Re: The first xy

[TBaehr@xxxxxxxx]

The two geniuses behind XyWrite were Dave Erickson and John Hild, colleagues at Atex who recoded the Atex typesetting system's front end and turned it into a word processor. John was the marketing brains, and Dave was the chief architect. I belive that, officially, John was president and Dave was chairman of the board of directors.

As with many software products, the "mature" realization of XyWrite was version 3, but the
gold standard for many was III+, which added spell-checking and a thesaurus.

The relationship with Atex remained, and XyWrite was marketed to many newspapers and magazines as
the means of getting authors' ideas into print. III+ was (and is) lean; it would fit on a 5.25"
floppy and would run on even the most modest laptop. Because data files were almost straight ASCII,
they were very small; transmitting stories over phone lines, even at 300 bits per second (called
"baud" in those days) was doable. Publications using XyWrite included the New York Times,
the Boston Globe, the New Yorker, and hundreds of others (including, ironically, many computer
magazines as they fell all over themselves to praise and hype WordPerfect and Word). XyWrite was
considered an industry standard and for a while was the best-selling word processor around.

The fall from the top started slowly and accelerated rapidly.

XyWrite IV was supposed to add limited but editable WYSIWYG capability to the DOS-based word
processor. Only a few type sizes would be available for on-screen display.

Part-way through development, Xyquest entered a partnership with IBM, which was seeking to replace
its clunky and outdated Displaywrite product. IBM had many good ideas, all of which slowed
development to a crawl: menus and dialog boxes were among them. IBM also imposed a lot of testing
and review overhead, further slowing development.

The new, IBM Common User Access-compliant user interface was built by modifying the core executable
(EDITOR.EXE) to accommodate renderings of dialog boxes and menus. The WYSIWYG concept expanded to
include fullly scalable Speedo typefaces.

The main programming for the interface, and much of the functionality of the new XyWrite was carried
out, however, in XPL, XyWrite's proprietary programming language that looked like a cross between
Basic and job control language. The interface staff never numbered more than 4, and the total
company headcount (including sales, tech support, engineering, and administration) never got much
over 60. At this time, WordPerfect was developing its Windows product (Win 3.0 at the time), with a
staff of 600.

Renamed "Signature," the Xyquest product was on the verge of (late) shipping when IBM
pulled the plug on all of its software partnerships. We had to put stickers over the IBM logo on our
boxes before we could ship our new word processor. And we didn't ship that many. Signature was slow
and buggy. Our customer base had eroded somewhat, and many XyWrite aficionados hated the new
interface (although the old command-line capability was still there).

The Xyquest strategy was to use the revenues from sales to finance further promotion. It never
happened. Sales declined, and money became extremely tight. There were layoffs. Xyquest looked for
another investor and eventually linked up with Kenneth Frank, a Baltimore lawyer and software buff
who wanted a text editor to form the basis of a vertical-market product for creating legal
documents. Kenny bought the company. Initially, Xyquest was a division of The Technology Group, but
the Xyquest name eventually disappeared. So did John Hild. Dave Erickson stayed on as chief
engineer, overseeing two subsequent versions: XyWrite 4 for DOS, a cleaned-up and much faster verson
of Signature; and XyWrite for Windows, a noble effort whose main feature (for some of us) was the
General Protection Fault. By the time of the Windows product release, the development staff was down
to about 10 people; I was the only interface programmer.

Our office space shrank; there was a rumor that Kenny had failed to pay our rent and we were on the
verge of eviction. Some paydays were delayed, with conflicting stories about whether the delays were
caused by computer or accounting glitches or lack of funds.

Development took on, sometimes, a surreal aspect. Kenny was a true enthusiast, a very hands-on
owner. Development was occasionally delayed while we added a feature or made some improvement he had
thought of.

We didn't all get along under the new regime (Xyquest had really felt like family to many of us).
The staff that remained started drifting away. I was one of the last to go, after three years
without a raise. I might have stayed, but a friend had lured me away to another small company that
showed more promise.

Now Dave is gone, and XyWrite is extremely unlikely to be developed further as a word processor.
Most of the intervening history has been chronicled in this forum.

Tim Baehr