Re: [idn] IDN eamples for testing

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

Paul Hoffman wrote:

>How many times on this list do we need to dredge up the 
>"transitional" debate? The IETF has lots of experience with the 
>failure of "transitional" strategies that involve multiple 
>supposedly-identical data.
>
>>Or must this be considered non-conformant to the standards?
>
>There is no standard for RACE; of course any use of it is non-conformant.
>

That is correct, but sending ISO 8859-1 is. The DNS RFCs allows any
octet values in a domain name.

Nor is it necessary to implement IDNA in applications. You can do
it in the DNS server and allow the clients be unmodified.

   Dan