[XeTeX] Using hebrew

[Will Robertson]

On 07/03/2007, at 22:33 , Peter Dyballa wrote:

> I read your answer. I also remember that it's possible to choose one
> of two renderers. When I am sure, and I can proof this with my
> source, that I did not use either of these explicitly, how can XeTeX
> report an inexistant error?!
>
> To me this indicates that fontspec has a flaw in reporting inexact
> error causes. Or causing inexact error messages.

?? One of the renderers has to be used even if you didn't specify  
one. And choosing one renderer over another will not cause XeTeX to  
report a missing font error message!

>  From one year ago when I was experimenting with the Lucida fonts in
> Java?

Of course. Don't you like bugs that you report fixed? I mentioned it  
in the announcement here, I'm pretty sure.

> I'd recommend to hide /AAT or /ICU from the error message! It
> can lead into wrong directions ...

That's not really possible. The choice between fixing a bug that  
doesn't happen very often vs. confusing people who don't know much  
about XeTeX was already made long ago, and I don't want to change  
things for the sake of it.

> Isn't there the command \XeTeXfonttype that could simply return -1
> when there is no font of the given name available (for the renderer,
> if specified)?

It is *very slow* to query whether a font exists. I believe that's  
why Jonathan hasn't implemented such a feature.

Will