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

Re: How should one handle condensed etc., variants?



>I was intrigued to read Rowland's comments on the limitations of the
>NFSS font selection scheme, especially his remarks on how font weight
>and width are lumped together in the ``series'' specification of a font.
>His thoughts coinside remarkably with similar recent findings of my own.
>
>The issues with weight and width are actually fairly well covered in
>_The LaTeX Companion_ (page 190) where it explains how the two parts
>are to be combined (weight is concatenated width, dropping any `m's,
>unless weight and width are both m, in which a single `m' is used).

I see - thanks.  Erm, to save me getting a copy out of the library (which
takes a few weeks), and then finding that it doesn't have the information I
need, is there a standard list of abbreviations used to width?

>But, neither _The LaTeX Companion_, nor any of the documentation I've
>found, seems to mention what to do if one has a font which fits more
>than one of the available (standard) font shapes (e.g., Italic Small
>Caps, for which I currently use {itsc} in my FD files). Combining two
>shape attributes into one raises similar issues to those Rowland was
>talking about with weights and widths -- it is a less than ideal solution.
>If I want to typeset an acronym (using \textsc) in the middle of of an
>emphasized piece of text, it shouldn't set the it in roman.

True enough - I'd say the NFSS needs at least three more selection axes:

weight: to allow for the selection of bold, light, etc., separate from the
informal/other variants there might be.

angle: to allow italic/slanted/upright/etc to be selected separate from
shape (small caps, or whatever)

width: to allow condensed/extended to be selected separate from series.

This doesn't take into account old style variants, which I've never used
and never really thought about.

So rather than:

   encoding, family, series, shape (now)

I suggest:

   encoding, family, series, weight, shape,  angle,     width

e.g.,

   T1         ptm   informal m/bx    sc/n   it/n/sl/ui  cn/ex

Anyone care to comment?

>Of course, things get even stranger when it comes to things like old-style
>digits, where for some reason the variation in shape gets put into the
>name of the font, which seems really wrong to me. (BTW, while I'm on
>old style digits, does anyone know if fontinst has the smarts to take
>the dollar, cent, percent, sterling, and any other symbols that need to
>be different for a text using oldstyle digits, from the `Small Caps and
>OldStyle Digits font' if such a font is available? [Or, at the very
>least does it take the dollar and cent signs out of an available expert
>font.])

I'm pretty sure that it won't on its own - I think you need to play about a
bit.

What I'd try is:

\installfont{fxxr8t}{fxxr8r,unsetstuff,fxxr8x,latin}{T19}{fxx}{m}{n}{}

The file unsetstuff.mtx needs to contain:

\unsetglyph{textsterling}

etc., to tell fontinst to forget that fxxr8a has set the glyphs you want
used from the file fxxr8x.mtx.

It might also work if you said:

\installfont{fxxr8t}{fxxr8x,fxxr8r,latin}{T19}{fxx}{m}{n}{}

If fxxr8x didn't set any glyphs you *need* from fxxr8a

Rowland.