On 1 Oct 2002, Eric Streit wrote: > How will the console know when it has to write up to down, right to left > . > > I can understand with only one langage, so it's possible to choose the > right one, but, imaging a top down writing (Mongolien), how will the os > (linux) write the answers (the prompt, the output of a find research). > And what to do with a text which contains 2 differents langages, like > english and Mongolian . How could any text processing program deal with > such a text. How could it know when to switch between the Two writing > directions, and how ( end of line) etc. And what to do with a really > mixed text ??
I don't know how English text would interact with traditional Mongolian, but in the case of English (or other Latin script) (ltr, ttb) embedded in traditional Chinese (or Japanese) (ttb, rtl), the English text would be rotated clockwise, like shown in the leftmost column of Table 7-10 (p. 352) in Lunde's _CJKV Information Processing_ (1999), or see the leftmost picture in section 3.3 of http://www.w3.org/TR/WD-i18n-format/ (CSS3 module: text; working draft 15 May 2002). It is the best solution for long spans of English texts, since not rotating the English text is not very readable (you only find English text like that in limited strings, like some store signs), and retaining English horizontally only works for maybe one or two letters (i.e., the width of a fullwidth CJK character). On the other hand, the kind of preferred embedding of English within vertical Chinese works better on paper, where you can turn the page to read the English (usually just a foreign name or vocabulary item they want to draw your attention to), than onscreen. But there'll be a clash between English and traditional Mongolian (ttb, ltr), since the flows are not compatible in the way that vertical CJK (ttb, rtl) could bend and become modernized to run like Latin ltr, ttb. Thomas Chan [EMAIL PROTECTED] -- Linux-UTF8: i18n of Linux on all levels Archive: http://mail.nl.linux.org/linux-utf8/
