Language Tags and Character Sets

Michael Everson everson at evertype.com
Tue Jun 3 13:20:53 CEST 2003


At 13:31 +0100 2003-06-03, Marion Gunn wrote:

>I also feel very strongly that cultures overlapping international 
>boundaries should be tagged by consensus between relevant national 
>bodies (in this case BSI and NSAI), rather than fall victim to 
>inexpert advice.
>
>>  I am not sure why this discussion goes into the Unicode list...
>
>That is a technically messy story. Do you recall my supplying 'quick
>brown fox' examples, showing Irish has a different Unicode/ISO 10646
>character set to Gaelic and Manx, which have their own distinctive
>charsets?

There are no "Unicode/ISO 10646" character sets. Unicode and ISO/IEC 
10646 *is* a character set. *The* character set. The languages you 
mentioned have different alphabetic repertoires. If you would like to 
learn more about alphabetic repertoires, please see 
http://www.evertype.com/alphabets

Having said this, your forwarding this discussion to the Unicode list 
was once again inappropriate, because language tags

>Well, there is some interest in the university here in extending 
>that approach to see if we could do the same with frequency measures 
>of IPA charsets

Please learn what the word "charset" means.

>to tentatively fix borders between dialects, treating sounds as 
>isoglosses, as it were, in tandem with our usual syntactic/wordstore 
>analyses.

Such dialectological work is laudatory, but the IETF languages forum 
is not the place to do such work, and such work is not really related 
to language tagging, at least not at this stage of the work as you 
report it.

>No idea if that would work, but it is something I wish to raise 
>first within a more local NSAI forum.

The dialects of Hiberno-English are not on the programme of work of 
that forum, of which I am convener.
-- 
Michael Everson * * Everson Typography *  * http://www.evertype.com


More information about the Ietf-languages mailing list