[XeTeX] \ifFontExists

[Mike Maxwell]

Apostolos Syropoulos wrote:
>> But how would we define "matches closely"? How do you measure the  
>> "closeness" of fonts...
> 
> This is an excellent topic for someone willing to apply fuzzy set theory to
> typography. The basic idea is to define an ideal font (i.e., a font where
> glyphs have the most standard form) and then to define similarity
> degrees, etc. 

Wasn't Donald Knuth's original concept of 'metafont' something along 
these lines?  And of course Douglas Hofstadter has worked on analogy, 
testing his ideas on fonts (his "letter spirit" project).  In both 
cases, though, I believe the idea is to *generate* fonts, not recognize 
similarities among existing fonts.
-- 
    Mike Maxwell
    maxwell at umiacs.umd.edu

    When he approached fundamental questions of astrophysics,
    Eddington did not try first to establish basic laws from
    which conclusions could be deduced (as Newton had done and
    as his rival James Jeans insisted on doing), rather he
    built a web of approximations whose results could be
    compared with nature.  --review by Owen Gingerich of
    Practical Mystic: Religion, Science, and A.S. Eddington