From asmusf@ix.netcom.com Sat Apr  1 20:27:22 1995
Received: from moose.cs.indiana.edu by whale.cs.indiana.edu
	(5.65c/9.4jsm) id AA04205; Sat, 1 Apr 1995 20:27:21 -0500
Received: from ix3.ix.netcom.com by moose.cs.indiana.edu
	(5.65c/9.4jsm) id AA21394; Sat, 1 Apr 1995 20:27:19 -0500
Received: from  by ix3.ix.netcom.com (8.6.12/SMI-4.1/Netcom)
	id RAA09992; Sat, 1 Apr 1995 17:24:36 -0800
Date: Sat, 1 Apr 1995 17:24:36 -0800
Message-Id: <199504020124.RAA09992@ix3.ix.netcom.com>
From: asmusf@ix.netcom.com (Asmus Freytag)
Subject: Ethiopian
To: dmulholl@cs.indiana.edu
Status: R

Unicode has run into difficulties every time it tried to diverge from 
existing PC and mainframe character standards in the definition of "what 
a character is". The recent Korean issues underscore this point.

In this light I would be interested in what you know about the encoding
of Ethiopian on PC's. Is it being used, how is it squeezed into
8-bit character sets. What are the issues.

My concerns are that many of the systems implementing Unicode will do
so concurrently with implementing 8-bit character sets for existing
system interfaces and libraries. As the Korean example has shown, there
is strong resistance where these solution differ from Unicode in what 
constitutes the atomic character. On the other hand, as implementation 
of Unified Han has shown, differences in the arrangement are of no 
practical concern.

A./

Asmus Freytag
ASMUS, inc.