From: Peter Constable (
Date: Tue Feb 22 2000 - 16:01:36 EST

       My only concern regarding the PUA would be in the event that
       major software developers promote some standard use of PUA
       characters and have that lead to software implementations which
       assume certain definitions of the PUA, meaning that users with
       particular (different) PUA needs encounter interoperability
       problems when using that software.

       If it can be guaranteed that software will always make no
       assumptions about PUA-encoded characters, there's no problem.
       But who can guarantee such a thing? Adobe has recommended to
       others that all presentation forms be encoded (using the PUA
       where not otherwise encoded), and apparently have made some of
       their assignments match those used by others like HP; Bitstream
       have indicated that they will use the same assignments as
       Adobe. We've heard that current work on Omega has gone this
       direction. I only worry that there's a trend unfolding that
       could lead to interoperability problems.


       From: <> AT Internet on 02/22/2000 11:51 AM

       Received on: 02/22/2000

       To: Peter Constable/IntlAdmin/WCT, <> AT
       Subject: Re: PUA

> -----Original Message-----
> From: James E. Agenbroad []

> Using the PUA for glyphs sems contrary to the our basic
> premises, but can
> we prevent consenting users from doing so? It seems
> contradictory to say.
> "Use these codes any way you wish" but "Don't use them for
> X". We could
> specify some likely undesirable outcomes of proposed PUA
> usages and hope
> they will heed the warning.

       Anyone can use the PUA for anything they want to. If it's
       useful for managing glyphs for some purpose, then why not? It
       should be a free arena. All bets are off as to the meaning of
       data in this area in plain text. The standard shouldn't
       restrict it's use in any way, other than the conformance
       clauses that are already there.

       --- Paul

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