> > Of course I don't insist "variant" problem is unique for CJK world,
> > since I don't know other languages. It is only on my native language
> > that I can talk with self-confidence. Hoever, even if there are some
> > languages or scripts which have same problem as CJK, the fact won't
> > deny CJK people's need. I.e., giving such examples is a wrong approach
> > to persuade CJK people.
>
> _Yes_ we know that language-dependent glyph selection needs to happen
> otherwise text may be illegible.
I don't even think it needs to happen at the XTerm level.
Having a unified single-style Han font is probably the optimal solution
for this level. All the visually distinctive glyphs have individual
codepoints in Unicode anyway. The only thing that is really annoying to
the normal CJK reader*1 is the switch between different fonts. There
is however no necessity to consider fonts as peculiar to one of the
CJK nations. Most existing and future fonts look nice and well
readable to all East-Asian nations. There is no need to have Msmincho
on a Japanese xterm.
> Please, work with us to fix this in a proper way. There is no need
> for the paranoid accusations that have been made. This is even needed
> for getting DIGITS displayed properly for urdu vs persian.
> How does this sound as an interim solution :
>
> * xterm will look at LANG (or whatever), and then use that in
> ADD_STYLE_NAME.
The most important thing is to have a complete Unicode font with
a consistent Han character style. If we have such a font in Msmincho
style, we can of course make that the default for Japan. However,
if there is no Taiwanese style equivalent, then we should rather use
a Msmincho-Unifont for Taiwan than use a big5 font. Completeness should
take precedence over regional style pecularities.
> (So if I do
> LANG=ja xterm -u8 -fn 9x18u
>
> It will pull out
>
> "-misc-fixed-medium-r-normal-ja-18-120-100-100-c-180-iso10646-1"
>
> And if I do
> LANG=ko xterm -u8 -fn 9x18u
>
> It will pull out
>
> "-misc-fixed-medium-r-normal-ko-18-120-100-100-c-180-iso10646-1")
>
> This is applicable for solving the problems with
>
> * zh/ja/ko
> * pl/fr
> * ur/fa
> * ro/tr
> * sr/ru
> * coptic/el
> * lots of other ones about how diacritics over latin letters are
> positioned
>
> We can work on a better way later, but we need people to work with us on
> that. For example, there is still no 18x18zh font, nor an exhaustive list
> of characters that have significant differences. (At least not that I am
I don't know of any single character that has variant differences
significant enough to pose a problem at the XTerm level.
> aware of). That's something people could be doing that would be far
> more valuable than complaining about semantics on a mailing list.
indeed.
-phm
-
Linux-UTF8: i18n of Linux on all levels
Archive: http://mail.nl.linux.org/lists/
