Comments on the Unicode Codepoints and IDNA Internet-Draft

Kenneth Whistler kenw at sybase.com
Tue Jul 29 02:31:51 CEST 2008


>
> On 28 jul 2008, at 23.59, Kenneth Whistler wrote:
> 
> > Certainly no more so that a clueless registry
> > that lets somebody register a Sumero-Akkadian Cuneiform
> > domain name or a Runic domain name.
> 
> Just remember that we not only talk about a clueless registry, but a  
> clueless (or a clever) registrant that register 3rd (or 4th) level  
> domains.
> 
> Question is, does registration of these characters that are PVALID in  
> -02, but requested to be not so, create harm if registered and used?

And the answer to that is no.

I tried to make this clear already, but let me restate.

What NIDA has indicated is a preference to use only the
11172 modern Hangul syllables for Korean.

As I said, that is fine. Nobody would object to such a
restriction for Korean domain name registration, I believe.

In fact, this is simply a reiteration of:

http://www.iana.org/domains/idn-tables/tables/kr_ko-kr_1.0.html

which lists LDH plus 11172 modern Hangul syllables as valid
for registration (and marks 2350 of them as recommended for
interoperability).

IDNA 2008 would not change that recommendation nor invalidate
anything already registered.

If somebody *else* chose to register an Old Hangul name
or use it in other levels, the protocol should not prevent
that, anymore than the protocol should prevent other PVALID characters
from being used in other contexts.

DENIC presumably does not allow any Unicode combining characters
in a registration. The list published as:

http://www.denic.de.en/domains/idns/liste.html

for allowable Latin letters is all precomposed characters
(and would remain as such under NFKC normalization under
the protocol definition for IDNA 2008).

Do we conclude from that restriction by the DENIC that
*all* Unicode combining characters could "cause harm if
registered and used" in some other domain registry, and
therefore mark them all as DISALLOWED in the table?

[And in case anybody in this discussion doesn't recognize
that as a rhetorical question, the answer to that is
also no.]

This is almost exactly parallel here to the Korean case.

Anybody who uses Hangul jamo to form an Old Hangul syllable
simply won't find that registered in the Korean domain, since
that NIC won't register that. Anybody who uses Hangul jamo
to form a modern Hangul syllable *will* find that in
the Korean domain, since the protocol forces the normalization
of such strings to the NFKC form (in this case = the Hangul
syllables that the Korean NIC allows).

--Ken




More information about the Idna-update mailing list