From: Peter Constable (
Date: Tue Feb 22 2000 - 17:22:46 EST

>Now, for a PRACTICAL discussion about the means of determining
       what those characters are, there SHOULD be added (if it isn't
       already there, that is) some means to tag those characters in a
       particular file as being "pua/microsoft" or "pua/apple" or
       "pua/DoofusCompanyInc" characters.

       Rather than such a registry, with all the associated problems,
       there is another alternative for how to make applications work
       well with end-user-defined PUA characters: develop a standard
       mechanism, including a standard file format, whereby an end
       user can specify just what the Unicode properties of their
       particular characters are. There would still be open questions
       as to how an app associates a particular datum with such a set
       of properties, this doesn't need to be part of a standard:
       applications want to do different things to manage their data,
       so developers need to have freedom in doing that. But a UTR
       that specified an interoperable way to describe PUA semantics,
       perhaps a standard set of interfaces developers could use,
       possibly a sample tool for compiling a set of PUA semantics,
       could open up the potential that any user could use their PUA
       characters in a wide variety of apps and have those apps
       provide appropriate behaviour for those characters.

       There would also be open questions about rendering of
       PUA-encoded presentation forms, but that is beyond the scope of
       this standard. It needs to be addressed as well, but in a
       different forum.


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