Re: [idn] IDNA applicability: per RR/Class or DNS-wide

["Eric A. Hall" <ehall@ehsco.com>]

on 6/11/2002 6:41 PM Adam M. Costello said the following:
> "Eric A. Hall" <ehall@ehsco.com> wrote:

>>The question formation differs, comparison is identical.
> 
> We're just using different terminology.

You're arguments have made it quite clear that you think this change
affects server behavior in some way, when it does not.

> Consider this scenario:  One user types a domain label containing an
> uppercase letter O-umlaut, another user types a domain label containing
> a lowercase letter o followed by a combining umlaut.  Do these two
> labels match?

How they are stored is all that matters. If they are stored as typed then
they will be represented by different octet sequences and will therefore
match as different sequences. If they are normalized to the same sequence
as part of a stringprep profile before they are stored then they can only
be represented by the resulting octet sequence.

> I think trying to work out the implications of throwing "maybe" into
> the mix is more trouble than it's worth.

There is no "maybe" here either.

-- 
Eric A. Hall                                        http://www.ehsco.com/
Internet Core Protocols          http://www.oreilly.com/catalog/coreprot/