Andre Engels wrote:

"Ralesk Ne'vennoyx" ralesk@livejournal.com schrieb:

...

Which I remember stated in a document too. Maybe what's suggested in the RFC (1766 and 3066 if I recall the numbers correctly), we should use the x primary subtag, and name it x-tp to avoid trouble (such as them getting a 3-letter code). Same goes for Lojban and Ido and Klingon which don't seem to have codes yet either.

Those have all been assigned codes, Ido even a 2-letter code:

Ido has 3-letter code 'ido' and 2-letter code 'io' since January 2002. Lojban has 'jbo' since Sptember 2003. Klingon has 'tlh' since February 2004.

See http://www.loc.gov/standards/iso639-2/codechanges.html

Same site does say "Both code lists are considered open lists (i.e., it is possible for new entries to be added to the lists)." by the way. (see http://www.loc.gov/standards/iso639-2/faq.html). What source does Brion have to claim ISO 639-1 is not, if I may ask?

Without necessarily addressing the immediate issue, it would be a good idea for consistently dealing with this kind of issue. 1. 2-letter ISO codes (ISO 639-1) should be applied strictly and no new ones should be invented by us. 2. 2-letter "w" codes (including w+numeral) that do not refer to a language (only "wa" and "wo" are so defined) should be reserved for interwiki links. (i.e. "wp" = wikipedia; "wq" = wikiquote, etc.) 3. 3-letter ISO codes (ISO 639-2) should be preferred when 2-letter codes don't exist. 4. 3-letter SIL codes should be preferred in the absence of a an ISO 639-2 code, providing that the choice does not create a conflict with any other existing ISO-639 code. (SIL codes should be treated as case insensitive) 5. 3-letter (case insensitive) codes of our own invention may only be used if none of the above applies, and the invented code does not conflict with any other existing ISO 639-2 code or SIL code.

Ec