Re: Question about new locale language tags

From: Doug Ewell (
Date: Fri Dec 22 2006 - 12:41:35 CST

  • Next message: Jony Rosenne: "Off-topic: Non-text localization"


    You have missed the entire long thread of discussion in the Language Tag
    Registry Update (LTRU) working group concerning extended language
    subtags, what they are, and why they are that way. They are correlated
    1-to-1 with ISO 639-3 code elements that are encompassed by an ISO 639-3
    macrolanguage. "min" has only one meaning in ISO 639-3, therefore it
    has only one meaning in RFC 4646bis.

    I suggest you read the LTRU archives, beginning at this URL, before
    jumping to conclusions:

    Be careful not to confuse "extended language subtags" (also known as
    "extlangs") with "extension subtags." They are completely different.

    Doug Ewell  *  Fullerton, California, USA  *  RFC 4645  *  UTN #14
    ----- Original Message ----- 
    From: "Philippe Verdy" <>
    To: "Doug Ewell" <>; "Unicode Mailing List" 
    Cc: "Peter Constable" <>
    Sent: Friday, December 22, 2006 2:15
    Subject: Re: Question about new locale language tags
    From: "Doug Ewell" <>
    > Peter Constable <petercon at microsoft dot com> wrote:
    >>> I note that Wikipedia currently uses "zh-min-nan" for Minnan
    >>> (independantly of the script used or the geographic region), not
    >>> "zh-nan" ; are there other "Min" variants?
    >> Given that "min" is the ISO 639 ID for a Philippine language,
    >> Minangkabau, which is quite unrelated to Chinese, "zh-min-nan..."
    >> would be a bit of a non-sequitor.
    If this is a Philipine language unrelated to Chinese, there isabsolutely 
    no risk that "zh-min" using the "min" extlang subtag in the "zh" primary 
    language context will conflict. Do extlang subtags have to have the same 
    meaning between various primary lang subtags? My opinion is that extlang 
    subtags have a meaning only in the context of the previous lang and 
    extlang subtags.
    So the "zh-min" and "min" full language tags will be unrelated; this 
    means that extlang will need to be registered contextually within the 
    primary lang tags in which they are valid and defined.
    A basic "universal" registration of "min" as a extlang from ISO639-3 
    will not work as expexted, and I wonder if this is a good idea to force 
    future extlang subtags to have universal meaning, and which benefit we 
    will have with such universal meaning! For me the syntax proposed in 
    RFC4646 for extlang subtags is already perfect like this and does not 
    force such constraint, so why adding it in the view of a draft 
    For me it's more important that RFC4646bis allows ISO639-3 languages to 
    be specified as registrable primary language tags (which can be used 
    without a ISO 639-1 or ISO639-2 prefix), and to define a policy for the 
    registration of ISO 639-3 codes either as primary lang subtags or 
    secondary extlang subtags, and how to resolve possible equivalence of 
    tags to a "canonical" tag.
    Merry christmas to all readers...

    This archive was generated by hypermail 2.1.5 : Fri Dec 22 2006 - 12:45:04 CST