Comments on the Unicode Codepoints and IDNA Internet-Draft

[Stéphane Bortzmeyer]

On Tue, Jul 29, 2008 at 01:27:15AM -0700,
 Vint Cerf <vint at google.com> wrote 
 a message of 38 lines which said:

> the issue might be better stated as "can these characters be so
> readily used to cause confusionas to merit their banishment at
> protocol level or warned against at "registration time" at all
> levels in domain name hierarchy?

But these two solutions are quite different. In the DNS, today, there
are no "banishment at protocol level" even between confusable
characters.

Guidelines for registries ("Allowing the registration of <U+xxxx> is
not a good idea") are a different thing. They are already used
(section 3.5 "Preferred name syntax" of RFC 1034) and the work of the
arabic script users is a good example of a work that should be used
that way for IDNA.

With "banishment at protocol level", we would never had the underscore
for SRV records, it would have been forbidden...