Re: [ft-devel] GX support now working

[Alexei Podtelezhnikov]

>>>> Looking more, looks like both TrueType and Type 1 / CFF use non-zero 
>>>> winding
>>>> rule. (I'm relieved!)  
>> There is common 'fill' and less common 'eofill' in Postscript.
> 
> Right.  But the question was about Type1 and CFF font formats, not Postscript
> itself.  Thanks to Dave Lemon, we now know that Postscript-flavored fonts
> leave the winding rule unspecified... 

It escapes me why the even-odd rule would ever be desirable. The XOR like 
overlap of otherwise correct shapes (without self-crossing in proper 
orientation) is weird in our context. Such overlap is lesser of the three 
errors.