Re: Question about new locale language tags

From: Addison Phillips (
Date: Wed Dec 20 2006 - 09:32:55 CST

  • Next message: Addison Phillips: "Re: Question about new locale language tags"

    While Stephane's reply is correct, I want to make sure that the
    situation is clarified.

    1. RFC 4646 is really officially done. "RFC 4646bis" is really
    "draft-ietf-ltru-4646bis", an Internet-Draft whose purpose is to add ISO
    639-3 codes to the available language subtags in the registry. This is
    what Stephane means below.

    2. RFC 4646 defines *language* tags. There are two standardization
    efforts aimed at using language tags as all or part of locale identifiers.

       - The Common Locale Data Repository (CLDR) project "here" at Unicode
    defines its locale identifiers in terms of RFC 4646-or-successor
    language tags. See, for example:

       -The Language Tags and Locale Identifiers SPEC at W3C has (had?) as
    one goal defining the relationship between language tags and locale
    identifiers (with an eye toward enabling Web services and application
    server technologies with consistent, open locale identifiers). See

    This latter effort could use more support and participation. The
    community has not supported this important effort (and other activities
    of the W3C I18N WG) very well of late and really should have an interest
    in doing so.

    3. Some implementations exist that use language tags to infer the
    locale. In particular, Microsoft's .NET APIs use "rfc1766" language tags
    as "Culture" (Locale) identifiers. And the ICU library implements CLDR.
    Other implementations may exist.

    4. The language tags (quoted below) in your note will all be valid under
    RFC 4646bis, which should happen in the next couple of months--with an
    exception. "i-ami" is a grandfathered tag and cannot be used
    generatively. Under RFC 4646bis, the tag "zh-ami-TW" would be valid.
    "i-ami" will be deprecated. For valid ISO 639-3 codes, see

    zh-nan-Latn-TW (Minnan using Latin script in Taiwan, aka. POJ)
    zh-nan-Hant-TW (Minnan using traditional Hanzi in Taiwan)
    zh-hak-TW (Hakka in Taiwan)
    i-ami-TW (Amis in Taiwan)
    5. Shorter versions of some of these tags are already registered (as 
    grandfathered). See:
    Hope that helps clarify things.
    Stephane Bortzmeyer wrote:
    > On Wed, Dec 20, 2006 at 02:46:33PM +0800,
    >  Arne Götje (???) <> wrote 
    >  a message of 60 lines which said:
    >> is there any timetable for implementing this RFC for locale tags?
    > Not that I'm aware of.
    >> I just read the rfc4646bis-02... may I assume that this is the
    >> latest revision of the rfc?
    > Certainly not.  rfc4646bis-02 is a draft ("work in progress") and very
    > much in flux. It has no authority. The only stable reference is RFC
    > 4646, which includes only ISO 639-1 and -2 (AFAIK, ISO 639-3 is not out
    > yet).
    > Sorry for the bad news, but it seems there is currently no standard
    > way to represent the fact that not all chinese languages are
    > "the chinese language".
    Addison Phillips
    Globalization Architect -- Yahoo! Inc.
    Internationalization is an architecture.
    It is not a feature.

    This archive was generated by hypermail 2.1.5 : Wed Dec 20 2006 - 09:34:52 CST