NetNews Usenet Archive 1992 #31/NN_1992_31.iso/spool/comp/std/internat/948

Internet Message Format | 1992-12-31 | 2.6 KB

Path: sparky!uunet!gatech!emory!wupost!spool.mu.edu!yale.edu!yale!mintaka.lcs.mit.edu!ai-lab!wheat-chex!glenn
From: [e-mail] (Glenn A. Adams)
Newsgroups: comp.std.internat
Subject: Re: Data tagging (was: 8-bit representation, plus an X problem)
Date: 31 Dec 1992 07:15:08 GMT
Organization: MIT Artificial Intelligence Laboratory
Lines: 39
Message-ID: <[e-mail]>
References: <[e-mail]> <[e-mail]> <[e-mail]>
NNTP-Posting-Host: wheat-chex.ai.mit.edu

In article <[e-mail]> [e-mail] (Masataka Ohta) writes:
>
>I heard that, microsoft's NT will have a locale mechanism so that
>it can print Japanes Han as Japanese and Chinese Han as Chinese,
>which is impossible with bare 10646/Unicode.
>
>Then, how can we have a file containing both Japanese and Chinese?
>

Ohta-san is making much too much about this issue.  First, there is
much less difference between a Chinese font's rendition of a given Han
character and a Japanese font's rendition of the same character than
he seems to imply here.  Indeed, in practice, it is merely a matter of
a font difference.  Furthermore, should a Japanese reader see a 10646
Han character displayed with a Chinese font, or, should a Chinese reader
see it with a Japanese font, it will still be legible in general.

If one wants a 10646 encoded text containing a mixture of Chinese and
Japanese to be displayed using different fonts for the Chinese and Japanese
parts, then a higher level protocol (rich text) must supply the information
needed to determine which text is Chinese and which is Japanese.  This can
be accomplished indirectly by using font runs, or directly by using some
form of language tagging.

This is no different from wanting to display English text with a mixture
of regular and italic faces.  10646 can't tell you which characters should
be displayed with an italic font either.  The key point is that, in the
absence of font shift information, one can still read the text.  Since this
is also true in the mixed Chinese/Japanese case, it is an equivalent problem.
10646/Unicode properly encodes only the "plain text" information which
allows for legible display.  Applications which desire more sophisticated
display will have to add other "rich text" information necessary for the
control of such advanced display.  The Unicode plain text model considers
display with multiple fonts to be a sophisticated display requiring
appropriate font tagging or other data (e.g., language tags) which allows
the proper font to be selected.

Glenn Adams
Cambridge, Massachusetts