From: Julian Bradfield (jcb+unicode@inf.ed.ac.uk)
Date: Sun May 24 2009 - 13:14:44 CDT
>I don't believe the Ethnologue does so. If it did, it would disagree with
>ISO and IETF BCP 47, in which en means any English; en-US, en-UK, ...
It seems rather silly to say you don't believe something which you can
trivially check.
http://www.ethnologue.com/14/show_iso639.asp?code=en
Yes, it does encompass other dialects such as US Englishes, but
they're listed under "also spoken in", not in the head definition.
However, that doesn't really matter - all that matters is that en is
not identical to en-US, and your transformation varies between types
of en.
>Often one has to make a choice; for example, if I ask for an 'en' web page,
>I need to get either en-US or en-UK, or en-CA, or en-AU, etc. If you know
>other information about the user, you may be able to pick the best one. In
>the absence of such information, the typical choice is to go with the
>variant with the most users: en-US for English, fr-FR for French, etc.
Different situation. If you claim to transform from en to X, your
transformation should be correct for anything that is en. If you can't
do that, because the transformation varies between subtypes of en, you
must include the subtype in the specified domain of your transformation.
Think of it in terms of subtyping in programming languages: if you ask
for an "en" Web page, then returning you an "en-UK" or "en-US" page is
fine, because "en-UK" is a subtype of "en". But if you have a function
that actually behaves correctly only on "en-US" arguments, it's unsound to
declare it to have argument type "en".
-- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
This archive was generated by hypermail 2.1.5 : Sun May 24 2009 - 13:17:23 CDT