Re: http charset labelling

Larry Masinter (masinter@parc.xerox.com)
Wed, 14 Feb 1996 00:50:59 PST


To: mohta@necom830.cc.titech.ac.jp
Cc: gtn@ebt.com, dupuy@cs.columbia.edu, uri@bunyip.com
In-Reply-To: Masataka Ohta's message of Wed, 14 Feb 1996 00:00:14 -0800 <199602140800.RAA07198@necom830.cc.titech.ac.jp>
Subject: Re: http charset labelling
From: Larry Masinter <masinter@parc.xerox.com>
Message-Id: <96Feb14.005101pst.2733@golden.parc.xerox.com>
Date: Wed, 14 Feb 1996 00:50:59 PST

> In Japan, for these 10 years, UNIX file names with Japanese
> Kanji has been widely available. But, it is not used at all.

For longer than 10 years, Xerox XNS servers have allowed file names
with Kanji names. I think the problem was that the UNIX systems were
missing several important features.

> One problem is that it is more difficult to type in the name in
> Kanji than ASCII.

It is difficult to type in the name in Kanji if you do not have a
Japanese typing system; fortunately, the Star and Globalview system
allowed you to select files by their names rather than having to type
them.

Now that ordinary users are getting such systems on their PCs finally,
the rest of the world might experience this convenience.

> The other problem is that systems with different locales
> can not share files.

This was not true; systems that did not have Japanese fonts had more
difficulty sorting directories, but could see the names of those files
that had roman character names.

I think a global international naming system for data objects is
possible and useful; I've used one. I miss it, actually.

However, I also don't think it is possible or useful to 'tweak' the
current URL structure to try to make it international. Unfortunately,
the URN groups don't seem to have this problem in their charter,
either.