Inconsistency in ISO 8859-1/8 tables

From: Markus G. Kuhn (kuhn@cs.purdue.edu)
Date: Thu Jul 10 1997 - 21:39:40 EDT


I have a question about the official mapping between ISO 8859-8
(Latin/Hebrew) and ISO 10646-1:

In <ftp://ftp.unicode.org/Public/MAPPINGS/ISO8859/8859-1.TXT> we find
as we expect

  0xAF 0x00AF # MACRON

but <ftp://ftp.unicode.org/Public/MAPPINGS/ISO8859/8859-8.TXT> says

  0xAF 0x203E # OVERLINE

This looks a little bit suspicious, given that both characters look
practically identical. Could this be an error in the file on the
ftp.unicode.org server, or is this according to the new ISO 8859-8:1997
really the correct character there and not MACRON? If it is really
OVERLINE in Latin/Hebrew, is there some deeper reason?

Slightly puzzled ...

Markus

PS: I ran into this while proof-reading Kosta Kostis' new character
set test pages on

  http://www.kostis.net/charsets/

with Netscape Communicator 4.0 under Windows NT. It turned
out that Netscape/NT now can display all Unicode characters that are
part of ISO 8859 except those that start with ARABIC or HEBREW. It
can also display all characters of CP437, CP850, CP1252, KOI-8, and
quite some more. I hope the latest Microsoft IE can do the same and
that those fonts will also be made available for Windows95. Then, we
could in relatively short time really use Unicode at least for all
European languages, and can declare ISO 8859 (except parts 7 and 8)
for dead on the Web.

What subset of Unicode does Netscape 4.0 display at the moment under
Windows95? Check out <http://www.kostis.net/charsets/>!

-- 
Markus G. Kuhn, Computer Science grad student, Purdue
University, Indiana, USA -- email: kuhn@cs.purdue.edu



This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:35 EDT