Re: Reality check - non-Unicode in Guinea-GTZ documents 2005

From: Richard Wordingham (
Date: Fri Aug 18 2006 - 15:42:12 CDT

  • Next message: Doug Ewell: "Re: Reality check - non-Unicode in Guinea-GTZ documents 2005"

    Doug Ewell wrote on Friday, August 18, 2006 7:24 AM

    > Richard Wordingham <richard dot wordingham at ntlworld dot com> wrote:

    >> For unsupported complex scripts on Windows, surely Unicode does have a
    >> crippling effect. Or does Vista allow non-Microsoft layout definitions
    >> for complex scripts?

    > By that you must mean "user-defined layout definitions." Do you have
    > anything in particular in mind?

    Old Myanmar script/encoding (up to 2005), new Myanmar script/encoding
    (2007?), New Tai Lue, Khmer, Lanna. Now, Khmer I believe is ready (in
    Vista?) and I think the old Myanmar script was nearly ready (probably
    subject to some hefty interpretation) - examples of chipping away at the
    list of encoded scripts not yet implemented in Uniscribe. (Lanna has no
    encoding yet.)

    Also U+0BB6 TAMIL LETTER SHA - it's available in Code2000, but the version
    of Uniscribe included in Windows XP does not treat it as a Tamil letter -
    preposed vowels do not combine with it properly. I'd be startled if this
    has not already been fixed for Vista.

    > Also, John Jenkins' glib remark notwithstanding, Microsoft does not have
    > to have a complete monopoly of the way Unicode text is rendered. Before
    > there was Uniscribe, there was SC UniPad, which could at least render RTL
    > Hebrew and connected, RTL Arabic. That was better than what had come
    > before. Unicode is not a platform-dependent technology.

    No, but one is limited to applications whose renderer can support the
    script. A widely used renderer is the obvious choice for a new application.
    It will be interesting to see how much support Graphite fonts get on
    Windows. There is already a browser that can render using Graphite fonts on
    Windows, so now Graphite fonts can be used for web pages targeted at private

    >> However, for complex scripts, Unicode plus Uniscribe will be restrictive.
    >> How will Windows users add new characters to their complex scripts? Hack
    >> fonts provided a solution - the Private Use Area seems not to be an
    >> adequate replacement.

    > Incomplete though they may be, Uniscribe (and non-Windows equivalents) are
    > a far more portable and available solution than hack fonts and private
    > encodings.

    Uniscribe is not a solution at all until the script support is made
    available by Microsoft. Once the support is there, yes, it should be a
    great enabler. Until then, while hack fonts suffer from their lack of
    standardisation, they may provide an immediate solution.

    As to the use of the PUA, how else are new characters supposed to emerge in
    an increasingly computerised age? Michael Everson was right when he said
    people had not stopped inventing characters - I would add that people have
    not stopped rejecting innovations either.


    This archive was generated by hypermail 2.1.5 : Fri Aug 18 2006 - 15:47:51 CDT