%(letter to the editor of TUGboat)
%(Barbara, I hope you appreciate my remark about "the lion's share"...)

Several recent reviews in {\sl Computing Reviews\/} have confirmed my
long-standing impression that lots of readers are sensitive to matters of
quality printing. So far \TeX\ has gotten good press in that journal; for
example, Dick Andree began review \#39,590 by saying ``You will certainly
be fascinated to see the excellent mathematical typesetting displayed in
this book, set using Don Knuth's \TeX\ system. It is worth examining for
this alone.''

Of course, such reviews are somewhat embarrassing to me, because the
lion's share of the praise should obviously go to the authors for the
wonderful things they wrote; the format is only one of the many things
that were well done.

This particular book---{\sl Practical Optimization\/} by Gill, Murray,
and Wright---was one of the first to be produced on Stanford's Alphatype.
I can recall being pleasantly surprised to discover isolated pages (of
unknown authorship) in our darkroom while I was working on {\sl
Seminumerical Algorithms\/}; later I found out that Gill, Murray, and
Wright were responsible for this fine work.  They took the time to ``go
the extra mile'' by combining superb mathematical exposition with numerous
refinements. For example, they added a unique appendix about ``questions
and answers''; they included excellent illustrations and tables; they
chose their notation carefully; they made a good index and bibliography;
they found wonderful quotations for the beginnings of each chapter.  If
\TeX\ had any part in this, it was merely to encourage the authors to
strive for such quality because they were more personally involved.

Lynn Steen's ``telegraphic review'' of Arthur Keller's {\sl First Course
in Computer Programming Using Pascal\/} is another instance of format
appreciation; he says, ``The elegance of the text is matched by the
elegance of its appearance: it was prepared and typeset at Stanford in
\TeX.''  [{\sl American Math Monthly}, January 1983, page C8.]

Conversely, Bob Fenichel's review \#40,719 in {\sl Computing Reviews\/}
speaks of a book that ``is photographically reproduced from the output of
the author's daisy-wheel printer \dots\thinspace. The article headers are set
randomly in at least two different fonts, while the font of the text, for
reasons that the author does not share, is the inhuman OCR-B.''

Such judgments are obviously matters of taste, and we can't expect
universal agreement. For example, I went to the bookstore to see the book
just mentioned, and I didn't find its format disturbing; indeed, I think
Adrian Frutiger did an excellent job, under the circumstances, when he
designed OCR-B.

Fenichel's review goes on to make a point that I think is much more
important:  ``There are errors in spelling and grammar on nearly every
page. Granted that some authors have special interest and competence in
copy editing, proofreading, and typographical design, are {\it all\/}
authors now to be required to have such interest and competence?  If not,
then how can a publisher, presented with so-called camera-ready copy,
reassert his traditional control of these matters? Should he refuse to
accept such copy, accepting only traditional manuscript or
machine-readable text?''

We must realize that fine formatting is only one of many aspects of book
quality. I hope the day will come when there are copy editors and book
designers familiar with \TeX, to deal with authors who are typesetting
their books with \TeX. Meanwhile, I want to encourage authors who are using
\TeX\ today to seek professional help, instead of assuming that publishers do
nothing but print, bind, and distribute books. It would be terrible if
\TeX\ were to lead to {\it decreased\/} quality because these other
aspects were being neglected.

Of course, there is no royal road to quality; editors can make mistakes
too. The immediately following review [{\sl Computing Reviews\/} \#40,720]
mentions that another book ``is particularly ill-served by its editors,
who have failed to correct a large number of errors in English usage.''

The best way to solve all of these problems is with teamwork. I hope that
\TeX\ will ultimately help to provide better means of communication
between people with different kinds of book expertise. At the moment we
are seeing a sudden shift in who has the ultimate power to input and
change copy; with care, we should be able to find an appropriate way to
distribute that power.

\rightline{---Don Knuth}