[XeTeX] Fontspec bug?

Will Robertson wspr81 at gmail.com
Thu May 18 00:08:22 CEST 2006

Good morning!

Thanks Jonathan, for your quick reply.

And I'm terribly sorry, Bruno, for the inconvenience. I've been a bit  
busy lately and while I've a new fontspec in the works, there's a  
couple of unresolved problems that I've yet to track down (who knows  
where they came from!).

On 18/05/2006, at 2:47 , Bruno Voisin wrote:

> However, even in this case I'm not quite certain that the use of
> Lucida should dictate whether or not the text font (not necessarily
> Lucida, for example here) is used for typesetting numbers in maths
> mode. I remember the introduction of the legacymaths mechanism in
> fontspec resulted from the use of the operators font for some
> diacritics in LaTeX by default, but I must admit I don't remember the
> whole story.

The approach I've taken is to use the legacymaths encoding to  
override everything that was previously assumed to be from the  
\mathrm font. This means Computer Modern will always be used for  
maths --- unless, that is, the maths font has been changed; for the  
time being, I'm aware of only Euler and Lucida that don't use virtual  

So in the fontspec code it goes something like this:

   override some stuff only for euler
     override some other stuff (actually, nothing!) only for lucida
     override everything!

Somehow, that \else slipped out somewhere along the line. And I've  
got no idea how I didn't spot the problem in the first place :(

Apologies again,


More information about the XeTeX mailing list