\def\CTAN{\textsc{ctan}}
\newcommand{\NW}{Norm}

\title{Book Review --- `Making \TeX\ Work'}
\author[Malcolm Clark]{Malcolm Clark\\
\texttt{m.clark@warwick.ac.uk}}
\begin{Article}

\begin{verbatim}
@book{Walsh94,
  title="Making TeX Work", 
  author="Norman Walsh", 
  publisher="O'Reilly & 
           Associates, Inc.",
  address="Sebastopol, CA",
  ISBN="1-56592-051-1", 
  pages=483}
\end{verbatim}

This is truly a niche market book. Until I read it, I didn't realise
there was a need for it. Basically it tries to describe the many tools
and tool fragments which exist for use with \TeX\ (and \MF). In
general terms we are talking of tools available through the Internet,
and most probably those available at your nearest convenient \CTAN\
host. It is therefore encyclopedic rather than especially
profound. You are unlikely to gain any new insights into the working
of \TeX\ and its friends, but you will probably meet lots of new
friends, whose existence you never even imagined (I excuse a couple of
people from this: almost inevitably our esteemed editor will be aware
of everything mentioned here, and probably has a list of other tools
which should have been included). Given the nature of the Internet
world, and \TeX, this can be no more than a snapshot which is probably
already out of date: for example, the IP address for
\texttt{ftp.tex.ac.uk} has changed.\footnote{For the curious, it is now
128.232.1.87 --- Editor}

The book was first published in April 1994, and by July had been
reprinted with `minor corrections'. The reprinted edition was one of
the goodies at last year's TUG conference in Santa Barbara. Thank you
O'Reilly \& Associates, Inc.

\NW{} describes \CTAN\ as `a fully-mirrored anonymous FTP hierarchy on
three continents' -- but the three continents are North America,
Europe and, er, Europe.  He makes the common mistake of equivalencing
UK with England. World Wide Web makes an appearance indirectly, rather
than as a \CTAN\ supported service (so much easier to use); to be fair
he does mention the WWW interface available through O'Reilly's server,
\begin{verbatim}
http://jasper.ora.com/CTAN/ctan.html
\end{verbatim}
If you try this \textsc{url}\footnote{If you are
confused by all this stuff, help may be at hand. We hope
that a future edition of \BV\ will discuss the
\TeX\ resources available through World Wide Web, and give
some clear pointers, as well as filling in some of the
background.}, you will find it would have been better
to go straight to
\begin{verbatim}
http://jasper.ora.com/ctan.html
\end{verbatim}
You may be concluding by now that we are strongly oriented towards electronic
networks. This is true, although \NW{} does list some ways of obtaining
distributions through more traditional means. But imagine you wanted to get
hold of em\TeX\ or Oz\TeX: \NW{} merely refers you to \CTAN\ for em\TeX\
and omits any contact at all for Oz\TeX. This despite the trailer at the
end of issues of TeXhax
with details of how to obtain this software on disc.
Similarly he seems to have missed the (admittedly fairly recent)
\textsc{cd-rom} offerings from both the Dutch-speaking users group and from
PrimeTime Software which contain gobs of material useful to those with and
without electronic connection -- let's not forget that those of us on
SuperJanet are still a minority. Many \TeX\ users still access a network (if
they access one at all) over a telephone line, and the thought of downloading
all these files, even if you can locate them, is daunting.

We are also fairly Unix oriented. This is hardly surprising, given
that this is an O'Reilly book, and they have made their name by
producing books, which if not 100\% Unix in flavour, are at least very
oriented towards Unix and GNU/Free Software Foundation. Thus
there is a sort of assumption that you can probably write
\emph{Perl}, and fairly readily modify the \emph{Perl} scripts
which \NW{} includes. He uses \emph{Perl} because it is a scripting
language available on all the platforms he discusses -- Unix, MSDOS
(and Windows), OS/2 and the Macintosh. Sadly he doesn't tell me how to
get hold of \emph{Perl}. It is this Unix bias which leads him to
assert that `Unix is probably the most common \TeX\ platform' and to
go on to note that the emphasis of the book is Unix workstations
running X11 producing output for PostScript and HP LaserJet printers.

\NW{} suggests that his description of \TeX\ is really for the systems
administrator who wants to know a bit more about this behemoth which
he has been persuaded to let occupy her or his system. The computer
naive, or just those wanting to get on with some documents are given
scant regard. Although he does include some discussion of MSDOS and
of OS/2, this is really because many tools have been ported to these
platforms from Unix. The inclusion of the Macintosh is intriguing, but
other than ghettoising it by giving it its own exclusive chapter out
of the sixteen, it makes little appearance, and the repeated themes of
command lines and restricted file names does demonstrate that it was
something of an afterthought. The really useful fact of Mac filenames
-- that they should not contain a blank space if you hope to use
\LaTeX, is omitted.

In general, there also seems to be a bias towards \TeX, rather than
the more accessible extensions like \LaTeX, although again \NW{} does
mention \LaTeX, Lollipop (a little surprisingly) and Texinfo
(unsurprisingly, although again demonstrating the Unix bias).

I have some misgivings when I start with \NW{}'s description `What is
\TeX'. The first aesthetic principle which he ascribes to \TeX\ is:
`The right margin is justified' (page~3). Besides being a rather
contested aesthetic principle, \TeX\ of course allows you to set
ragged right, or ragged left (or even both). He also claims `ladders
are avoided'. None of my books on typography mention ladders: I assume
them to be equivalent to rivers, and while this is certainly a claim
by Knuth \& Plass, \emph{and} I have yet to see rivers created by
\TeX, I'm still not convinced of its universal truth.

I'm made uncomfortable by the first table (page~9), where among the
special characters for plain \TeX, \NW{} suggests using \verb+<+ for
!` (the usual recommendation is to access it by \verb+!`+), and
\verb+>+ for ?` (usually \verb+?`+). This smacks of the shortcut `it
works!' approach which always makes me nervous. He also notes the use
of \verb+|+ to obtain an em-dash! I begin to despair. This way lies
madness. And why would you want a `lone' tilde character or an
underscore in your document? Yes I know they are widely used in Unix
for paths in directory and file names, but that's hardly mainstream
\TeX\ usage.  I would be wary of defining a command \verb+\big+ in
plain \TeX\ (as on page 100), since it could easily confuse maths
typesetting (try \verb+\big\lbrace+ after you have redefined
\verb+\big+). And it must be incorrect to maintain (page 21) that if a
document contains `no forward references then it can be formatted in
one pass'. If it contains no references at all, yes, this is true, but
if there are any references (i.e.~\verb+\label+ commands) the
\texttt{aux} file has to be written out fully and read again on the
next run to resolve the reference(s). In reality this is hardly a
problem. I've never yet been able to write a document in one `pass'.
There is always some other reason for having to run \LaTeX\ again.
Where it does matter is with a document you either receive, or one you
pick up from a server. You have to know to run it \emph{at least}
twice, and perhaps more times. Basically, keep running it until
\LaTeX\ stops carping.

To me a section entitled, `\TeX\ for beginners' contains a bizarre,
almost oxymoronic, concept. I no longer believe that you should be
taught \TeX, until you have mastered \LaTeX. If you need to go further
than \LaTeX, then it is worthwhile finding out how \TeX\ works. Most
of us don't really need to know first about macros and registers, far
less `token lists' (unless of course we already have a few computer
languages to our credit). There also seems to be some confusion in
describing what \TeX\ is. \NW{} writes ``\TeX\ allows you to separate
markup and layout. Logical divisions in the text (chapters, sections,
itemized lists, etc.) are identified by control sequences''. Yes this
is true, but it is closer to a description of \LaTeX.  On page~14 we
have a supposed comparison of \TeX\ and \emph{troff} commands,
except the \TeX\ is \LaTeX.  In passing, why \emph{troff}? Surely
no-one would select \emph{troff} over \LaTeX\ given a free choice.
Ten or fifteen years ago, and tied to Unix, maybe there was a contest,
but today?  Similarly the \TeX\ macros on page 43 are for \LaTeX\ (to
be fair, he does say that they are from the \CTAN\ directory
\texttt{macros/latex/contrib/misc} so I should have been able to work
that one out by myself.  If \NW{}'s contention was that to talk of
\TeX\ was to imply \LaTeX, I wouldn't mind, but there is just too much
jumble here, and I'm getting confused. Am I just quibbling? I
don't think so. The seeds of confusion are being sown.

Having said all this, I think the book is very useful, and it contains
answers to many questions. Just ignore the bits about \TeX\ and
concentrate on the tools. It almost answered my query about PFM files
and how to convert them to \texttt{tfm}s. Simply by recording where
many things are in the \CTAN\ archives is a great boon. I can happily
spend all day searching through \CTAN, but my chances of coming across
something useful are slim. Now I have a better idea of what is there
and my searching is more directed. In some areas \NW\ does go into
reasonable detail -- for example the installation of em\TeX, or
describing how to use new PostScript fonts in \TeX. He also usefully
spends a little time and effort describing how to get pictures into
\TeX, describing the problems and pitfalls, as well as actually doing
it. It is a useful addition to my library of \TeX\ books, if only
because it more easily enables me to answer those `is there a public
domain driver for \textit{xxx}?' questions.  

I'm not sure I enjoy the
format very much. Paragraphs without indentation and a paragraph
separation of about one line height make me shudder a bit, but many
manuals have this sort of form, and maybe that's what \NW\ wanted. I
don't think it shows \TeX\ off to advantage, although he did make a
real effort by eschewing Computer Modern and using Garamond instead.

\end{Article}