On 2013-03-22, Uwe Stöhr wrote: > Am 20.03.2013 21:58, schrieb Guenter Milde:
>> Thinking about it, I realize that also the basic support for other languages >> is optional -- my Debian system has lots of "texlive-language-*" packages. >> I don't think we should completely disable using a language that is not >> properly supported, but a warning would be in place. > How would this be possible. We would then have to check if the ldf-file > exists. But this can exists anywhere and we don't have access to the > LaTeX-file database. This warning only makes sense for babel, not for > polyglossia. The check should be done during LyX (re)configuration. AFAIK, there are lots of tests for LaTeX files, so this must be doable. >>>> So, maybe for the time being it could be better to just exclude >>>> Kazakh from the "Cyrillic languages" to let the "textcyr" feature >>>> work. With the current state of textcyr, the fix should be far more easy --- change the default latex input encoding to ASCII. > Today I asked the Kazakh colleague and he told me that nobody yet uses > Latin script for Kazakh. There had been some attempts by the > government but without success. So we can treat Kazakh as Cyrillic > language. >> Yes, Kazakh seems only supported with Polyglossia and should be marked as >> unsupported with Babel. > But Kazakh is not supported by polyglossia. I already marked Kazakh as > unsupported by babel _and_ polyglossia. In this case, I'd comment the whole Kazakh section out: # not supported by babel, not yet supported by polyglossia # Language kazakh # GuiName "Kazakh" # BabelName kazakh # Encoding ascii # LangCode kk_KZ # End Mind, that global language options are safely ignored, but in-document language changes require a valid language. I.e. you may have a Kazakh document that works but never a Kazakh word in a non-Kazakh document. >> Some test revealed, that the "textcyr" and "textgreek" features do not >> consider the current language. > I see now but I am not familiar with this code part of LyX. Perhaps > Enrico can state about this. I CCed him in your bug report. OK. > However, what do you think about supporting Friulan? I'd wait until a) there is a feature request, or b) there is a check for the language definition file. Otherwise, you add a "trap": you can easily mark a word as Friulan but this makes your document uncompilable with Xe/LuaTeX as well as for most TeX installations. Günter
