Re: ISO 15924 text revision : missing reference to ISO 639-3/-5

From: Michael Everson (everson@evertype.com)
Date: Sat Jan 24 2009 - 04:10:43 CST

  • Next message: Richard Ishida: "Urgent call for clarification of Armenian numbering rules"

    On 24 Jan 2009, at 02:13, verdy_p wrote:

    > The current text of ISO 15924 should be updated in light of ISO
    > 639-3 and -5.

    Why?

    > "Where possible, the first three letters of the four-letter code
    > corresponds to the three-letter code. Preference is given to the
    > Bibliographical codes given in ISO 639-2 in deriving the codes
    > specified in this International Standard. [...]
    >
    > Dans la mesure du possible, les trois premières lettres du codet à
    > quatre lettres correspondent au codet à trois lettres. La
    > préférence est accordée au code bibliographique de l’ISO 639-2, dans
    > la dérivation des codets attribués dans la présente Norme
    > internationale."

    This text does not prohibit use of ISO 639-3 or 639-5. But at the same
    time there is not a great deal of pressure on ISO 15924. Codes are
    added slowly, even conservatively. Indeed "where possible" and "dans
    le mesure du possible" gives the Advisory Committee the freedom to
    depart from 639 entirely.

    > This is clearly not enough, given that the ISO 639-2 registry is
    > nearly closed, but most other scripts named after a language will
    > most probably come from ISO 639-3 (or possibly ISO 639-5 if the
    > script is used by a family of language whose name matches the script
    > name).

    This describes a non-existent problem.

    > However, I did not check all the existing script codes whose name do
    > not match a language of ISO 639-2 but can
    > match a language of ISO 639-3 or family in ISO 639-5. There may
    > exist exceptions now, notably for scripts that are still not
    > encoded in Unicode/ISO 10646 but that already have a ISO 15924 code
    > (Lydian, Lydian, Mandaic, Osmanya, Pahlavi, Parthian...), or whose
    > truncation to 3 letters may collide with the ISO 639 alpha-3 codes
    > of unrelated languages (Cirth, Italic...)

    I'm fairly sure there isn't a problem. To modify the text of ISO 15924
    would require a re-balloting the whole standard. I don't see a
    compelling reason to do this.

    Michael Everson * http://www.evertype.com

    This archive was generated by hypermail 2.1.5 : Sat Jan 24 2009 - 04:14:43 CST