Lars Gullik Bjønnes wrote: > - All output: > - text (perhaps finished?)
Not at all. I sent a simple patch that almost finished it, but in discussions it turned out that it was the wrong way to do it. Don't you remember? I think we have no choice but use ucs4/docstring for all strings that can contain unicode internally. > - Change the internal gettext machinery (message.C, B_, _) to > use unicode. Could you do that please? I think that Jean-Marc's idea of a temporary _ucs4() function makes a lot of sense. > | 2) How can I fix this endianness problem on windows? Why should the endianness on windows (x86) be different than on linux (x86)? > | 2-a) how do I convert the lyx file to unicode so I can load them in > | English rather than Chinese? > > On window? pttrt. On linux use iconv: > iconv --from-code=latin1 --tocode=utf8 --output \ > UserGuideUTF8.lyx UserGuide.lyx Don't do that, lyx2lyx will do it for you. Reading and writing .lyx files works well. Make sure that you change the file format manually to 249 of all files that were saved in utf8 when the file format was still 248. Georg
