[XeTeX] Font variant selection in Fontspec/Mathspec options incomplete

Ross Moore ross.moore at mq.edu.au
Thu Nov 11 20:12:07 CET 2010


Hello Pinfeng (or is it Christian?)

Sent from my iPad

On 12/11/2010, at 4:48 AM, "Christian Schmidt" <c.schmidt at dlgs.ioer.de> wrote:

> Hi Will,
> 
> Thanks for this hint. I updated to this last fontspec version (2.1d) and now it brings no error message anymore and bold italic is set correctly in ordinary text...
> 
> Though, this leads me to the additional note, that in math mode---which is typically set in italics---, the use of \mathbf{x} does not produce a bold italic variant but rather only a bold upright variant of x in this case.

This is the way it has always been in LaTeX.
It should not be changed, else this will affect millions of existing documents.


> If I use the \pmb{x} or \bm{x} commands instead for instance, XeLaTeX produces a bold italic looking variant of the letter x. In closer inspection, however, the bold italic x is nothing else than a normal italic x superimposed on another one with a slight offset...

Yes. This is a hack to compensate for lack of math fonts.
This kind of thing should no longer be necessary with proper OTF math fonts, but you need new macros to refer to the desired characters.
This is what Unicode-math and math spec are for. They completely rewrite how LaTeX handles mathematics with such fonts. It is a massive job, not yet completed.


> 
> Is there a way to use the regular \mathbf{...} command to have bold italic math?

Isn't there a command \mathbfit or similar spelling?

But better is to define a macro for the concepts that require use of such bold symbols;
E.g.  \newcommand{\xx}{\mathbfit{x}} 
and use these within the body of your document.
Be careful to use names that cannot conflict with standard macro names in other TeX or LaTeX packages, even if you do not use those packages. Otherwise you will have trouble when you collaborate with others, or send your work for journal publication, say.

> 
> Thanks,
> Pinfeng



Hope this helps,

        Ross


More information about the XeTeX mailing list