unicode Consortium have already been through the SC/TC problem once, and
to
> > explicitly state this fact, the Unicode Consortium explicitly sent a
liason
> > statement to the IETF stating this fact.
> >
> > When all of these things was presented to the CNNIC people, the rough
> > consensus in the wg on what to do, the information from external
sources,
> > and the liason statement from Unicode Consortium, the response is:
> >
> > - The wg don't understand the problem
> > - The external sources don't understand the problem
> > - Unicode Consortium don't work on these issues
> > - ISO have nothing to do with Chinese Characters
> >
> > And we go back to (c) again from (e).
> >
> > Sorry, but this is absolutely NOTHING but a delaying process from
ignorant
> > people.
> >
> > Note that I am nowhere above saying that SC/TC is a problem. Just like
> > other equality (define the word equality whatever way you want) when you
> > use more than one script -- and in some cases even with one.
> >
> > I am as you can see extremely tired on this wg not making process just
> > because one group of people is blocking the discussions.
> >
> > This wg is not making any progress.
> >
> > Absolutely NOTHING has been changed in the proposals which are on the
table
> > the last couple of months, and as in all processes, we _could_ have been
> > able to discuss things which we still have to polish here and there. For
> > example, the part of the SC/TC issues which have to do with mappings
from
> > local charsets to Unicode, use of Unicode in other protocol elements
than
> > domain names etc.
> >
> > So, at the moment it is CNNIC which effectively see that the world can
not
> > use Chinese Characters in the DNS.
> >
> >      paf
> >
> >
>