<[EMAIL PROTECTED]> writes: > On 22 Jan 2001, zhaoway wrote: > > > It's not just locale. Say, if I want read Chinese and Japanese at the > > same time on the same XTerm, UTF-8 will do it, GB18030 won't. Glibc > > can of course even support GB2312, but if XTerm use GB2312 locale, it > > won't be able to read Japanese (whatever encoding) then. So you will > > have to use UTF-8. Then Gb18030 has quirk. > > GB18030 of course can do it! I can edit Chinese, Japanese, Korean, > Russian and many other language in a gedit window with GB18030 locale!
Oh, yeah. To summarize it, you have better chance to bet all of the other part of the world will most probably prefer UTF-8 than GB18030. Back to the problem, if every good country will get out their own set of standards to do what UTF-8 being supposed to do, then suppose I download an article describing MULE techinicals with a Chinese filename encoded in, unfortunately, the _good_country_abcde_'s STD-12345, now back to the problem, what will GB18030 locale do to help? If you're using wget to get that article, your command line with that filename is probably in a horror, and even worse, the filename of the file you've just downloaded onto your filesystem will neither be in a good state either. And the worst things is, how will you input that filename encoded in the _good_country_abcde_'s STD-12345 from your GB18030 locale? So you know that is Chinese, you can read it, you recognize the nice glyphs, and you can't even get it! 8-P So, again, my question is, what does GB18030 provide to us, which cannot be solved with UTF-8, or Unicode surrogates? (The current version of Unicode is not perfect, I agree. But there're no fixed obstacles there, fortunately. And I agree the pressure from China on Unicode is also good and necessary too.) > > Oh, man, FontSet is cool, but even cooler is UTF-8 locale and iso10646 > > fonts. FontSet can, cough, _not_ support GB2312 and Big5 in the same > > IRC window. If XChat use UTF-8, and we all use UTF-8, then we can > > (people from HK, TW, CN) chat at the same time in #debian-zh. Man, > > it's not locale here that matters. It's the distinction of characters > > here I'm talking about. > > I use ISO10646-1 font under GB18030 locale, but the problem is, there > are very few complete ISO10646-1 font available. We should use fontset > instead of a incomplete ISO10646-1 font for a UTF-8 locale either. > And not all of us can use UTF-8 locale on there system. As far as I can say, the _on-the-fly_re-encoding_ layer of the new XFree86 4 will have a far more better chance to win out. Which means, you can (will?) use -gb2312.1980-0 + -big5-0 + -iso8859-1 + whatever as an on-the-fly -iso10646-? font which will eliminate all the needs for the FontSet. And with this technique being introduced into XFree86 (IMHO, _this_ is the state of the art of XFree86, BTW) you will be able to use -iso10646-? fonts in GB2312 locale too. So with all of these questions, if anyone could tell me the benefits of using GB18030 over GB2312, UTF-8, etc. I will thank you very much. -- zhaoway -- | This message was re-posted from [email protected] | and converted from gb2312 to big5 by an automatic gateway.
