Re: visible glyphs for U+2062 and similar characters

From: Andrew C. West (andrewcwest@alumni.princeton.edu)
Date: Tue May 13 2003 - 07:49:43 EDT

  • Next message: Ben Dougall: "Re: encoding checker"

    On Mon, 12 May 2003 20:44:14 +0000, jameskass@att.net wrote:

    > Visible glyphs for control characters can be useful in plain text
    > editing/input and for filling out Unicode charts. But, if this is
    > messing things up for other applications, those glyphs could be
    > removed from Code2000.

    I think that it is a good idea for fonts to provide visible glyphs for invisible
    characters (as do many fonts, not only Code2000), and it should be up to the
    rendering engine to not display an invisible character where it is inappropriate
    to be displayed visibly.

    For example, Code2000 provides visible glyphs for U+180B through U+180D
    [MONGOLIAN FREE VARIATION SELECTOR ONE ... THREE]. With the latest version of
    Uniscribe the visible glyphs for FVS1..3 are displayed when the characters are
    in isolation or set within non-Mongolian text, but when the characters are set
    within Mongolian text Uniscribe does not render FVS1..3 at all (i.e. they are
    invisible). This behaviour is extremely useful :

    1) It allows invisible characters to be invisible when set in the appropriate
    context (e.g. within running Mongolian text for FVS1..3)

    2) It allows visible representation of invisible characters in Character Map
    applications, etc.

    3) It allows visible representations of invisible glyphs in metalanguage
    contexts.

    4) It allows the user to easily notice cases where an invisible character has
    been inadvertently placed or left in an inappropriate textual context.

    Similar behaviour could be implemented for U+2062 [INVISIBLE TIMES] and other
    invisible characters by the rendering engine (e.g. do not render U+2062 between
    numbers, elsewhere render the visible glyph).

    In short, visible glyphs for invisible characters are indeed useful in some
    contexts, and I would be sorry to see them disappear from Code2000.

    Regards,

    Andrew



    This archive was generated by hypermail 2.1.5 : Tue May 13 2003 - 08:52:30 EDT