\title{Adventures of a new \TeX nician}
I have been a \TeX\ user for just over a year now. My experiences are 
rather different from those of most \TeXline\ contributors, who seem 
to jet around the world telling Donald Knuth how he should have done 
it!

I was a prime target for the \TeX\ virus: a mathematician; someone who
typed his own papers, because seldom satisfied with what secretaries 
produce; and someone on an academic salary who could not contemplate 
buying expensive equipment. In the early 80s, I took the first 
tentative step into wordprocessing, putting aside an old portable 
typewriter in favour of a ZX Spectrum. It was liberating, but not 
quite perfect. For example, many of the misprints in my papers at that
time came about because I'd left space for a handwritten symbol, the 
space had come at the end of a line, and my eye had skipped over it.

So last summer, when I chaired the problem session at a conference and
was asked to edit a collection of problems, I had the perfect 
opportunity. A reasonably short document, with plenty of easy 
challenges like typefaces (bold for authors' names, italic for postal 
addresses, typewriter for email addresses, and a mixture of plain 
roman and mathematics for the text), but nothing too hard like fancy 
page layouts. One problem involved large matrices, so I had to read 
Chapter 18 of the \TB\ earlier than I expected.

By the end of the year I was facing a bigger challenge, producing 
camera-ready copy for a book. My decision to use plain \TeX\ rather 
than \LaTeX\ was made then, as the publisher sent me by email the 
formatting macros for the house style. Apart from one point I'll 
return to, I never regretted this decision. I had to change the 
formats a bit to get a page layout I liked, and also to add running 
heads, two-column index, etc. I also had to start devising sneaky 
tricks. For example, the publisher's format capitalised section 
headings; one of my section headings was a formula, which meant 
something quite different when capitalised. Did you know that 
`uppercase' affects math italic? For example
\begintt
\uppercase{The case when $x\to\infty$}
\endtt
gives \uppercase{The case when $x\to\infty$}.
Is that reasonable?

Other unexpected difficulties arose. I started producing my class 
problem sheets in \TeX\ but, though they were perfectly legible, the 
students seemed not to like them. One said that such immaculately 
typeset problems were more intimidating than my usual scruffily 
handwritten sheets.

First impressions of \TeX\ are inevitably a joy, both at the beauty of
the printed output, and at the way the designer thinks like a 
mathematician. (After all, he is one.) Being able to type a formula in
almost exactly the way I would {\it say} it to a colleague, and have 
it come out looking the way I would {\it write} it while saying it, is
a great boon. (There are some exceptions, e.g. having to type |\bar 
x| where I would say ``x bar''.)

The other great advantage of \TeX\ strikes you soon afterwards. Since 
its input is in plain \ASCII, you can send it by email, and your 
collaborators can have almost instantly a copy of your paper. Email is
great for academics, speeding up international collaborations 
immeasurably, but its defect used to be that you couldn't send 
mathematical formulae. Now \TeX\ has become the {\it de facto\/} 
language for this, being used in the middle of plain text by \TeX-%
literate mathematicians. These bits of \TeX\ are never even compiled
(except in the human brain).

Inevitably, after a while, you come up against limitations. I'll 
mention just one. You can draw simple diagrams in \LaTeX\ but not in 
plain \TeX. So, in my otherwise complete book manuscript, I had to 
send diagrams on separate sheets of paper, to be pasted in to spaces 
on the pages. (Some diagrams were beyond even \LaTeX, so I drew them 
by hand.) Wouldn't it be possible for someone to write plain \TeX\ 
macros (or copy the \LaTeX\ ones) for using the line and circle fonts?


In fact, the benefits of doing this would be a system more flexible 
than the \LaTeX\ `picture' environment. Often, I want to say 
something like `graphs having no subgraph of the form blah', where 
`blah' is a small graph (with, say, 3 or 4 vertices). Such an object
would take up no more space than a $2\times 2$ matrix (and Knuth gives
you two methods to include matrices in a line of text). You could 
design your own with \MF, but then you immediately make your file non-%
portable.

Geometers now habitually use diagrams to describe classes of 
geometries. These diagrams are composed of nodes (which may be solid 
or hollow) linked by edges (which may be single, multiple or directed),
both nodes and edges having optional labels above and below. If the 
diagram is a `string', this is a good exercise in writing macros.  But
if it branches, I don't know how to draw it in \TeX.

Being, as I said, an impoverished academic, my next computer after the
Spectrum was an Atari ST. Within a month of buying it, I was delighted
to see a public domain version of \TeX\ and \MF\ announced in a 
magazine. The PD library supplying it (the South-West Software Library)
also set up a helpline of experts to assist novices with problems. I had
so many problems getting the system running that, at the end of  it, I
was ready to offer my services as an expert! Now the working  system is
a great aid to productivity. I type the input on an old  Tandy portable (I
spend a lot of time commuting by train), debug and  preview it on the ST,
and laser-print it on the departmental network.  One small but important
ingredient is Tempus, an excellent  programmers' editor on the ST. With
its structure check, I can catch  all those missing closing braces!

From my experiences, I wonder how many potential users have been 
scared off by the steepness of the learning curve, not so much of 
\TeX\ itself, but of the mechanics of using it (installation, screen 
and printer drivers, etc.) I recently acquired a German shareware 
version. First time round, the ARC file was corrupted and wouldn't 
unARC correctly. Months later, a replacement came. This time, the 
`install' program re-wrote a file on my boot disk, putting the 
computer into a crash-and-reboot cycle which didn't do my nerves any 
good at all.

To conclude: I think that Knuth's `health warning' in \MFbook\ 
applies equally to \TeX: once you have started using it, you'll never 
look at a typeset page in quite the same way, and you'll never ever be
satisfied with your own efforts!
\author{Peter J. Cameron}
\endinput