MIME-Version: 1.0
References: <199703181933.TAA09294@rasha.math.uci.edu>
            <199703232040.VAA25736@puma.npc.de> <3341B518.2781@math.uci.edu>
            <199704051803.UAA09645@puma.npc.de>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-ID:  <3347E43F.41C6@math.uci.edu>
Reply-To: Mailing list for the LaTeX3 project
              <LATEX-L@RELAY.URZ.UNI-HEIDELBERG.DE>
Date:         Sun, 6 Apr 1997 10:58:23 -0700
From: Marcel Oliver <oliver@MATH.UCI.EDU>
Sender: Mailing list for the LaTeX3 project
              <LATEX-L@RELAY.URZ.UNI-HEIDELBERG.DE>
To: Multiple recipients of list LATEX-L
              <LATEX-L@RELAY.URZ.UNI-HEIDELBERG.DE>
Subject:      Re: International documents
Status: R

Joachim Schrod wrote:
>
> >>>>> "MO" == Marcel Oliver <oliver@MATH.UCI.EDU> writes:
>
> MO> 2. There may be situations where letterspacing is
> MO> required (by some higher power), and TeX/LaTeX is
> MO> not well equipped to satisfy such demands.
>
> Why? TeX/LaTeX is well equipped for such tasks. If one needs it, one
> needs a font that's specially designed for this demand (ligatures must
> be taken care of, etc.) Then one switches the font for emphasis, just
> like wo do normally. VF files might be a first step towards such
> fonts, though I'm rather sure that stems must be changed if we change
> glyph distances.

That's what I mean.  It requires an effort which is in no
proportion to the desired effect.  In two instances where
I was hacking some class files in the past which demanded
letterspacing (for reasons beyond my control, which I
therefore don't want to justify) I lost lots of time getting
it to work.  Both instances could eventually be handled
by Phil Taylor's letterspace package, or modifications
thereof, but I wouldn't call this "well equipped", it's
essentially a clever hack.

Moreover, switching to letter-spaced fonts has two
disadvantages:

- A class cannot be written independent of the fonts
  used in the document (unless a letterspaced font family
  will become part of the standard font set, which is
  probably overkill since in most instances letterspacing
  is considered "bad" typesetting anyway...)

- Apart from the example I gave where letterspacing
  is used as emphasis (which I mentioned not because I
  want it, but because I have seen it used by some
  "authority"), the amount of spacing may not be known
  in advance. Therefore there may be a need for run-time
  letterspacing.

I think there is no reason for being dogmatic
about letterspacing, because TeX/LaTeX may be
used for purposes not originally imagined by the
authors.  It should be at the discretion of the
class or package designer to employ or not to
employ particular features, not of LaTeX itself
to make certain features practically a pain.

  Marcel

> PS: On the topic of typographical quality, Marcel wrote
>
> MO> At least the books by Taeubner Verlag do it like this.
>
> If you mean the Informatik-Reihe -- they don't typeset at all, but
> take camera-ready copy from the authors. I don't know about other
> series, but I would not take the output of a computer books publisher
> as a shining example for typesetting. Several well-known large
> publishing houses belong to my customers -- many producers there
> (responsible for house styles, etc.!) don't even know what a Geviert
> (german for `quad') is... Neither would I take the Duden as a model,
> as brought up by Richard. By explicit will of its authors its content
> is descriptive, not prescriptive.
>     Books like `Kaeufers Setzerlehrbuch', used for education of
> Typesetter's Masters (correct? Setzermeister), books by Tschichold,
> Aicher, Morris, etc., are better to learn from.

I am talking about Teubner's (sorry about the spelling in my
previous mail) Reihe "Mathematische Leitfaeden", in particular
Heuser: Funktionalanalysis and Wloka: Partielle Differentialgleichungen.
Both boods are typeset books from the 80s of this century, not
"camera-ready" low-budget productions.  I leave it to the experts to
judge the design.