[XeTeX] stacking diacritics without mark-to-mark

[Mike Maxwell]

On 3/12/2014 10:19 PM, Andrew Cunningham wrote:
> Although personaly I'd consider such a solution a poor hack compared to a well
> designed font that is fit for purpose.

I won't disagree.  We've been told by the publisher, who original built the font 
(or more likely subcontracted it) that the problem is a shortcoming of their 
OpenType font files, that could only be eliminated by revising the font as a 
whole.  I'm inclined to tell them to go for it, or else we'll use a font that 
does work (Charis SIL comes to mind).  That would have the unfortunate 
consequence that the first book in our series would use the publisher's font (we 
didn't have the stacked diacritic problem in that book), but the rest of the 
books in the series would use some other font.

But in order to convince them to make such a change, I need to understand better 
what is involved in revising a font to make mark-to-mark positioning work.  I 
gather that only the diacritics need the two mark-to-mark points (top and 
bottom) to be defined, not every character--correct?

Is it done at the character level, or at the glyph level?  Does it need to be 
done separately for each point size/ weight/ style that the font supports?

And do those attachment points need to be defined manually, or is there a way to 
automate that process?  Can this be done in a tool like FontForge, or does one 
need specialized tools that only the font manufacturer would have?

I'm assuming that base characters already have top and bottom marks, but that 
may not be a safe assumption.  The problems we've seen so far have been with 
stacked diacritics.

And of course there may be a better forum than this to ask this question.
-- 
	Mike Maxwell
	maxwell at umiacs.umd.edu
	"My definition of an interesting universe is
	one that has the capacity to study itself."
         --Stephen Eastmond