Abdelrazak Younes wrote:
Martin Vermeer wrote:
On Sun, Oct 28, 2007 at 11:56:51AM +0100, Abdelrazak Younes wrote:
Martin Vermeer wrote:
On Sun, Oct 28, 2007 at 09:37:36AM +0200, Martin Vermeer wrote:
On Sun, Oct 28, 2007 at 02:27:25AM +0200, Dov Feldstern wrote:
Martin Vermeer wrote:
...
I would not be happy about this as I said. Elaborating, the
problem of needing in this case to switch encodings is due to
a LaTeX limitation. If LaTeX were Unicode-capable, we could just
write, and the typed chars would come out right by
themselves.
AFAIK, This is one of the goal of XeteX which will be supported in the
upcoming MikteX-2.7.
Correct. XeTeX is definitely something to watch in this area. But it's
going to be a while before we can really switch to XeTeX, and until then
we need solutions.
Switching based on characters would allow us to automate the switch
when Unicode is used, so we don't have to (ab)use language for that.
This is what I proposed before 1.5.0. You're right, for example you
don't need to know the language to properly typeset Farsi or Arabic (at
least in theory), the same thing applies to CJK. At the time Dov and
Georg had some very good argument in favor of setting the language
unconditionally though (and spellcheck was one of them).
Well, LaTeX needs language information in order to typeset correctly.
Even for things like spacing around punctuation, which is different
between French and English, even though they both use the same encoding.
Certainly for things like RTL/LTR. So language is necessary information.
So is Encoding. Encoding can perhaps be inferred from the characters
themselves, but language cannot. (This Sonnet project sounds
interesting, I am not familiar with it. Thanks for the link, Abdel!)
However, this is not our real problem right now. Our real problem is
being able to transmit the language and/or encoding switches to latex.
Once I know how to do that, enabling the language and encoding switches
in LyX will be relatively simple...
Dov
