summaryrefslogtreecommitdiff
path: root/info/digests/texline/no11/ap.tex
blob: f40a27b423ae7f91399395d725e9fa48327ac3f7 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
%\def\rbox#1{%
%\leavevmode\setbox0\hbox{\strut#1}\special{dvitops: gray reverse 1}%
\def\yen{Y\kern-0.65em\raise0.25ex\hbox{$\scriptscriptstyle=$}}
\def\quotedblbase{\lower1.4ex\hbox{''}}
\font\teeny times at 4pt
\font\eight cmr8
\font\larger helvetica at 16pt
\def\mark{$\triangleright$}
\def\TA{\TeX{}-APS}
{\parindent7pt
\centerline{\bf A brief review of \TA\ fonts}
\smallskip
\noindent
If there is a gap in the world, someone somewhere will fill
it sooner or later; the gap in question here is the choice of fonts
available to a \TeX\ user. Donald Knuth's remarkable achievement in
creating a parameterized font creation system, and a complete example
family called {\em Computer Modern Roman}, has not impressed the
commercial typesetting world who want both their old favourite fonts,
and (more importantly) a choice of typefaces to suit every occasion.
Knuth's fonts suit one sort of occasion (mathematical academic papers)
extremely well, but they have not matched the visual variety available to
users of `professional' typesetting machines. It is to the enormous
credit of Adobe that the fonts they licensed with most \PS\
interpreters {\em have} satisfied the vast majority of users. So why
are we \TeX\ users not all taking advantage of \PS\ fonts?
There are various reasons:
\item{\mark} Everyone who
produces a \PS\ printer has to pay a fee to Adobe, and probably
additional fees for fonts; this cost gets passed to the user.
\item{\mark} Even cheap laser printers are beyond the purses of many people,
who must use dot matrix printers.
\item{\mark} Most screen previewers and editors still rely on bitmap font technology
of a low standard, even in modern graphical user interfaces -- the
vast majority of X Window or Microsoft Windows users, for instance, still
suffer from ghastly fonts; use of a \PS\ printer produces a
mismatch between screen and page. Obviously there is a {\em lot} of work
in this area (Folio and TrueType scaleable fonts, Display \PS\
implementations like the NeXT and Sun's OpenWindows), but for most \TeX\
users, it is still very convenient to get their fonts from the same
source (\MF) for screen and previewer.
\item{\mark} Even those who {\em have} found the fonts on their \PS\
printers have difficulty in adapting \TeX\ to their use; again, there
are plenty of examples of how to do this (the work of Frank Mittelbach
and Rainer Sch\"opf on font selection schemes, reported in recent
\TUGboat s, is clearly the way to go), but still the issue drags on. Not
the least of the issues is how to logically lay out fonts -- in the way
\TeX\ does, the way Adobe does, or following some international standard?
\smallskip
\noindent
Into this maelstrom leaps Richard Kinch, energetic creator of Turbo\TeX,
with a compromise -- \TA\ fonts. These are a set of the `standard' \PS\
scaleable typefaces fossilized into \TeX\ {\tt pk} bitmap formats at
commonly used sizes from 5\,pt to 25\,pt. This means you can use them with
any device for which you have a {\tt dvi} driver (screen or printer).

The \TA\ fonts arrive on 10 360\,k MSDOS floppy disks at a cost of abour
\quid 200; they are  also
available on 720\,k disks. They come with a good clear installation
manual, and soon flood your disk with 6\,Mb of files. The widespread
`zip' compression, and `cpio' archiving sofware, are used, and source
for these programs is provided; the user can either follow the
recommended installation by typing a few simple commands, unpack the
fonts by running the utilities by hand, or install the utilities on
another operating system and do the work there -- I found it easy to
perform the unpacking on a \Unix\ workstation, for instance. It was a
pleasure to find this clean and comprehensive set of installation
options.

The initial release of the fonts (1.0) had a serious flaw in the 
design sizes built into the files; this is fixed in release 1.1, and 1.0
purchasers get a simple program to run which fixes matters. I could find
no other problems.

Along with huge numbers of {\tt pk} files come a set of {\tt tfm} files
for \TeX\ to digest. Kinch has made a number of far-reaching decisions here:
\item{\mark} The layout of the fonts is the same as that of {\tt cmr} for all
the normal fonts; the naming and characteristics of the fonts seem to
correspond with the practice of ArborText. This layout is {\em not} the
same as the default layout of \PS\ fonts.
\item{\mark} For compatibility with older drivers, there are only 127
characters in each font. \PS\ fonts have 256 or more glyphs, so Kinch
has had to abandon many characters; these are mostly composite
characters like \"A, but also include many useful odd things like \yen\
or \quotedblbase.
\item{\mark} There is no real advice on how to integrate the fonts into daily
usage to replace Computer Modern -- it is curious that the manual is
typeset in CM! The examples given are for plain \TeX, and \LaTeX\ users
will find no example of either a reworked {\tt lfonts.tex}, a style
file, or a pointer to vital ideas like those of Mittelbach and Sch\"opf.

The fonts for which there is no obvious CM equivalent (Symbol font and
Zapf Dingbats) are {\em not} mapped, but are instead each split into
two fonts, one with characters 1--127, the other with chararacters
128--256.

Now, all this is fine if you simply install the fonts on your PC and
use them all the time with a LaserJet printer or the like. But what if
you are after something more sophisticated? We bought the fonts for
one specific purpose, providing screen preview facilities for students
using \LaTeX\ on Sun workstations, who normally print to a LaserWriter
Plus. It would work nicely if we abandoned the builtin fonts, and sent
\TA\ bitmaps to the printer, but that would be silly, so we have to
set up a system whereby the previewer uses \TA, but the printer driver
uses builtins. This is not easy because the font metric files we use
normally, built by AFM to TFM conversion programs, assume that there
are 256 characters in a font, and may put things like ligatures in the
top half of the font. If we use \TA\ {\tt tfm} files, the font
encodings in the \PS\ we generate do not match up with the
metrics.  It {\em is} possible to build the right set of files all
round, but it means a permanent loss of the top 127 characters. Building virtual
fonts seems like hard work. In addition, the breaking-up of the Zapf Dingbats
into two fonts makes a dual system impossible -- again, virtual fonts are the
only answer. 
\topinsert
\vskip3.5in
\endinsert


This may sound like special pleading, but it is a sensible use of the
fonts, and it would be much appreciated if future releases provided
{\em all\/} the characters, without any mapping at all. Richard Kinch
has in fact promised that he {\em will\/} revise the fonts, provide new
ones, and offer a complete {\sc iso}-like encoding (it is under discussion
what scheme he should use).

What else is wrong with all the riches which Kinch has laid in front of
us, and which are illustrated in the figure, in normal use on a
PC? Should we  switch over to \TA\ fonts and cheap printers?
\item{$\bullet$} Kinch provides no solution to the problems of small
      caps. There is absolutely no excuse for this, as they are in
      very widespread use, and their absence causes irritation. They
      are of course easy to generate by manipulation of builtin
     \PS\ fonts.
\item{$\bullet$} You cannot have  arbitrary sizes like 4pt {\teeny like
a mouse's writing} or 16pt {\larger quite substantial, eh}; nor can
you perform non-linear scaling to squeeze or stretch the letters.
\item{$\bullet$} You cannot buy extra fonts like  Optima
\item{$\bullet$} You
cannot re-encode fonts to get new layouts (although Knuth's new virtual fonts
solve this problem) 
\item{$\bullet$} You cannot perform useful tricks like
rotated 
or reversed out text which
any publisher's designer might request at any time.
\item{$\bullet$} You cannot take advantage of the fact that you have a \PS\
      typesetter by also doing all your figures and illustrations with
      the many excellent graphical packages which produce encapsulated
      \PS.

So the \TA\ fonts are necessarily a poor man's alternative to the
standard method; but very few people use \PS\ font facilities to any
greater degree than is permitted by \TA, and will be delighted with the
results on the common HP LaserJet printers. My greatest worry is that
developments like this will marginalize the \TeX\ user even more; the
vast majority of casual users will go on quite happily with PageMaker,
Ventura, Word and other excellent programs, leaving \TeX ies as a small
band of fanatics typesetting maths. \TeX\ can survive in the real world,
but I seriously doubt that \MF\ can now ever succeed; lets
celebrate \TeX\ as the premier formatter, and leave fonts to page
description languages like real \PS.

If \TA\ fonts sound like what you have been waiting for, they are a good
product and good value for money; they can be obtained from

{\eight\baselineskip10pt\settabs2\columns
\+Kinch Computer Company&Text Formatting Company\cr
\+501 S Meadow Street&Suffield Works\cr
\+Ithaca&1 Suffield Road\cr
\+NY 14850&London N15 5JX\cr
\+\tt fax:0101 607 273 0484
&\tt\quad tel:081 802 4470\cr}}

\author{Sebastian Rahtz}
\endinput
\clearpage
\thispagestyle{empty}
\begin{tabular}{l}
\fontsample{times}{Times} 
\\
\fontsample{palatino}{Palatino} 
\\
\fontsample{ncs}{New Century Schoolbook} 
\\
\fontsample{avant}{AvantGarde} 
\\
\fontsample{helvetica}{Helvetica} 
\\
\fontsample{courier}{Courier} 
\\
\fontsample{bkman}{Bookman} 
\\
\end{tabular}
\author{Sebastian Rahtz}\date{Department of Electronics and Computer
Science, University, Southampton S09 5NH, email: {\tt spqr@uk.ac.soton.ecs}}
\begin{table}
\begin{tabular}{ll}
\em builtin \PS\ fonts & \em \TA\ fonts\\
\hline
\begin{tabular}{l}
\fontsample{times}{Times} 
\\
\fontsample{palatino}{Palatino} 
\\
\fontsample{ncs}{New Century Schoolbook} 
\\
\fontsample{avant}{AvantGarde} 
\\
\fontsample{helvetica}{Helvetica} 
\\
\fontsample{courier}{Courier} 
\\
\fontsample{bkman}{Bookman} 
\\
\end{tabular}
& %\psfig{figure=texaps.ps}
\\
\end{tabular}
\caption{\TA\ fonts, compared with builtin PostScript fonts}\label{demos}
\end{table}