[comp.fonts] fonts with numeric characters of varying widths?

nvi@mace.cc.purdue.edu (Charles C. Allen) (02/28/91)

I was very surprised to find that the Tekton font from Adobe has
numeric characters of varying widths.  This renders it useless for
many intended uses here.  It is the first font I've run across where
this happens.  Do other people have comments on this?  Should I be
surprised?

Charles Allen                           Internet: cca@physics.purdue.edu
Department of Physics                   HEPnet:   purdnu::allen, fnal::cca
Purdue University                       Bitnet:   cca@fnal.bitnet
1396 Physics Building
West Lafayette, IN  47907-1396          talknet:  317/494-9776

rcd@ico.isc.com (Dick Dunn) (02/28/91)

nvi@mace.cc.purdue.edu (Charles C. Allen) writes:
> I was very surprised to find that the Tekton font from Adobe has
> numeric characters of varying widths.  This renders it useless for
> many intended uses here.  It is the first font I've run across where
> this happens.  Do other people have comments on this?  Should I be
> surprised?

It's not surprising to find it in a display face.  I suppose you've hit the
problem because you wanted it as a text face rather than a display face?

It's a type-designer's tradeoff:  It's unnatural to make the figures
uniform width, the "1" being particularly unpleasant.  However, if you
don't, you can't set columns of figures which align properly.  Therefore
the normal decision is to have uniform widths for faces intended for text
use, otherwise possibly not.  I did a quick check on the Adobe faces I've
got; the ones which have non-uniform-width figures are:
	Boecklin
	Cottonwood
	Fette Fraktur
	Freestyle Script
	Trajan
	VAG Rounded Bold
...all display faces.  Curiously, good ol' Zapf Chancery Medium Italic has
uniform-width figures.  I wonder if that was an Adobe change to the face
back in the earlier days of their work.

Some typeface designs have two sets of figures--lining and non-lining (cf
my question last week on readability).  Adobe has this in at least one
"expert collection" I've looked at.  I'd assume that in such a font, the
lining figures would be uniform width and the non-lining wouldn't.
-- 
Dick Dunn     rcd@ico.isc.com -or- ico!rcd       Boulder, CO   (303)449-2870
   ...But is it art?

rcd@ico.isc.com (Dick Dunn) (03/01/91)

A little more about the problem posed by nvi@mace.cc.purdue.edu (Charles
C. Allen):

> I was very surprised to find that the Tekton font from Adobe has
> numeric characters of varying widths.  This renders it useless for
> many intended uses here...

Depending on what sort of software you're using with the fonts, you may be
able to work around this:  You can add a little PostScript to give you a
font which works like the base font except that the figures will be
constant width.  The re-encoding basic idea is in the Red book - they show
just a width adjustment, but you should do side-bearing adjustment also.
For example, pick a width (either digit 0 or widest digit) and set up a
/Metrics entry with [sidebearing,width] entries for each digit, using the
chosen width and a sidebearing calculated as
	original sidebearing - 1/2(new width - old width)
It's possible to write this as a general PostScript procedure which will
gather all the crud out of the font itself (looking for Encoding, doing
stringwidth and charpath/flattenpath/pathbbox for each digit character,
etc.) for any arbitrary font, but it's a lot easier to write something
which massages the appropriate part of the .afm for the font into a (much
shorter) piece of PostScript to adjust that one font.
-- 
Dick Dunn     rcd@ico.isc.com -or- ico!rcd       Boulder, CO   (303)449-2870
   ...But is it art?