IDN uses unicode because...

JFC Morfin jefsey at
Wed Jul 9 15:53:37 CEST 2008

At 02:56 09/07/2008, John C Klensin wrote:
>Just for the information of anyone who was curious about why
>text on the "why Unicode?" subject was present at all, it was in
>response to feedback and questions from a few communities who
>argued that Unicode was not particularly optimal and so why
>didn't we use X, for some value of X (real, potential, or

Having been the most vocal on this matter I think I can tell:

- as long as IDNA was supposed to the be the only way to support 
indigenous names this was a problem. The responses given by James and 
Vint about ML-DNS remove that concern.

- the argumentation introduced by Debbie's ISO/TC46 failed NWIP last 
year raised interests in some security areas (police, passports, 
banks) that could be better addressed through the security oriented 
universal code that RFC 5242 tries to joke about.

- I would however suggest to use ISO 10646 rather than Unicode for 
serveral reasons :
   - political in several countries and cultures
   - to clearly define the technical perimeter being involved
   - proponents/searchers/projects towards a security oriented code 
belong to the ISO, JTC1 and Open Norms & Standards IGF spheres rather 
than to IETF/Unicode spheres.
   - for consistency with further developments in connex semiotic 
areas. They will probably be more extended ISO and ONSIGF than at Unicode.

This also permits to welcome the inputs of the numerous Unicode 
members and of their employees as contributions rather than negociations.

"IDNA uses the Unicode character repertoire, for continuity with 
IDNA2003" clearly shows that the use of Unicode or alternatives was 
not to be considered by the WG, speeds up its conclusion and will 
clarify the implementation debate.

More information about the Idna-update mailing list