Re: Fw: [idn] I-D ACTION:draft-ietf-idn-jpchar-00.txt

[Yoshiro YONEYA <yone@po.ntts.co.jp>]

From: "Mark Davis" <mark@macchiato.com>
Subject: Re: Fw: [idn] I-D ACTION:draft-ietf-idn-jpchar-00.txt
Date: Sun, 26 Nov 2000 17:05:17 -0800

> Nameprep includes normalization to NFKC. That will map, for example,
> FF66 => 30F2
> So many, if not all, of the Japanese mappings that you have in
> http://www.i-d-n.net/draft/idntabjpcanon10.txt
> are already covered.

Yes, right.  I think NFKC is preferable for Japanese.  But it is not
decided yet as NAMEPREP normalization rule.  I think it needs clear
definition although NFKC will be chosen, because of easy understanding,
and preparation for succeeding composition defined in idntabjpcomp10.txt
that is not fully covered by NFKC.

BTW, I made a big mistake in idntabjpcanon10.txt :-(
I missed to include U+2212 (MINUS SIGN) alternative to U+FF0D
(FULLWIDTH HYPHEN-MINUS).  Please add

2212;1.1;002D

on the top of idntabjpcanon10.txt.

I don't know the reason why, but in code conversion mapping table
between Unicode and JISX0208, FULLWIDTH HYPHEN-MINUS in JIS mapped
into MINUS SIGN (U+2212) in Unicode.

-- 
Yoshiro YONEYA <yone@po.ntts.co.jp>
           aka <yone@nic.ad.jp>