Hi,
At Mon, 05 Feb 2001 11:56:48 +0000,
Markus Kuhn <[EMAIL PROTECTED]> wrote:
> What are you talking about? Japanese people will always use Japanese
> Unicode fonts to display plaintext (unless they use multi-lingual
> software for multi-lingual applications). This is easy to implement.
Did you read my mail? I already wrote about this. Yes, large companies
like Microsoft, Apple, and RedHat can release "Japanese version" of
their OSes with Japanese fonts.
However, community-based distributions like Debian and FreeBSD cannot
take this approach, because they can not and should not release
"Japanese version". I said in the previous mail that Debian's aim is
a single distribution for all over the world, where all what users
have to do is to set LANG variable properly.
> In
> fact, XFree86 4.0 shipped with *ONLY* a Japanese Unicode font, there
> still isn't a Chinese one.
Do you remember a discussion in i18n@xfree86 mailing list that
whether we need separate version of -misc-fixed-medium-r-normal-*
font Chinese, Japanese, and Korean?
http://www.xfree86.org/pipermail/i18n/2000-November/000422.html
http://www.xfree86.org/pipermail/i18n/2000-November/000424.html
http://www.xfree86.org/pipermail/i18n/2000-November/000426.html
I think someone may develop Chinese fonts, because Chinese (Taiwan)
people also complain.
> Nobody in Japan will ever or has ever been
> bothered by Chineses glyph shapes used to display Japanese text!
I agree with you that Japanese people who use "Japanese version"
OSes (Windows, Macintosh, and preconfigured commercial Linux) won't
be bothered, if they use "Japanese version" of OSes and don't want
to read Chinese or Korean texts on the OSes.
However, such usage kills the merit of Unicode. Unicode should
be able to be used for multilingual plain text.
> Any
> fear to the contrary is a false misconception by a few geeks who misread
> the Unicode 1.0 and 2.0 books, which unfortunately used only one single
> Chinese font to render the CJK section.
No. We need some mechanism to distinguish Chinese, Japanese, and Korean
version of the same codepoint in Unicode.
> In case you didn't notice, >90% of Japanese
> word processing has been done in Unicode for many years now (thanks to
> the market dominance of a certain word processor produced in Redmond,
> which has been fully Unicode based since the 1997 release).
I use it. Since I use Japanese version of Windows which includes
Japanese fonts, I can use it. Note that the knowledge that Unicode
is used in Windows is a geek knowledge. Unicode is used as an
internal encoding and average users don't know it.
> ISO 2022 is just generated when a message leaves the
> system via SMTP, because the developers were told that this is still
> necessary for compatibility with older Unix email tradition even though
> the message is converted back to Unicode on most receiving systems.
Why do you mention ISO-2022? I already wrote this problem is not
related to ISO-2022. If you are interested in promoting Unicode,
you should invent a way so that plain Unicode text can distinguish
Chinese, Japanese, and Korean versions of the same codepoint.
> "Average Japanese people" have been using mostly M$-Word with Unicode
> for how long now? How many "average Japanese people" still consider
> Linux distributions useless crap software because it doesn't support
> Japanese as simply and easily as the Unicode based competition products
> from Redmond do?
Oh, you understand well why Linux is not so popular yet in Japan
because you are annoyed everyday and everyhour by softwares which
don't support Japanese, your daily language! :-p
I kindly teach you the real reason. There are a lot of softwares
which don't support multibyte encodings.
> We are aiming to get Japanese character processing just as smoothly
> integrated into the very fabric of the Linux infrastructure as ASCII,
> and I am certain that neither so-called ISO 2022 implementations nor any
> other existing Japanese Unix legacy practice is going to help
> accomplishing this in a globally sustainable way. We need one single
> encoding,
I see. I know you are interested in promoting Unicode. I don't
deny it. However, today's Unicode is too poor for CJK people.
Therefore, what you should do is to make Unicode better than
ISO-2022 in every fields. In the point of simplicity, Unicode
is better. However, this cannot deny the weak points of Unicode.
Fortunately, Unicode is a developing encoding. Please admit the
weak points of the current version of Unicode and try to improve
Unicode to overcome the weak points. Don't pretend Unicode is
perfect.
> and language or font style tagging are independent, optional,
> and orthogonal issues to that.
If you are saying about Roman or Italic, I agree. However,
CJK Han Unification problem is "I cannot read a text written
in my mother tongue" problem. I don't want to use Chinese
dictionary to read a plain text written in my mother tongue.
---
Tomohiro KUBOTA <[EMAIL PROTECTED]>
http://surfchem0.riken.go.jp/~kubota/
"Introduction to I18N"
http://www.debian.org/doc/manuals/intro-i18n/
-
Linux-UTF8: i18n of Linux on all levels
Archive: http://mail.nl.linux.org/lists/
