Re: Welsh Collation

From: Richard Wordingham (richard.wordingham@ntlworld.com)
Date: Thu Apr 20 2006 - 14:10:23 CST

  • Next message: Mark Leisher: "Deep structure of writing"

    Michael Everson wrote on Thursday, April 20, 2006 at 5:59 AM
    re: [indic] Welsh Collation (was: Discussion about two Telugu consonant
    signs...)

    >>>At 22:37 +0100 2006-04-19, Richard Wordingham wrote:
    >>>>I seemed to be in a minority of one when I suggested that in Welsh
    >>>>'Llangollen' should be spelt with a CGJ so that the sequence 'ng' would
    >>>>not be confused with the Welsh 'letter' 'ng', which sorts between 'g'
    >>>>and 'h'.
    >>>
    Michael Everson wrote:
    >>>The Welsh do
    > not use CGJ.

    The current, natural meaning of CGJ (= combining grapheme jibber?) was only
    formalised in March 2005 (Unicode 4.1.0). Some of us are using applications
    that completely predate CGJ (introduced in Unicode 3.2.0, March 2002) e.g.
    Word 2000 on Windows 2000 at my workplace, and many private individuals are
    still using Windows 98 or ME. On that system, Internet Explorer 6.0
    displays CGJ as the missing glyph. IE 6.0 also renders CGJ as the missing
    glyph on my Windows XP SP2 system, whereas Firefox does not, but that could
    be just be a font or font selection issue. (The IE 6.0 behaviour may not be
    totally unreasonable, given that CGJ may affect the rendering in the Hebrew
    and Fraktur scripts.) Both browsers work fine when Code2000 is specified as
    the font, and Word XP seems to handle the non-display of CGJ fine - better
    than it handles ZWJ!

    Given these circumstances, it is too early for CGJ to be in general use with
    the Latin script anywhere, though the Welsh Language Board does say that
    software should understand it
    (http://www.bwrdd-yr-iaith.org.uk/download.php/pID=66182.1). At present,
    one can only safely use it when one knows the applications that will be
    called upon to interpret it.

    This does not mean that CGJ should not be used in Welsh, merely that it
    cannot be used until installations catch up with Unicode.

    Richard.



    This archive was generated by hypermail 2.1.5 : Thu Apr 20 2006 - 14:18:55 CST