Area codes [was Re: Updated! LANGUAGE TAG REGISTRATION FORM : es-americas]

Peter_Constable@sil.org Peter_Constable@sil.org
Fri, 6 Sep 2002 11:14:35 -0500


On 09/06/2002 03:26:30 AM Scripts2 wrote:

>I still think Michael Everson is correct...

>It strikes me that people looking at this list should request some
>sort of standard code for "Areas" covering more than one country.



On 09/06/2002 05:55:23 AM John Cowan wrote:

>This code already exists and is in use by the U.N., which coordinates
>it with the numeric codes they assign for ISO 3166.  As we have recently
>learned, 3166 letter codes are not stable, whereas the numeric codes are.
>Here's the list of Area, as you call them, or "macro geographical regions"
>as the U.N. has it:

>019 Americas


So, will people be happier if someone simply requested to register
"es-019"? Or would they be happier if there were an RFC that simply allowed
people to use "es-019" without having to register it along with an
explanation of what was intended by it, and suggestions on how
imiplementers should handle it? As I have said all along, what is most
important in this is that it's clear to everybody what a tag denotes, and
that that denotation is something that implementers can know what to do
with.

It seems to me that we're getting hung up on a misconception regarding what
the purpose of these metadata tags is. They are not intended to make a
statement about some community of speakers out in the real world. Rather,
they are intended to make a statement about the data to which they are
attributed. People are objecting to "es-americas" because it doesn't
correspond to a coherent community of speakers, but that's not what it or
any other "language" tag should be doing. The intent of such a tag, as is
true for every tag sanctioned by RFC 3066, is to make a statement about
data to which the tag is attributed. Using "es-americas" as an attribution
of data has (assuming good documentation) a clear meaning, and a practical
use.

(I could go on with an argument that exactly the same issues exist with
de-1996 as with es-americas -- there isn't a coherent entity out in the
real world that it corresponds to, it's just a useful descriptor for
certain content -- but have to catch a plane and don't have time to develop
it, so only offer the outline.)


On 09/06/2002 03:26:30 AM Scripts2 wrote:

>Don't make language tags and subtags do a different job as well.

Nobody is asking to have language tags do a different job. What is being
sought is is a tag to describe a particular linguistic characteristic of
certain content.




- Peter


---------------------------------------------------------------------------
Peter Constable

Non-Roman Script Initiative, SIL International
7500 W. Camp Wisdom Rd., Dallas, TX 75236, USA
Tel: +1 972 708 7485
E-mail: <peter_constable@sil.org>