\def\and{\item{$\bullet$}}
%----------------------------------------------
\centerline{\bf\TeX\ Users Group 1987 Meeting }
\centerline{\bf 1. View towards Mt.~Olympus (WA)} 
\medskip\noindent
Among the (many) virtues of \TeX\ are the facts that it is from the hand of a
single man, that it is not going to change and that it has a single, effective
User Group. The 1987 annual meeting of TUG took place this year from August
24th--27th at the University of Washington in Seattle (home of TUG past-%
president Pierre Mackay), and where, it seemed, better to garner the latest
from the horse's mouth?  Heart in mouth, your intrepid correspondent
accordingly entrusted himself and his copy of {\sl The \MF book} to TWA's
shoddy care and flew (the movie was {\it Hoosiers}, 
starring Gene Hackman, and the
food was a curious mess based on peanuts, which I came to learn was all TWA
{\it ever} offer vegetarians)  
off to the rainy State of Washington. To the surprise
of most people in sight, Seattle was bathed in sunshine and remained so
throughout our stay. Accommodation in a student `dorm' was in a bunk-equipped
bedroom reminiscent of a Townsend Thoresen cabin, and the nautical theme
continued as the c.\ 160 assembled
TUGgers gathered for a cruise on Lake Washington; beer and
wine flowed, though the culinary delights of Cajun Shishkebab and barbecued
salmon proved a little frustrating for the non-carnivore. Initial impression
of the participants was later proved correct---mostly from north American
universities, but  a healthy presence of commercial implementors and users,
and a sprinkling from countries all over the world. The British presence was,
to put it mildly, dominated by the Open University---my final headcount showed
4 OU, 1 Soton, 1 Oxford and 1 Wellcome Institute; lets hope the accounts
department of the OU don't ever find out.

Day 1 started as it meant to go on, early and hard. 7.15 breakfast showered
the bleary-eyed punter with eggs, bacon, pancakes, potatoes, fruit salad,
donuts, toast, coffee {\it ad nauseam}. Formal registration (and more donuts and
coffee, which appeared to come from a specially developed {\it cornucopia\/}
throughout the conference) produced the longed-for TUG 87 memorial coffee mug
and a program generously filled with \TeX\ vendors' adverts. The first proper
session opened with Christina Thiele 
(Carleton), who described her experiences
of small journal production with \TeX\ and electronic submission of MSS
without markup from authors; the field of linguistics brought nasty problems
of layout and fonts (there was subsequent discussion of where to find an IPA
font), but the interest of the session was the evidence of \TeX\ in heavy use
by non-experts and in a humanities setting, which was one of the themes of the
conference.

Moving onto more detailed matters
Silvio Levy (Princeton) followed with a description of his work on
\MF ing a Greek alphabet for a new Dictionary of Modern Greek. Greek has
special problems, such as the large number of 2, 3 and 4 letter ligatures used
in classical Greek  (e.g.~for prepositions), and the multiple accents on some
characters, and Levy uses a 256 character font to accommodate lots of
possibilities, including special combinations of sigma and all other letters,
to cope with the idiosyncracy of Greek's changing form of sigma if it comes
last in a word. 

After a more traditional digression on how to port \TeX\ to an
IBM 6150 (using PL8 and AIX `mapped files') 
by  Richard Simpson of IBM (with an
interesting aside on the use of Metafont to draw graphs by creating a special
purpose, very large, character representing the graph),  Walter Andrews and
 Pierre Mackay continued the alphabet theme with Turkish. 
The problem there was
accentuation so heavy that the |\accent| 
primitive cannot cope, and the only
solution was an accented font; discussion of the problem of sensible input
with Turkish characters on-screen provided the curiously useful idea of using
the printer driver family of a conventional word-processor to generate \TeX\
from wp formatting commands by `printing' to a file---subsequent experiments
with PC-Write show that this is a very workable system for font changes. 

Our
European problems were as nothing, however, compared to the horrors of
Japanese, which came next;  Nobuo Saito 
described the general problems, and the
existing transliteration and printing schemes, 
while  Yasuki Saito talked about
one of four actual \TeX\ implementations---J\TeX. The large number of
characters in a font was dealt with by fooling \TeX\ into thinking it was
dealing with a family of fonts, which together made up one single Japanese
font; the actual fonts are generated from existing raster files to |.gf| format.
Input uses the JIS code (2 bytes per character), and a small change to \TeX\
forces it to take characters in pairs. While Saito listed problems and areas
for improvement, it was clear that this was a viable system of practical use.

After more donuts, we moved on to device drivers; the chief stars here were
 Nelson Beebe, whose driver family had frequent mention, 
 Stephan v.\ Bechtolsheim (currently working, incidentally,  
on a book described as an `exegesis of \TeX'),
who has nearly completed his new DVI to \PS\ driver, and  Robert
 McGaffey, who led a panel discussion 
on standards for \dvi\ drivers (it struck
me during this that discussion of \dvi\ driver standards was by men, and
discussion of standards for macros, prompted by Christina Thiele, was by
women). The panel was formed into a committee to take serious action, with the
likely outcome being a set of standards which \dvi\ drivers should follow, at
different levels; those with serious thoughts on this were invited to send
them to McGaffey.

The evening of the first day was occupied by a barbecue hosted by
Addison-Wesley, with extraordinary entertainment from the Seattle Banjo Club,
whose average age seemed to be greater then three score and ten, yet who kept a
frightening vitality. As A-W's corporate image stopped them serving alcohol,
and since yet again the carnivore was in the ascendant, I can only relate
that the walk to and from the event across a research ecological preserve,
including a small secluded duck pool,
provided the best moments of my trip over the Atlantic.

Day 2 started with the commercial section, where exhibitors were given ten
minutes each to expound on their wares, which we were later able to inspect.
Thus I noted: \nl
\and {\bf Arbortext} with {\it Publisher}, their {\it wysiwyg}-cum-generic markup
front-end to \TeX, which offers a great many features (a good
interactive equation editor, and table builder) without {\it quite\/} working
yet as we would expect (for instance, cross-referencing is not supported, nor
is import of external \TeX). In any case, it costs real money on Sun
workstations! Arbortext also noted their new Sun font editor, an unsupported
Sun/Vax `dvicrt' for crude display on {\sc ASCII} terminals, and the fact that their
previewers now use Adobe screen fonts for those who want \PS\ fonts.
\and {\bf Digital Composition Corp.} 
were newer entrants to \TeX ery, who had made
their own PC port (complete with |WEB|, no less!), were working on text wrap
around figures and had a `dvi2ps' with merge of HPGL, and even referred to an
\SGML\ preprocessor.
\and {\bf FTL systems} produce Mac\TeX, which seems to be a more stable and
advanced product than \TeX{\sc tures}, concentrating on support for \PS\ fonts
(they have traced the CM maths fonts with Fontographer to generate \PS\
downloadable outline fonts), online \TeX\ and \LaTeX\ help and translation out
of MacWrite and Word; they also use `Page One', a system of simple generic
codes for authors which are converted to \TeX.
\and {\bf Docusoft Publishing Technologies} were offering \TeX write, 
a special PC
editor for \TeX, with a WordPerfect-like interface; they gave away evaluation
copies of a version which did not yet have many special \TeX\ features in it;
Emacs users can imagine the sort of thing.
\and {\bf K-Talk Communications} had a full-feature WordPerfect to \TeX\
converter; it was hard to get excited, but this sort of thing is very useful
and should become more popular.
\and {\bf Kellerman and Smith} wheeled on the said Smith, who was excessively
laid back with a Mac, but did mention that they could start shipping Lucida
for \TeX tures soon (with a  {\it complete\/} set of maths fonts).
\and {\bf Personal \TeX}, with the charismatic 
 Lance Carnes, talked about their
licensing of Bitstream fonts (a large library of raster fonts), the imminent
marketing of Montreal's multilingual \TeX\ (with European hyphenation
patterns, a
matter of great importance to many \TeX\ users in the humanities), new,
simpler, table macros and new device drivers. It has to be said that PC \TeX\
came over as having the edge on Micro\TeX, in terms of development and service
to the users.
\and {\bf St\"urtz AG} offered \TeX\ printing on a Monotype Lasercomp, and
promised to implement Knuth and Mackay's concept of a left-to-right Hebrew
\TeX.
\and {\bf \TeX nology} had Macro\TeX, 
a new, almost-finished, macro package at the
level of \LaTeX\ or \AmSTeX, with some interesting features such as
multi-page tables.
\smallskip
After  a set of Site Coordinators reports (pretty dull, to be honest), Nelson
Beebe talked about his driver family, which featured  a uniform interface for
all devices and much sharing of modular elements of the code. I think it was
also Beebe who mentioned in passing that \TeX\ and \MF\ are almost
exactly the same number of lines of code, a cause for thought! More
entertaining fare was to follow, however, 
from TV Guide's  David Ness, whose
talk was in some ways a high point of the conference; TV Guide is the weekly
listing of all the TV programs in the US, and its production is basically
database driven. Ness's eventual aim is to use \TeX\ for this highly complex
piece of typesetting, but the importance of his talk was its viewpoint from a
`real' organisation choosing \TeX\ on performance rather than availability or
price considerations. I cannot resist quoting TV Guide statistics---20 million
copies per issue, listing in 108 editions the 
output of 2000 stations; {\it Radio
 Times\/} eat your heart out. Silvio Levy returned to complete the day with an
outline of his new CWEB system for literate programming in
C --- not the combination of C and |nroff| that has
circulated in Britain, but C and \TeX. Maybe it will even catch on!

Before the wine and cheese provided by PC \TeX\ (some very nice local wines
and good cheese --- now you see why their product is better than Micro\TeX),
`birds of a feather' did the requisite flocking. I joined a flock of people
interested in special editors and markup, 
with \TeX\ {\it per se} being put in its
place as a formatting engine. One interesting fact I elucidated in this
happily-familiar discussion was that the American Association of Publishers
has defined its own \SGML\ tag set and is promoting it heavily.

In the evening, a small group made a pilgrimage to see the Seattle Mariners
play the Toronto Blue Jays in the Kingdome; the Mariners' fielding was
excellent, but their batting couldn't catch fire and the Blue Jays (as
was expected) won with no real strain. Did you know the Kingdome is the
largest concrete dome in the world? Thanks to  David Parmenter (DEC) for
explaining the rules.

Day 3 started with a problem session run by Barbara Beeton, who deftly fielded
some incredibly obscure hackings. 
The `humanities' theme was pursued by  Allen
 Dyer (a Maryland Lawyer), who described how his state Bar Association was
standardising on \TeX\ for the good of all. Last, but not least, your
correspondent presented a paper by Les Carr on the generation of \PS\ outline
fonts from \MF. The {\it cul de sac} reached by the actual project did not
stop a good number of people wanting to know more, and confirmed my suspicion
that for many \TeX\ people \PS\ is the way forward, possibly creating an
unfortunate divide between the pragmatic \PS ers and those who stand by CMR
and \MF.

Closing sessions gave a chance for more jokes from \TeX an TUG President Bart
Childs, presentation of \TeX\ `oscars' for deserving people, a framed Duane
Bibby drawing for Barbara Beeton, and the election of David Ness as Treasurer.
TUG seemed to be flourishing and happy, so we went away.

What were my overall impressions from the meeting?
\smallskip
{\narrower
\item{a.}An increasing number of people are 
simply {\it users} of \TeX, and are not
interested in how \dvi\ drivers work; in the USA at least I got the feeling
that the main thrust of usage was on PCs and Macs, using the prepacked systems
from conscientious vendors. Many of them are finding the underlying algorithms
of \TeX\ can be used for an extraordinary variety of tasks.
\item{b.}People who {\it are} 
interested in \dvi\ drivers are on the verge of getting
their act together to ensure that these programs do their job. One can only
applaud!
\item{c.}The commercial world is less interested in \TeX\ as a visible
product than as a reliable engine. It was interesting to hear FTL say that
they did their best to {\it conceal\/} \TeX\ in Mac\TeX, which meant almost
discouraging the use of CMR (they are an extra in Mac\TeX).
\item{d.}Americans drink too much fizzy stuff and coffee,
and eat too many donuts.
\smallskip
}

\medskip
Readers who want to follow up any of the matters discussed above should first
of all look at the latest TUGboat, which covers a number of them, but are
welcome to get copies from me of brochures, addresses etc.\ I picked up --- write
to:
{\obeylines\parindent2em
Sebastian Rahtz,
Department of Electronics 
\qquad and Computer Science,
University, 
Southampton SO9 5NH
{\it email} {\tt spqr@uk.ac.soton.cm}}