Re: Defined Private Use was: SSP default ignorable characters

From: Doug Ewell (dewell@adelphia.net)
Date: Thu Apr 29 2004 - 02:55:13 EDT

  • Next message: Theo Veenker: "Unicode_Radical_Stroke"

    Peter Kirk <peterkirk at qaya dot org> wrote:

    > Or, since a system must use some font to display the text, in many
    > cases it will be able to determine which font is currently specified
    > for display of this text and use the properties (if any) defined in
    > that font. Of course this doesn't work for entirely plain text
    > processing processes, not tied to display. But presumably in such
    > cases it would be possible to set as a parameter for the process which
    > font to take the properties from. When such processes are rewritten to
    > take advantage of this new character properties mechanism, it would
    > probably be easier 9and certainly quicker!) to get them to accept a
    > new parameter than to get them to search through perhaps hundreds of
    > fonts on the system for the ones which support each PUA character
    > individually.

    FWIW, in my sample Ewellic-script Web pages, I indicated the encoding in
    two different ways:

    (1) with an XHTML comment in the source code:

    <!--
    This page is written in the Ewellic script using the encoding proposed
    for the ConScript Unicode Registry:
            http://www.evertype.com/standards/csur/ewellic.html
    -->

    (2) indirectly, by requesting a specific font that supports it, using an
    inline style sheet:

    <style type="text/css">
    <!--
    body { font-family: Code2000 }
    <etc. />
    -->
    </style>

    There's no other common, standard, or non-markup way of indicating how
    these PUA characters are to be interpreted (not to imply that there are
    hundreds of competing PUA assignment schemes). I suppose I could have
    used the private-use ISO 15924 script codes described at
    http://users.adelphia.net/~dewell/conscript-15924.html, but those have a
    few problems of their own:

    (1) they're just as private and non-standard as the PUA assignments
    themselves

    (2) they only specify the script, not how that script is encoded

    (3) the list needs to be revised anyway, to restrict codes to the new
    and smaller range from Qaaa through Qabx.

    -Doug Ewell
     Fullerton, California
     http://users.adelphia.net/~dewell/

    This archive was generated by hypermail 2.1.5 : Thu Apr 29 2004 - 03:37:36 EDT