[Almost OT] Re: Hangul jamo issues - are jamo sequences legitimate?

John C Klensin klensin at jck.com
Wed Jan 10 10:50:41 CET 2007


FWIW...

I am assuming that your reports of what is routinely done in 
Korea are correct.  Certainly you are in a better position to 
know than I am.  However, to repeat something I've said before 
--and that is, I think, just a variant on what Ken is saying-- 
"people write things like this" is neither a necessary nor a 
sufficient condition for those things being permitted in domain 
names -- especially if there is the slightest chance of issues 
arising when the same "word" or other string can be written in 
multiple ways that are not resolved by NFC.

We really need an informed opinion from NIDA on this --an 
opinion that is informed by expert Korean linguists and that 
balances whatever desires exist for particular strings with the 
desire for safety-- and I am anxiously awaiting the comments 
that I believe we were promised soon.

     john


--On Wednesday, January 10, 2007 11:12 +0900 Soobok Lee 
<lsb at lsb.org> wrote:

> On Tue, Jan 09, 2007 at 05:41:44PM -0800, Kenneth Whistler
> wrote:
>> Soobok Lee said:
>>
>> > Not only hangul syllables but also jamo letters form hangul
>> > words.
>> >
>> > Such use is not an "illustrative" use of jamo letters.
>>
>> But it does seem to me that this is going very OT for the
>> purposes of this list.
>>
>> In any language or writing system you can find unusual
>> edge cases that can formally be considered part of the
>> mechanisms for writing "words".
>
> Currently, you can see "KI-YEOK NI-EUN DI-GEUD" very
> often in booknames like (ABC in ascii world). that is
> not an edge case in Korea.
>
> At that time of 15th century, jamo letters were inter-mixed
> with hangul syllables, as joining characters between
> hangul nouns, as you can see HUN-MIN-JEONG-EUM standard
> document cited above.
>
>>
>> For example, U+0026 AMPERSAND is rather widespread in the
>> representation of wordlike abbreviations in English. See,
>> for example:
>>
>> http://en.wikipedia.org/wiki/Dungeons_&_Dragons
>>
>> Now it seems to me that someone might want to be able to
>> register "AD&D.com" or the like, but it is most unlikely
>> that IDNA would be extended to include "&" for that.
>>
>> Even more obviously, the apostrophe is a mandatory and
>> widespread part of multiple orthographies, including
>> English and French, but is totally out of bounds for IDNA.
>
> But, jamo letters do not
> cause  such punctuational problems like that of ' or &  in
> URI/IRI.
>
> use of jamo letters in hangul words is close to
> use of F in "IMF". You should pronounce that "F" as "ef",
> unlike the case of "f" of "fine".
>
>
>>
>> It strikes me that a perfectly reasonable position to take
>> for Korean for IDNA to allow any of the 11172 Hangul
>> syllables (or the jamo sequences that are canonically
>> equivalent to them, which would resolve, through nameprep,
>> to the same strings), and not to attempt to restrict any
>> usage of Hangul with CJK ideographs. That would seem to
>> easily handle the 99.99% case for Korean without any
>> particular difficulties.
>
> Yes, I estimate  that hangul syllables will cover
> 99% of user demands for Korean IDN label.
>
> But, such jamo use (currently marginal) in korea is increasing
> ,especially among teenagers.
> ( http://hompy.hangame.com/themeBlog.nhn )
> And we already have jamo-containing words/names.
>
> Moreover, U+31xx compat jamo letters are the only input method
> for jamo chars under NFC and KSC5601. We have no direct input
> method for U+11xx,which is not in KSC5601->UNICODE table.
>
> So, U+31xx and U+11xx both should be allowed in labels.
>
> Soobok
>
> _______________________________________________
> Idna-update mailing list
> Idna-update at alvestrand.no
> http://www.alvestrand.no/mailman/listinfo/idna-update






More information about the Idna-update mailing list