Re: [idn] Determining equivalence in Unicode DNS names

[Patrik Fältström <paf@cisco.com>]

--On 2002-01-26 23.40 -0800 Stuart Cheshire <cheshire@apple.com> wrote:

> Then I'm afraid I don't see your objection to my original proposal, which 
> was *precisely* what you just said, namely: the DNS protocol doesn't 
> determine if two names are equal; let the domain name holder decide.

That was not what you said. You said:

> It seems to me that the solution is to give up on the idea of a single 
> global set of rules, and instead let each name server be authoritative 
> for the equivalence rules for the zones for which it is authoritative.

That is a completely different thing, and that was what I objected to.

The servers need to have exactly the same rules.

>> The changes are very controlled, and no changes are made which have
>> the impact you tell.
> 
> I wish. I have to sit through the meetings where the file system guys 
> debate how to handle each rule change, and its impact on forward and 
> backward interoperability.

Give one example after Unicode 3.0. (There is one change, and you might
catch that one, but that change was deliberate, and made in cooperation
with IETF....i.e. it was decided to make that change to 3.1 just because
IDN had not taken off yet.)

Before 3.0, sure.

And, yes, your filesystem guys have digged a very deep hole in the ground
where they are currently, trying to get up. I am thinking of the ability of
use of UFS or HFS(+) for MacOSX, where things like case insensitivity in
HFS+ is biting them. I have had some problems with Perl CPAN install
mechanisms which rely's on the fact that "Makefile" and "makefile" actually
should be two different files in "normal" UFS....so there are some problems
with some install scripts from CPAN.

So, I understand the discussions you have inhouse...I don't want to be
there.

  paf -- Very happy MacOSX user!!!