[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

cmex5,6 and cmbex parameters

Hi all,

some weeks ago when Matthias released version 0.58, I raised some
questions concerning the choice of METAFONT design parameters for
cmex5,6 and cmbex5-10.

While I more or less suggested so start out with cmr5,6 parameters 
and just add the appropriate fontdimen code, Matthias went further 
and applied some changes to |cap_stem|, |cap_curve|, |cap_ess| 
and |jut| based on he differences between cmex7,8 and cmr7,8.

Now, guess what I found when I had another look at the font programs
of big operators and big delimiters in Volume E?  Suprising as it 
may seem, it turns out that none of these parameters plays any role
in the design of the extensible characters.  In other words, it simply
doesn't matter what values one choses, since the designs are based 
on the lowercase parameters |stem| and |curve|, while the serifs on 
the product and coproduct signs depend on |cap_jut| rather than |jut|.  
Who would have guessed so without looking it up?  Certainly not me!

BTW, In case you're interested, the story is different it comes to the
Euler extension font.  Here, the differences between euex and cmex
affect the parameters |tiny| and |hair|, which do indeed play a role 
in the design of the Euler-style integrals and summation sign.

Cheers, Ulrik.


P.S. Speaking of the Euler layout, did anyone notice that the bold
`less' and `greater' symbols from eurb10 look quite different from 
the non-bold version from eurm10?  Unless there's some serious bug
hidden somewhere in fonts/ams/euler/eurbch.mf, I'm tempted to say 
that these characters look like they were designed to be used as
French quotes in text rather than as math symbols.  Anyone knows?

IIRC Knuth does mention in ``Typesetting Concrete'' that half the
characters in Euler, i.e. the bold version, remain untested and
haven't been debugged yet, but I'm not sure if he was referring 
only to glyph placement  or the Metafont coding in general.