\def\PS{\textsc{PostScript}}
\title{Malcolm\rq{}s Gleanings}
\author[Malcolm Clark]{Malcolm Clark\\\texttt{m.clark@warwick.ac.uk}}
\begin{Article}
\newcommand{\TL}{{\em \TeX\kern-0.1emline}}
\subsection{Rest in peace}
For a number of years, the \lq{}maverick\rq{} (so described by Adrian \lq{}no relation\rq{}
Clark) occasional publication \TL\  made its infrequent appearance. The last
issue, number 14, came out in February 1992. As the erstwhile editor I now
reluctantly admit that it has ceased publication. I had hoped that it might
survive to a ripe old age (like issue 20), but a variety
of reasons compel me to abandon it. \TL\ was always a free
publication, so at least I don\rq{}t feel that there is a vast
subscription paying audience out there which will feel aggrieved
at not receiving the next edition. After all, part of the deal
was supposed to be that if you wrote for it you were then
on the subscription list. To the list of contributors, sorry,
your subscriptions have lapsed. To the rest of you, well, you
know you get what you pay for.

There are now plenty of other ways to disseminate information about
\TeX\ and family. Estimable publications like \BV, \emph{Lettres
  GUTenberg} (no, not \emph{Lettres fran\c cais}), TTN and the Dutch
  group\rq{}s MAPS more than adequately fill any holes that might have
  been left by \TL, and have the great virtue of being related to
  various groups. I hope that this encourages people to belong to
  those groups.

Of course there are other ways to spread the word. The electronic
distribution of information is an obvious alternative. Why, for
example, don\rq{}t we have all these journals and newsletters (or annals)
up on WWW (World Wide Web)? TTN is already available in electronic
form, but I think I can see the day when it is available in (say) \PS\
form so that I can just call it up on the screen in front of me.  The
problem here lies in one of the recurrent problems of material
available on the Internet -- who would bother subscribing to TUG or
the UK group if one of the key benefits of membership, the
publications, was already available electronically? I still prefer
bits of paper (I can easily take them home, read them in the bath,
make notes on them -- and the batteries rarely run flat), but paper
may be a luxury we can\rq{}t really afford. I\rq{}m moderately hopeful that
gadgets like the Sony DiscMan, together with emerging technologies
like Adobe\rq{}s Acrobat may make portable electronic books a realistic
alternative.


\subsection{Carousel?} 
A couple of years ago, Adobe Systems started
to introduce their \lq{}Multiple Masters\rq{} technology. At the time it was
not wholly clear what their agenda really was. Multiple Masters allow
you to retain the \lq{}look and feel\rq{} of a printed (or screen) document
when you do not have access to the \lq{}correct\rq{} fonts.  Imagine you had
been sent \BV\ electronically, but without the embedded Baskerville
fonts which are part of its corporate identity. Exactly what would
happen when you tried to view it or print it depends on lots of
things: at the worst you might not be able to view it at all;
alternatively you might have some substitution which retained
character widths, or character positions. The effect is that you would
end up with something which was frankly unpleasant. The content would
still be the \BV\ you have come to love, but its appearance would be
well below the standard which we have come to expect.

Adobe\rq{}s solution was to develop a scheme where two font
families were introduced which would attempt to emulate
the characteristics of the \lq{}missing\rq{} font(s). All this means is
that they would have the same character widths and heights
(slants, weights or whatever) so that the colour of the
\lq{}page\rq{} would be similar. Of course, there was no intention that
this would be an exact reproduction -- just that it would preserve
characteristics like line breaks and the other physical attributes
of the page. The two base fonts introduced were Minion and Myriad
(now renamed Adobe Sans and Adobe Serif, I think). These were
placed in the public domain.

Originally Adobe\rq{}s marketing people (probably to put us off the true
scent) were suggesting that this was the start of a wide range of
Multiple Master fonts. In passing, this is where Yannis Haralambous in
the last edition of \BV\ was misled. He states that -- \lq{}These (MM) are
extremely complex and memory consuming, but still much poorer than
\MF\ created fonts\rq{}: this may be true, but it is totally irrelevant
and misses the point entirely; this is not what was going on at
all. It was all a prelude to Acrobat, or, as it was originally
code-named, Carousel. Multiple Masters were simply an enabling
technology which had to be in place before Acrobat could be
released. I had originally supposed that part of the reason behind MM
was font licensing. The copyright of fonts depends on the country in
which you sit. See, for example, Chuck Bigelow\rq{}s article in \TUB\
(\lq{}Notes on typeface protection\rq{}, vol.7(3), 1986). It is by no means
certain that you have the right to transmit the details of any
arbitrary font electronically: by doing so you may be in breach of
copyright (there is no use pleading that Bodoni has been dead for more
than 50 years, or Baskerville for that matter -- they never held the
copyright anyway). Another reason not to do this is that encoding or
embedding the fonts takes up an enormous amount of space. If you are
transmitting a book sized document, the extra for fonts is no big
issue: if you are transmitting something relatively short, it could be
the biggest item. There are therefore legal and practical reasons for
this route.

Once we have an Acrobat document, it can be transmitted electronically
without difficulty, and can be read by anyone with an Acrobat
reader. Such readers seem destined to be made available with every
machine, virtually as part of the operating system. It no longer
matters what software was used to prepare the original -- you will
still be able to see what the author intended, without having that
software yourself. There are other little advantages: the one I like
most is the post-it notes that you can attach to the document. For
example, this could mean that as an author I send my Acrobat document
(which of course started life as \TeX\ or \LaTeX) to my publisher,
whose editors would annotate via the post-it notes. I then revise the
returned document. It all starts to sound very plausible. And yes, it
does work.

Another pleasing feature is the ability to search the Acrobat
documents. This works reasonably well, although, as you might expect,
ligatures (and I dare say, diacriticals) don\rq{}t yet work properly. Of
course, years and years ago you could search a dvi file for text
strings with ArborText\rq{}s previewer, with just the same sorts of
restrictions. I think it is always good when the innovations
introduced through \TeX\ and its associated products becomes
incorporated into mainstream technologies like this. It shows how
\TeX\ leads the way.

\subsection{Knuth, the video}

Donald Knuth is now available on video: University Video
Communications, based at Stanford, now have a tape of Knuth\rq{}s
\lq{}Computer musings\rq{} in which he discusses the One-Way Associative
Law. Sounds just the thing for a stocking filler next Christmas.


\subsection{Sheep stealing in Barnet}
From time to time a chestnut re-appears. A half remembered quotation
from Frederic Goudy is trotted out: \lq{}\lq{}anyone who would letterspace
lower case would steal sheep\rq{}\rq{}. There are many who concur with the
spirit of this statement, but unfortunately it isn\rq{}t what Fred
said. Matters have been confused recently by the publication of a book
by Erik Spiekermann \& E M Ginger entitled \lq{}Stop Stealing Sheep\rq{},
which does contain the erroneous quotation.  Fortunately they also
print the correct version. What Goudy actually said was \lq{}\lq{}anyone who
would letterspace black letter would steal sheep\rq{}\rq{}. Where does that
leave us? In \BV\ 3(2) I noted that Eric Gill suggests letterspacing
lower case as a substitute for italics. While I know of no rumours
that Gill actually {\em stole\/} sheep, that might have been the least
of their worries.


\subsection{Public domain does not mean user unfriendly}

In another surprising outburst (one by-product of the remarkable
mud-wrestling competition held with Berthold Horn, judged by most
observers to result in a draw), Yannis Haralambous suggests in a
footnote that \lq{}public domain software is never as user-friendly as
commercial\rq{}: I beg to differ. I have the doubtful pleasure of a
Windows machine on my desk. One of its few really excellent pieces of
software is Pegasus, a Windows-based electronic mailer. It is simple,
straightforward, and almost a pleasure to use. A true pleasure to use
is Eudora, another mailer, on the Macintosh. Both of these are public
domain, and I challenge anyone to carp significantly about their user
interface.

In fact, I\rq{}m rather surprised by Yannis\rq{} assertion that \lq{}important
innovations (in \TeX\ etc) have always appeared first in public domain
software\rq{}: I have only to think back to the release of \TeX\ on the
pc: this was not in the public domain but was the result of
significant and independent work by Lance Carnes and David Fuchs --
maybe it was not an important innovation, but to me, freeing \TeX\
from the restrictions of academic mainframe computers and releasing it
to the personal computer level seems a quantum leap forward. Even
Yannis\rq{} beloved virtual fonts were released first by ArborText, many
years before they became commonplace. There are countless other
examples around -- think of Blue Sky\rq{}s innovative Lightning Textures,
or Michael Vulis\rq{} V\TeX. This continued denigration of the importance
which commercial vendors have had in the adoption and spread of \TeX\
smacks of a re-writing of easily verifiable history. They were there:
they supported us all for many years until public domain versions of
the program finally became robust and reliable. I believe they still
have a place in the development of \TeX\ and \TeX\ tools. I just hope
they don\rq{}t get so frustrated by the lack of understanding and respect
they receive that they abandon the game entirely.


\subsection{Hints \& Tips}

At the first meeting that this group held at RHBNC one of the \lq{}events\rq{}
was a \LaTeX\ tables workshop. One of the problems that always comes
up with tables is the incredible contortions that you have to go
through to make the floating tables appear just where it is that you
want.  It would be naive to assume that this is going to be one of the
problems that will disappear with \LaTeX3. In the first place it is
extremely difficult to come up with a sensible and coherent set of
rules which describe the variety of conditions and exceptions to
figure and table placement; in the second place, the \LaTeX\ defaults
are perhaps not the best chosen. At least we can do something about
that, although there are rather too many parameters to be able to come
up with a definitive optimal set of values. However, at the RHBNC
meeting, Geeti Granger of John Wiley \& Sons provided a set of values
which she suggests had been used with more success at John Wiley than
the defaults. She suggested the values in the following table:

\begin{tabular}{lrr}
parameter & default & suggested \\
\hline
topnumber & 2 & 2 \\
bottomnumber & 1 & 2 \\
totalnumber & 3 & 4 or 2 \\
dbltopnumber & 2 & 2 \\
\verb|\topfraction| & 0.7 & 0.9 \\
\verb|\bottomfraction| & 0.3 & 0.5 \\
\verb|\textfraction| & 0.2 & 0.1 \\
\verb|\floatpagefraction| & 0.5 & 0.8 \\
\verb|\dbltopfraction| & 0.7 & 0.9 \\
\verb|\dblfloatpagefraction| & 0.5 & 0.8 
\end{tabular}
\end{Article}
\endinput