[XeTeX] fontspec v1.1

Bruno Voisin bvoisin at mac.com
Tue Oct 19 08:39:02 CEST 2004


Le 19 oct. 04, à 04:37, Will Robertson a écrit :

> Well, it's not like I invented the encoding itself, just the name. 
> After all, XeTeX uses unicode. The only advantage of using a separate 
> encoding from Unknown is that we get a little osxenc.def file to play 
> with. At the moment all it says is
>
>> \ProvidesFile{osxenc.def}
>> [2004/10/04 Encoding definition for Mac OS X fonts]
>>
>> \DeclareFontEncoding{OSX}%
>>    {\newcommand\UTFencname{OSX}%
>>    \RequirePackage{utf8accents}}{}
>> % This is all that is declared since Ross Moore's utf8accents.sty
>> % package supercedes the LaTeX functionality of \DeclareTextAccent,
>> % \DeclareTextSymbol, etc. (I assume).
>>
>> \endinput
>
> Whereas in the other encoding files it contains all of the commands 
> that define accents and so on, which is what utf8accents does. Users 
> shouldn't need to know about utf8accents as far as I'm concerned - all 
> it does is make XeTeX behave as they expect. (When people switch 
> encodings in LaTeX, they don't need to say \usepackage{t1accents}, for 
> example.)

Ah, now I understand, the OSX encoding means utf8accents.sty (whatever 
its name may evolve to in the future) is implicitly loaded. That makes 
sense.

> But I think the encoding matter should be sorted out as soon as 
> possible so that we don't start proliferating files with an encoding 
> name that may change in the future. As Ross says, there is nothing in 
> the font definitions that would change if XeTeX was ported to another 
> platform, so maybe my first choice of "OSX" as the encoding name was a 
> bad bad idea.
>
> Perhaps XU or XUCS would be a better idea --- "XeTeX Unicode", or 
> something along those lines. Whom does one consult when creating new 
> encoding names?

I've tried to look into teTeX's doc with not much luck. Inside 
/usr/local/teTeX/share/texmf.tetex/doc/fonts/fontname/fontname.dvi you 
find, at page 23, "9u" for Unicode-compatible, but that's, I think, for 
the TFM and VF file names in standard TeX installs, not for the 
encoding itself.

There's the new [UTF8] option of {inputenc}, but I couldn't find it 
documented inside 
/usr/local/teTeX/share/texmf.tetex/doc/latex/base/inputenc.dvi (I'm all 
speaking of TL2004, Gerben's new experimental setup still in beta 
test). Looking inside /usr/local/teTeX/share/texmf.tetex/tex/latex/base 
you find files ot2enc.dfu, t2{a,b,c}enc.dfu, x2enc.dfu connected with 
UTF-8 encoding, but I couldn't find inside utf8.def how their names are 
generated when calling them, what relations the names OT2, T2 and X2 
exactly bear to UTF-8.

So in summary I don't know which encoding name, if any, is supposed to 
be associated with UTF-8 in standard LaTeX.

(A helpless) Bruno



More information about the XeTeX mailing list