Re: [idn] new I-D: Safely Encoding of likeness information into ACE label version 0.2

["Soobok Lee" <lsb@postel.co.kr>]

----- Original Message -----
From: "Adam M. Costello" <amc@cs.berkeley.edu>
To: <idn@ops.ietf.org>
Sent: Monday, July 30, 2001 1:01 PM
Subject: Re: [idn] new I-D: Safely Encoding of likeness information into ACE
label version 0.2


> Soobok Lee <lsb@postel.co.kr> wrote:
>
> > Even If we normalize U+6046 into U+6052 ( or U+30AB -> U+529B ), my
> > I-D's likeness encoding will allow recover U+6046 (or U+30AB) as long
> > as applications do not casefold the produced ACE label.
>
> Today's ASCII domain names already allow visually similar characters
> that don't match (0O 1lI 2Z 5S 6G) and people have lived with it for
> a long time, so I don't think we should feel obligated to eliminate
> this problem in IDNs, especially if it means significantly increasing
> complexity and decreasing efficiency, which it appears to do.
>
> AMC
>
Choosing a disambiguating  font may eliminate such similarities. But,
We should not forget the problem around borrored glyphs that look identical.
Cherokee 'A' and Latin 'A' and latin 'a' and cyrillia 'a' look identical than
look similiar.
I haven't ever seen any identifier system that has duplicate  atoms (in BNF):
  IDN   :=  (atom)+
  atom :=   'a'|'a'|'a'|'b'|'b'|'c'|'c' ......


Soobok Lee