What a Locale is....

[Michael Everson]

At 10:31 -0700 2003-04-15, Mark Davis wrote:
>Jon was arguing that script should not be part of RFC 3066. (It already is,
>in yi-latn, BTW).
>
>The point that my reply was making is that the only real technical rationale
>for having two language code standards (639 and 3066) are that for the
>first, written form is not material, while for the second it is material.

No, for the second it may be material. 3066 is for extensions to 639. 
These may have nothing to do with script. Further, 639 distinguishes 
between Romanian and Moldavian, and between Serbian and Croatian, 
which is arguably a script distinction. Sweeping statements don't 
help us I guess. ;-)

>RFC 3066 already allows fine distinctions in orthography, such as "theatre"
>vs "theater", or German orthographic conventions. The RFC is deployed in a
>wide variety of contexts, such as in HTML and XML, where it is absurd to
>think that users would care about such fine distinctions as "theatre" vs
>"theater", but then not care which script the language was expressed in.

Orthography is a distinction which may affect language identification 
if not identity.

>Thus my message was an example: in English, but a mixture of scripts (of
>course, I didn't do the transliteration by hand, I cheated with
>http://oss.software.ibm.com/cgi-bin/icu/tr/ -- and it is only a
>transliteration, not a transcription).

I know. I sight-read the Armenian which impressed even me. :-) But it 
was cute, not in itself convincing. (The real arguments are more 
convincing.)
-- 
Michael Everson * * Everson Typography *  * http://www.evertype.com