ud at art-satz.de
Mon Aug 25 15:21:31 CEST 2003
I've recently thought about the kerning problem in T1 or LY1 encoded TeX
fonts. As I'm not very experienced with fontinst I would like to ask the
experts if there're some kind of high level commands to simplify the
task of generating kern tables for T1/LY1 encoded fonts.
The point is of course that most font companies had 'forgotten' to
supply kerning tables not only for the 7-bit part (exception: the new OT
fonts). I think of something comparable to Adobe's class kerning in OT
fonts. Otherwise it will be to much work setting up a proper kerning
table for T1/LY1 fonts which not only works for standard pairs but also
with accented chars. And T1/LY1 fonts will always be somewhat
inconsistent to have pairs, e.g., for 'Ta', 'Te', 'Ti', 'Tu', etc. but
not for 'Taacute', 'Teacute', etc. AFMTOTFM by YandY does something via
a command line switch. But this has the serious draw-back that ALL
accented characters will get the same kerning values as the unaccented
counterparts (all-or-nothing principle). Which is of course not
desirable, just look at 'Yi' and 'Yigrave'.
I think---if it's not already there---this would be a very useful
fontinst feature especially for high quality setup and maintenance of
Easiest would be to have some kind of table like below
%left char(s) right char(s) value
T [a aacute atilde aring ...] -80
T [adieresis agrave acaron acircumflex ...] -30
[V W Y] [c d m n ...] -75
-- yielding -->
T a -80, T aacute -80, T atilde -80, T aring -80, ...
T adieresis -30, T agrave -30, T acaron -30, T acircumflex -30, ...
V c -75, V d -75, V m -75, V n -75, ...
W c -75, W d -75, W m -75, W n -75, ...
Y c -75, Y d -75, Y m -75, Y n -75, ...
What do you think about it?
More information about the fontinst