[XeTeX] Polyglossia, babel and xelatex

[Ulrike Fischer]

Am Wed, 11 May 2011 17:57:13 -0400 schrieb Alan Munn:

> In another thread, I apparently provided some misinformation about
> whether babel was compatible with xelatex or not. I was
> apparently under the false impression that you should always use
> polyglossia with xelatex.   I'm starting a new thread to ask the
> following questions:

> Can someone clarify exactly when babel should or shouldn't be used with xelatex?

No ;-). It depends on the languages involved. Each language
definition file is an independent "style" file, often written by
different authors. They even can clash when used together (french
e.g. had some problems with ngerman regarding \nonfrenchspacing),
and if and how they clash with xelatex and/or polyglossia must be
decided individually.

Until now xelatex + babel + ngerman works fine for me.

> Can someone clarify when polyglossia is the only solution (if that's ever the case)?

When non-western scripts are involved. In this cases the babel
language files include font encoding switches (e.g. greek uses LGR,
russian some cyrillic encoding). This clashes with the use of
"unicode fonts" and fontspec. 

It can still be possible to use in a document e.g. babel + ngerman
and polyglossia + greek. But without trying it is not possible to
say.

The main problem with polyglossia is that for some language (e.g.
french) its language file is much less sophisticated. The other
problem is that an adaption for luatex is needed.   

-- 
Ulrike Fischer