It relates bug-709,
JISX0201 parts are shown with fonts selected in the "Fonts for Japanese"
in the preference menu now.
It results in that the ASCII parts in Japanese pages are also shown with
fonts selected in "Fonts for Japanese".
Can we make the ASCII parts in Japanese pages use the
setting in the "Fonts for English" in the preference menu?
In X, we can use iso8859-1 for ASCII parts,
jisx0201, jisx0208, jisx0212 for Japanese pars, and so on,
so it is natural that we can do so.
# that is, for codes from \uFF80 to \uFF9F,
# we can use font family selected in the "Fonts for Japanese",
# for codes \u00xx, we can use the settings in the "Fonts for English".
It is difficult to prepare times/helvetica/cursive/fantasy
fonts which includes jix0201...
------- Additional Comments From Kazuhiko Koike 2001-06-15 12:21 -------
I cannot handle this, so leave this to Katakai.
------- Additional Comments From Masaki Katakai 2001-07-06 16:22 -------
We can put RFE for this.
But I don't think it will be coded as the default feature.
It is likely that this feature can be enabled with
some words in prefs.js.
In Windows, Japanese fonts are used for ASCII in japanese pages,
so in a way the situation is the same as in Linux.
But I am in the side of the implementer,
so could you someone register the RFE in bugzilla.mozilla.org?
Not bug, but RFE, please.
Some suggestion is;
1. Now fonts as lang=ja are jis201,208,212,
so we can add iso-8859-1 at the head of this list.
2. First try to use x-western fonts, and then try to use
fonts of lang=ja.
------- Additional Comments From Kazuhiko Koike 2002-01-19 21:10 -------
------- Additional Comments From Katsuhiko Momoi 2002-01-27 14:15 -------
Matsumoto, Katakai, this is RFE only for Unix, right?
In Windows and Mac, Many Japanese fonts have
Western and Japanese characters in a good balance,
so mixing Japanese fonts and Western fonts may
corrupt the balance of characters.
------- Additional Comments From Shoji Matsumoto 2002-01-27 15:18 -------
In regard to this problem,
I can say this is only for UNIX.
In a practice, I think we should
select fonts for jisx0201, 0208 separately as in NC4.
but, for example, if we follow the IE,
we may follow how the windows can select the fonts
when we specify face='arial, MS PGothic',
In some word processors, we can specify
separate fonts for English part and Japanese part.
(I wonder how CSS handles the situation...)
I imagine that, Windows first tries to use arial, then MSPGothic,
In a extreme, in DTP, we can select separate fonts
for English, Math, Marks, Hiragana, Katakana, Chinese character.
And in a future, Unicode which includes CJK is popular,
we need to specify certain part of some font for some code of one language.
# still dreaming...
# If we specify CJK in a CJK-mixed document
# using language tag cn, ko, ja,
# then we should be able to specify Roman part of Japanese as jaRoman,
# Hiragana Katakana part of Japanse as jaHirakata,
# Chinese character part of Japanese as jaKanji...
------- Additional Comments From Masaki Katakai 2002-01-28 10:51 -------
I agree a comment from Momoi,
which is in bugzilla.mozilla.org
Logically, we are correct in the current method.
The problem is;
The problem of jisx201 of Japanese in UNIX platform.
1. jisx201 fonts dont have good ASCII glyphs.
2. only a few typestyles have jisx201.
3. some jisx201 fonts dont have bold or italic styles.
4. there are no proportional jisx201 fonts.
the problem that we dont have bold, italic, proportional fonts
is a big one.
And another is the compatibility problem, which nags many.
It is the comaptibility between NS4.x and old Moziilla.
1. NS4.x uses ASCII fonts.
2. Old Mozilla also does.
So many are uncomfortable with the difference in the appearance.
Personally, I think the current method is correct in a theory,
so it is okay to remain as is.