[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [idn] Proposal for creation of new gTLD for IDN

To: "A. Vine" <avine@eng.sun.com>
Subject: Re: [idn] Proposal for creation of new gTLD for IDN
From: Mark Davis <markdavis@ispchannel.com>
Date: Fri, 07 Jul 2000 11:15:25 -0700
CC: Deuk-Kul Jang <dkjang@smind.co.kr>, idn mailing list <idn@ops.ietf.org>
Delivery-date: Fri, 07 Jul 2000 11:15:14 -0700
Envelope-to: idn-data@psg.com
Reply-To: Mark Davis <mark@macchiato.com>

The proposal requires an ordering of all of the subject letters, which means that they are
*not* simply allocated according to Unicode. According to the proposal, it requires a
compaction of each of the characters for the language. Presumably this would be still in
Unicode order, but would start at some arbitrary point and remove holes. And the values for
German could be different than French, Italian, Spanish, etc.

While this proposal would offer more compaction than a code-based approach, in
implementation, this would require a table per language (or language/country); at least
several hundred, with updates for languages coming in periodically (since they couldn't all
be covered the first time around).

That makes this approach a bit scary. Even if an algorithmic approach is more complicated,
it doesn't have the problem of handling multiple tables.

Mark


"A. Vine" wrote:

> At the very least, if the presumption is that the domain name is in a particular
> language, then the 2-letter language code should be used, not the 2 letter
> country code:  ko = Korean, ja = Japanese but kr = Korea, jp = Japan.
>
> But I find this proposal inappropriate for IDN.
> --
> Andrea Vine, avine@eng.sun.com, iPlanet i18n architect
> "Stew my foot and call me Brenda"
> --From "Angry Kid" by Aardman Animation
>
> Deuk-Kul Jang wrote:
> >
> > Dear Mr. James Seng.
> >
> > Thank you for your help. you are right.
> >
> > Dear John.
> >
> > Let's assume that two languages, Korean and Japanese, are spoken in your
> > country, and +ACI-.kr.icom+ACI- +ACI-.jp.icom+ACI- are allowed respectively.
> > Then you can have Korean domain of +ACI-----.kr.icom+ACI- by registering in Korean,
> > and Japanese domain of +ACI-----.jp.icom+ACI- by registering in Japanese.
> > And to your monitor, xxxx.kr.icom will be displayed in Korean,
> > +ACI-xxxx.jp.icom+ACI- in Japanese, and the others in English.
> >
> > If you want all kinds of domain to be displayed in original language, you
> > can do with a proper application, I think.
> >
> > sincerely yours
> >
> > -----+xtC8+A- +ulTC3MnA------
> > +vPSwuA- +wKy3jA-: James Seng +ADw-jseng+AEA-pobox.org.sg+AD4-
> > +vBuylA- +wKy3jA-: John C Klensin +ADw-klensin+AEA-jck.com+AD4-
> > +zDjIcA-: Deuk-Kul Jang +ADw-dkjang+AEA-smind.co.kr+AD4AOw- idn mailing list
> > +ADw-idn+AEA-ops.ietf.org+AD4-
> > +sKDJ3A-: 2000+sUQ- 7+xtQ- 7+x3w- +rgjGlMd8- +xiTWxA- 12:08
> > +yBy6qQ-: Re: +AFs-idn+AF0- Proposal for creation of new gTLD for IDN
> >
> > +AD4-Okay, let me be the devil advocate again, this time, trying to defend
> > +AD4-the draft :) No, this does not mean I like the draft. Lets just say it
> > +AD4-is for academic reasons...
> > +AD4-
> > +AD4-Back on the basic of Jang's proposal. One of the merits Jang bring to
> > +AD4-the table is that his encoding format are more compact. The reason it is
> > +AD4-able to be smaller because it makes certain presumation on the domain
> > +AD4-names, namely, you dont mix different language in a single label. It
> > +AD4-furthers compact it by using a base36 rather than the commonly used
> > +AD4-base32.
> > +AD4-
> > +AD4-(CIDNUC/RACE compression also make the same presumation, compression
> > +AD4-occurs only when the labels are from the same script 256 blocks so I
> > +AD4-guess you can say CIDNUC also make this presumation indirectly).
> > +AD4-
> > +AD4-Hence, Jang's requires an tag to identify the language/script use (just
> > +AD4-like CIDNUC use the first 8bits specifying the 'block' or D7 otherwise).
> > +AD4-His proposal is to use kr, jp etc as a tag per label.
> > +AD4-
> > +AD4-John C Klensin wrote:
> > +AD4APg- I'm not sure I fully understand it, but this proposal seems to
> > +AD4APg- confuse +ACI-language/encoding used+ACI- with +ACI-country in which language
> > +AD4APg- is spoken+ACI-.  By doing so, it causes:
> > +AD4APg-
> > +AD4APg-         +ACo- problems for users of other (presumably still
> > +AD4APg-         non-ASCII) languages in that country,
> > +AD4APg-
> > +AD4APg-         +ACo- worse problems for countries with several
> > +AD4APg-         official/national languages,
> > +AD4-
> > +AD4-Jang's email suggested having TLD kr.icom, jp.icom, cn.icom etc. This
> > +AD4-add another presumation that the domain name itself is one
> > +AD4-language/script (not just per label). Note that kr.icom, jp.icom does
> > +AD4-not mean korea or japan or china, but Korean, Japanese and Chinese.
> > +AD4-
> > +AD4-So, if we are willing to accept these presumation, then supposing to
> > +AD4-register a Japanese name, you would then registered it under .jp.icom
> > +AD4-irregardless where you are located. Similar for countries with several
> > +AD4-offical languages (e.g. Singapore), different registration occurs under
> > +AD4-different xx.icom.
> > +AD4-
> > +AD4APg-         +ACo- and still worse problems for countries whose companies
> > +AD4APg-         and organizations wish to participate in international
> > +AD4APg-         e-commerce or discourse and hence to exhibit multiple
> > +AD4APg-         languages to the rest of the world.
> > +AD4-
> > +AD4-Yes. One of the problem is that you will not be able to mix Chinese,
> > +AD4-Japanese, Korea, French, German etc in a single label or domain. On the
> > +AD4-other hand, for sanity check, do you +ACo-really+ACo- want to have such a domain
> > +AD4-name?
> > +AD4-
> > +AD4-ps: Having said all these, let me say that I do not like a technical
> > +AD4-solution which makes certain assumption on the domain names, especially
> > +AD4-mixing language/script. My reason is that I am a Internet Engineer, not
> > +AD4-a Social Engineer, so I am not in any position to say what is right or
> > +AD4-wrong or how anyone should do their naming.
> > +AD4-
> > +AD4--James Seng
> > +AD4-
> > +AD4APg- Or am I missing something?
> > +AD4APg-
> > +AD4APg-     john
> > +AD4APg-
> > +AD4APg- --On Thursday, 06 July, 2000 17:46 +090- Deuk-Kul Jang
> > +AD4APg- +ADw-dkjang+AEA-smind.co.kr+AD4- wrote:
> > +AD4APg-
> > +AD4APg- +AD4- Hallelujah    Hallelujah    Hallelujah
> > +AD4APg- +AD4-
> > +AD4APg- +AD4-
> > +AD4APg- +AD4- This proposal is concerning an Internet Draft for an encoding
> > +AD4APg- +AD4- for IDN
> > +AD4APg- +AD4- (http://www.ietf.org/internet-drafts/draft-dkjang-idn-00.txt),
> > +AD4APg- +AD4- that represent a method for internationalized domain name as
> > +AD4APg- +AD4- below.
> > +AD4APg- +AD4-
> > +AD4APg- +AD4- -------------------------------
> > +AD4APg- +AD4-      When a multilingual domain name is converted to the
> > +AD4APg- +AD4-      traditional one. 1. The 'multilingual key' is included
> > +AD4APg- +AD4-      automatically in converted names. 2. The 'language key'
> > +AD4APg- +AD4-      is included automatically in the converted names.
> > +AD4APg- +AD4-...
> > +AD4-
> > +AD4-

Prev by Date: Re: [idn] Proposal for creation of new gTLD for IDN
Next by Date: Re: [idn] Proposal for creation of new gTLD for IDN
Prev by thread: Re: [idn] Proposal for creation of new gTLD for IDN
Next by thread: Re: [idn] Proposal for creation of new gTLD for IDN
Index(es):
- Date
- Thread