From: Peter Constable (
Date: Wed Feb 23 2000 - 08:08:06 EST

>This is just one of the possible uses of PUA, and one that
       does not cause PUA code points to be stored or transmitted.
       What's wrong with it?

       Some possible concerns:

       - This entails creating fonts that have the presentation form
       glyphs accessible in the cmap via PUA codepoints. That, in
       turn, means that users *could* encode documents in terms of
       those presentation forms, and that very likely means that some
       users will do so. And those documents are only interchangeable
       with others that follow the same use of the PUA.

       - Smart font technologies, such as AAT and OpenType, can handle
       complex script rendering without making use of PUA codepoints,
       and also provide additional benefits (e.g. supporting features
       such as multiple levels of ligation). Software that uses an
       algorithm such as you described will be limited to only
       proprietary fonts, and misses out on the benefits provided by
       smart font technologies.

       - Since such a font contains presetation forms for Arabic (or
       Bengali, or whatever) encoded in the PUA, then its ability to
       support custom characters that an end user might need is
       limited (depending upon the number of PUA codepoints consumed).

       Not to say that, therefore, this shouldn't be done. This is
       certainly a valid use of the PUA. Any attempt to get widespread
       adoption within industry of any particular use of the PUA,
       however, is a bad idea.


This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:59 EDT