Re: 28th IUC paper - Tamil Unicode New

From: Adam Twardoch (
Date: Mon Aug 22 2005 - 16:48:56 CDT

  • Next message: Richard Wordingham: "Re: 28th IUC paper - Tamil Unicode New"

    Richard Wordingham wrote:

    > So how do I get it to live up to its 'responsibility' to support an
    > Indic conlang living in the PUA?

    As the name says, PUA is intended for "private use". Neither the Unicode
    Standard nor the applications have any knowledge about the writing
    system, the directionality or other properties of characters that use
    PUA codepoints. To avoid redundancy of mapping information in the fonts
    and reduce their complexity, the OpenType architecture uses the two-tier
    model. An application needs to have knowledge about which writing system
    a character belongs to, what is its directionality etc. -- and this
    information is by definition not available for PUA. Other technologies
    such as AAT (Apple Advanced Typography) have such information contained
    in the "prop" and "Zapf" tables that are included in the font so you can
    define arbitrary reorderings even for PUA characters using AAT. This
    will work on Mac OS X in Cocoa applications. However, experience has
    shown that the effort to develop an AAT state machine that would
    reliably perform all the complex script processing within a font was
    quite substantial -- this is why the OpenType technology turned out to
    be far more successful than AAT. The limited flexibility of OpenType
    also means limited complexity.

    There are open source OpenType Layout processors available: IndiX and
    ICU Layout.
    Since they're open source, you can extend them to cover your own PUA
    usage schemes.


    This archive was generated by hypermail 2.1.5 : Mon Aug 22 2005 - 16:49:46 CDT