Re: [idn] Comparisons of the proposals

[Dan <Dan.Oscarsson@trab.se>]

James wrote:
> Now, moving down to the wires, for DNS packets, UTF-8 seem most appropriate.
> But if 8-bit dirty data is allowed, then why not localized encodings?
You mean allowing something else than UTF-8 on the wire?
First you have to tag each string with information about what encoding
is used, second you have to make every DNS server handle those encodings.
Very difficult. We do not want character set taging. That is one thing that
makes handling MIME in e-mail so difficult.
What you could allow is to have a local DNS server
recognise a local character set, if it can be detected to not be UTF-8, to
support local needs, but not on the wire everywhere. My DNS server should
probably accept ISO 8859-1 apart from UTF-8, as that works fine and will
enhance local adaption. But I cannot expect a Chinese DNS server to
support ISO 8859-1.


> Then on the zonefile, why couldnt it be store in localized encodings?
It is really up to the local DNS server implementation. It would work fine
for many places to have the zone file in a local encoding. The DNS
server will just convert in into UTF-8 before sending it on the wire.
Actually the DNS server need not use UCS internally, it can use a local
character set as long as it handles the on the wire protocol correctely
and does character handling correctely (like string matching).


> > charcters. It is based on ideas from CIDNUC and UTF-5.
> > It is only intended to encode BMP. Leaving BMP you
> > must go international and use UTF-8.
> 
> One question. Why limit it to only BMP since from the algo, it is possible for
> it to go beyond BMP...

With some extra rules it can be extended outside BMP, I optimized it
for BMP. Even if we see this type of ASCII compatibility encoding
as support for the old ASCII world, it can be used even in international
contexts when the local character support if not able to display all
the characters. So I should probably add those rules in.

   Dan