Re: [dnsext] we need help to make names the same, was draft-yao-dnsext-identical-resolution-02 comment

[Masataka Ohta <[e-mail]>]

John Levine wrote:

> The pattern language for encoding variant names would probably be more
> complex than people would like, but if we are serious that all the
> spellings of Greek words are equivalent, or traditional and simplified
> Chinese are equivalent, I don't see how anything simpler could do the
> job.

I'm happy that you seems to have reached the same conclusion
as mine:

: For precise definitions of extended case insensitivity, we need
: a formal language for extended case insensitivity description.

But, then, the issue is purely about character encoding which
is better handled by clients outside of DNS.

Note that people working on secure DNS (I'm not) won't welcome
proposals to support yet another type of wild card.

Or, if you think pattern should be locale dependent, I agree
that its support must involve DNS protocol. Locale information
must be supplied from DNS, at least.

						Masataka Ohta