[XeTeX] XeTeX for Linux (first experiencies)

[Yves Codet]

Le 8 juin 06, à 13:26, public at heslin.eclipse.co.uk a écrit :

Sorry about the late reply. My silly antispam misinterpreted something 
in your message (something to do with LaTeX?).

>
> I'm afraid I think that the difference between the simplistic Babel
> patterns in grhyph.tex and Filippou's patterns (which are the ones
> called elhyphen -- has Robin got confused about this?), are that the
> Babel patterns are wrong, and Filippou's are right (or at least try to
> DTRT).  I would prefer holding off at the moment uploading a solution
> that we know to be incorrect.

Robin didn't mention "elhyph.tex"; the "ellhyph-ety.tex" suggestion is 
mine. Maybe I'm perverted by Sanskrit; as far as I know, no edition of 
Sanskrit texts ever introduced etymological patterns which are totally 
unknown in Sanskrit manuscripts. So I'm inclined to think that 
"grhyph.tex" patterns aren't wrong; some people care about etymological 
patterns, some don't (which is my case, particularly if manuscripts 
don't use etymological breaks). One could argue that hyphenations like 
ὑπε-ραλγῶ are misleading, but only for beginners (and maybe beginners 
who are still mislead by such hyphenations should study handbooks, not 
texts :). That's why I think it would be a good things if users had the 
choice between etymological and non etymological patterns, according to 
their taste.

> As for the issue of ancient vs. modern, IIRC the mechanisms of Babel
> require that greek and polutonikogreek share the same hyphenation
> patterns, since they are defined as dialects of the same language.  
> This
> restriction is one of the many unfortunate aspects of Babel's handling
> of Greek.

I agree. Ancient and modern Greek should be separated in the 
"substitute_of_babel" package for XeTeX.

Best wishes,

Yves