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

Re: PostScript font installation: my evolving tools...



Concernant « Re: PostScript font installation: my evolving tools... », Melissa O'Neill écrit :
« 
« Just what is a skewchar/skewkern anyway, and how does one ensure that
« it is honest?
« 

with an eye! that's where computer fail! skewchar is usually specified
by the encoding you intend to comply with. skewkerns shift the accents so
that they look good (scaling what i'd compute as the italic correction
[which for me is simply the |sinus| of the italic angle times the
height of the char] is usually honest).

« I have two ideas for creating math letters. One is to follow a scheme
« like Thierry's, which is to decide we are only prepared to tollerate
« negative sidebearings up to a certain limit (I think it may be wrong
« to elimiate them entirely, `f' in CMMI10 has a right sidebearing of
« -62 AFM units and `j' has a negative sidebearing of -13 AFM units).

well remember that the `width' of f in cmmi is not its advance width
but only determines the position of subscripts, the actual advance
width is rather that declared width + the italic correction (if i
recall well, sure that Berthold Horn will fix me). hence it's natural
that it is not symetric. What is crazy with TeX's math is that you
have to disasemble TeX's behaviour to produce metrics that are not
intuitive at all...

There are 2 endeavours on which i choked (because of my weak
abilities, as always): the first was to have a real working math font
tuned for tight setting in math in any purposes (as required with
minion e.g.) i indeed made something similar  to what you imagine
below: make the math fonts look metrically similar to cmmi, so that
TeX behaves honestly with them, leaving much too much loose
spaces. The second one is to have a working set of math fonts with old
style digits yielding proper spacing & placement of
super/sub-scripts. I didn't understand where they failed, but all my
experiment failed (although nothing beats 2n+3\pi/4 in OsF!).

« 
« The other is to take a pair math and text fonts (e.g. CMMI10 & CMTI10
« or LucidaNewMath-Italic & LucidaBright-Italic), which I shall call the
« source fonts, and attempt to steal their wisdom and apply it another
« font, which I shall call the candidate font.  First, we make the candidate
« text font `more like' the source text font.

   Thierry Bouche.       -----       thierry.bouche@ujf-grenoble.fr
          http://www-fourier.ujf-grenoble.fr/~bouche/