What rules have been used for the current list of codepoints?

Martin Duerst duerst at it.aoyama.ac.jp
Fri Dec 15 10:02:23 CET 2006


At 19:18 06/12/14, Vint Cerf wrote:
>I have been following along, perhaps with less understanding than many, but
>I continue to have concerns that we are not always distinguishing that which
>is needed for expressive natural language, and that which is safe, stable,
>and secure for Internet domain names.

Hello Vint,

To some extent, what is safe, stable, and secure for domain names
will have to sort itself out over time. Some typical examples
from ASCII: 1/l/I, O/0, or 8/B. If we wanted to be really, really,
on the safe side, we would have eliminated numerals and upper case
ASCII completely from the DNS. Even whether they are necessary
to express natural language is debatable.

The above example shows, in my opinion, that radical elimination
of everything that looks similar even slightly isn't the right
direction. Security issues and expressibility issues have to be
balanced. Using good fonts (which make the letters in the above
letter groups clearly distinct) will help. Ultimately, we will
end up with solutions that will have some restrictions
(an ASCII example is that you can't use an "'" in a domain name,
so e.g. O'hare.org isn't allowed), but don't unduely limit
natural language, in particular in cases where characters
might look similar to outsiders but not to 'native' users.

Regards,    Martin.



#-#-#  Martin J. Du"rst, Assoc. Professor, Aoyama Gakuin University
#-#-#  http://www.sw.it.aoyama.ac.jp       mailto:duerst at it.aoyama.ac.jp     



More information about the Idna-update mailing list