JFC (Jefsey) Morfin jefsey at
Sat Feb 12 20:29:50 CET 2005

At 19:18 12/02/2005, Mark Davis wrote:
>I want to point out that there are currently 338,345 valid language tags
>according to RFC 3066. The great value of RFC 3066 is that people *didn't*
>have to go through a registration process for the overwhelming majority of
>these cases.

Yes, this IS the difference. RFC 3066 acknowledges the use of ISO 639 on 
the Internet, like RFC 1591 acknowledges the use of ISO 3166. It permits 
special cases to be registered in addition, like RFC 1591 permits special 
cases to be addressed like ".ps", ".eu", ".su" in accordance with the NTIA 
agreement with ICANN and the IETF IANA MoU with ICANN and ISOC.

The French tags you quote make no problem being used. They would make in 
immediate problem (same as the Draft) if they were registered because there 
is no reason for that in the RFC 3066 vision. RFC 3066 and 1591 accept 
codes from external existing systems with their well defined purposes and 
maintenance rules, and permit to add a few of them if necessary. RFC 
3066bis creates a brand new system and wants to impose it where it does not 

You say that the IBM and Microsoft requested registrations are needed. May 
be could you document for why they are specifically needed (the whole 
reason chain) - if this is not a proprietary reason - so we could better 
analyze and discuss the issue on concrete examples?

I will copy a response I send to John Klensin on the IETF list which makes 
a summary of the multilingual issue, together with the remarks on the IANA 
to Peter Constable.


More information about the Ietf-languages mailing list