department.
“They don’t,” Odendhal said. He and others around Newman had hoped that
the Kays would sound out the need for products scientifically before
turning the engineers loose. Odendhal may or may not have been right. If
marketing men at the large computer companies were so smart, how come
Kaypro and Osborne had been the first firms to make portables usable in
business? And what about all the IBM clones? Couldn’t the marketing
men’s me-too-ism have stifled innovation? Then again, another company
dominated by engineers—Texas Instruments—had flopped in the
home-computer market because it didn’t pay _enough_ heed to marketing.
The Kaypro Corporation, at least, still boasted a dealer network of more
than a thousand stores in mid-1984. But some stores in my area had
dropped or downplayed the Kaypro line in the earlier part of the year.
One reason was that just about all the Kaypro machines were still 8-bit
and didn’t run the new IBM-style software. Also, ComputerLand and other
big-name chains, a growing part of the business, complained that margins
on Kaypros were ten percentage points under competitors and that the
machines' software hurt sales of off-the-shelf programs. So ComputerLand
shunned Kaypro.
It was a clear case of the interests of the consumers being at odds with
those of the computer dealers. For the Kaypro II at the time was a
remarkable bargain; Andy Kay had lowered his _retail_ price in March to
$1,295.
What a contrast to the overpriced PC_jr_ from IBM. Just rumors of the
machine—nicknamed the Peanut—had kept the entire industry on edge. It
bore the three magic initials. And as the pros said, IBM was compatible
with IBM. Only it didn’t work out that way. The PC_jr_, at least the
first version, couldn’t run some electronic spreadsheet software and
other important programs written for the IBM. The introduction in
November 1983 had been anti-climactic in other ways. _Jr's_ little keys
were like chiclets, horrors for the touch-typist, and by the time you
bought your software and added a second disk drive, you’d be paying
hundreds of dollars more. Even after IBM upgraded the PC_jr's_ keyboard
in summer 1984 and lowered the price to $999 with one drive, the Kaypro
II was still superior for word processing. So were other machines. With
two disk drives and WordStar, the Sanyo MBC550—a somewhat IBM-compatible
desktop from Japan—listed for just $1,400. Not that the Sanyo lacked
limitations of its own. Its keyboard, too, was no match for the
Kaypro’s.
In 1984 the Kays introduced the New Kaypro II, which offered only one
disk drive but sold for just $995 and allowed for another to be added
for a few hundred dollars more. Anyone using the New Kaypro II for
business would be foolhardy to depend on just one drive. Not
surprisingly, the $995 machine inspired jibes that Kaypro was guilty of
using the same gimmickry as IBM and Apple had been with their
single-disk computers.
Moreover, all the Kaypro models, despite their low prices, made _some_
customers pay for their bargains with headaches.
Take Chris Jensen, a writer in Cleveland. He futilely tried for months
before he could send his stories to his newspaper from home via the
phone lines. His dealer had sold the Kaypro IV, promising that Jensen
could. “It’s really scummy to put this out,” the _Plain Dealer_ reporter
said of his machine, “and know that the communications software has
serious bugs in it. It’s the kind of thing that the government would
kick the shit out of the automobile companies for doing.” A Kaypro
employee admitted that the communications software for the Kaypro IV had
come out with bugs, but not those of the kind that were plaguing Chris
Jensen. He confirmed another of Jensen’s complaints: that the Kaypro IV
model with the modem couldn’t use the most popular communications
software at the time. Coincidentally or not—shortly after Jensen griped
to me, the _New York Times_, and computer writer Peter McWilliams—Kaypro
released SuprTerm version 5.3. And that did the trick for him. The
machine delighted Jensen in other ways, but he still wondered about all
the ordinary customers out there who were unwilling or unable to wage a
media campaign to get their software up to snuff.
Jensen’s story was atypical—many Kaypro owners still spoke as if their
computers were micro manna.
Still, the new Robie, introduced in early 1984, might be a sign of
troubles to come. It was unique, a desktop selling for $2,300 with two
2.5-megabyte _floppy_ disks. But disks didn’t run as quickly as a hard
disk, and as of late 1984, the Robie just hadn’t caught on. The Robie
looked so ugly—ungainly, plastic cheap, fit for a discount store—that
even some value-minded buyers might have thought twice about bringing it
into their offices. It was a computer that only a designer could love.
And it was just an 8-bit machine.
In the 8- versus 16-bit battle, people were accusing Kay of living in
the past. It was Osborne all over again, some might have said—just like
Kay’s old rival babbling on and on about the virtues of a machine with a
tiny screen. This may have been unfair. For simple word processing, for
instance, an 8-bit machine could be a terrific bargain. But many buyers
still gravitated to the 16-bitters, worrying that future software might
not appear in the Kaypro CP/M format.
Kay’s attitude didn’t help. He was showing less of the adventurousness
that had led to his hard-disk portable.
And he now felt that the Kaypro II was just like the old Volkswagen bug
and would always be around—an ominous parallel considering VW’s recent
fiascos in the auto market.[13]
Footnote 13:
Kaypro later dropped the Kaypro II and replaced it with the $1,595
Kaypro 2X. It’s still a II, though, except for some improvements such
as more disk space. The 2X’s disks store almost 400K each.
Some software programs endured because enough people learned and liked
them; and the Apple II in one form or another would last; but the Kaypro
II? Even with hundreds of thousands of users? What about the ill-fated
Osborne? No, computer buyers normally spurned antiques. Within a year or
two Andy Kay might change his mind and ponder whether to update the
Kaypro II drastically or kill it.
In November 1984, well over six months later than Kaypro had promised
investors, a 16-bit IBM-compatible machine appeared.
“He’d rather not call it an IBM clone,” said an industry observer. “He’d
rather call it just a 16-bit computer. It’s a matter of pride.”
The 16-bit model, a portable with a 10-megabyte hard disk and 256K RAM,
looked somewhat like the Kaypro II and IV. It had an =expansion slot=
into which you could plug a device to hype up the performance of the
machine or expand its versatility; and it offered free WordStar along
with other software. At $3,295 it sold for $1,700 less than a hard disk
portable from Compaq. But it faced stiff competition from machines such
as the hard disk desktop from Leading Edge, which was discounted to as
low as $2,000. The 16 just wouldn’t wallop rivals the way the original
Kaypro II had.
And it shared some of the drawbacks of the Compaq and many other IBM
clones—especially the keyboard. The left shift key was in an awkward
position like the IBM PC’s, with a weird, nonalphabetical character
separating it and the “Z”; imagine the reaction of touch typists who had
enjoyed the Kaypro II’s close resemblance to the traditional Selectric
keyboard. Many makers of 16-bit machines had aped the horrid PC
keyboard, causing one reviewer to liken them to idiot savante piano
players who imitated the mistakes of true geniuses.
The 16, moreover, showed Kaypro’s internal weaknesses. Dan Berger of the
San Diego recalled how Kaypro had futilely struggled for more than eight
months to design the computer in-house. The project had flopped. Kaypro
farmed the design out to a San Diego firm, which did the job in a
fraction of the time. Earlier in 1984 Andy Kay had told Berger that
Kaypro was spending only 1 percent of revenues on research and
development, compared to the industry average of about 10 percent. How
frustrated he must have felt paying an outside firm. Osborne had had to
do that, but technically Kay had placed himself and his company in a
different league.
Kaypro was heavily depending on the 16. The company hoped it would
provide 40 percent of income for the year ending August 31, 1985.
Early signs were good. A small dealer in Virginia, for instance, The
Disk Connection, sold six Kaypro 16s in one week—half to existing Kaypro
owners and half to new customers. Some people had demanded an IBM clone,
bungled keyboard and all; now Kaypro had given it to them with a
vengeance. Appearing late, however, the 16 had lost much of its price
advantage. In another six months IBM might well sell its XT hard disk
model for as little as the new Kaypro portable.[14]
Footnote 14:
“XT” stands for “extended.”
The 16, if successful, would triumph at a helpful time. In the 1983
budget year, the company had earned $12.9 million on $75 million in
sales; the next budget year it had _lost_ $267,683 on sales of about
$120 million—blaming everything from advertising costs to price-cutting
and inventory problems. No longer was Kaypro a success story. It
remained strong in some stores specializing in computers; but the share
of the whole domestic market for personal micros would be 2.5 percent in
1984—a drop from 1983.[15]
Footnote 15:
Kenneth Lim of Dataquest, Inc., the San Jose, California marketing
research firm, is the source of the 2.5 percent estimate quoted in the
_Wall Street Journal_, November 14, 1984.
In the long run, moreover, unless Andy Kay made major production
changes, his computers might lose what price advantage they did have.
Automation was improving rivals’ quality control and whittling down
their labor expenses. Eventually, they’d be just a tiny fraction of the
cost of a computer. Apple already had built a $20 million plant to crank
out the Macintosh computer, using Japanese-style manufacturing savvy;
Macs even helped oversee the manufacturing of Macs. “I don’t even know
if the Kays could automate their plant,” said the ex-employee who’d
accused the family of running a feudal society. “You’ve got these four
buildings no wider than a single room. They have a lot of windows, and
that’s great for the people—but hard on computer parts. Often people
placed piles of parts on the lawn overnight when they didn’t have space
elsewhere.” The Kays, to be sure, had put up a semiautomated warehouse
designed by Andy’s daughter Janice and built by his son-in-law’s
construction firm. “But,” said the ex-Kaypro employee, “that’s not the
same as automating the whole plant.” And Andy’s selection of the
architect and construction firm was further evidence of the Kays’
tight-grip management style.
This lack of flexibility could especially hurt them in responding to the
threat from lap-sized machines with flat screens. Sooner or later they
would menace the Kaypro-sized “luggables” in the business market.
About 120,000 flat screens sold in 1983, and although they weren’t
moving off the shelves as rapidly as expected in 1984, sales still might
exceed several hundred thousand.
Some showed 16 lines at one time. That wasn’t as many as the 24 on the
Kaypro II screen, but enough for light-duty word processing. And soon
others would do for the heavy-duty kind and a slew of additional uses.
Thousands of traveling reporters and executives were already using the
Radio Shack Model 100 introduced in 1983—despite its tiny built-in
memory (no more than 24K in a typical configuration) and small screen (8
lines only 40 characters wide).
But where had Andy Kay hoped to get his first flat-screen model? From
Mitsui & Co. in Japan, which would offer it to other U.S. companies,
including perhaps some with marketing clout far greater than his. The
Kays had thought they’d be the only people selling the note-book-sized
computer. By late summer 1984, half a year after the original hoopla,
Kaypro should have been shipping this miracle machine, but a spokesman
in September could not even tell me if the Mitsui deal was still alive.
Then word came out that the deal had fallen through and a flat screen
Kaypro wouldn’t appear until 1985.
A rival company, Morrow, Inc.—headed by George Morrow, a witty, balding
ex-grad student in math, one of Silicon Valley’s better hardware
gurus—had meanwhile acted more successfully. In late 1984 it began
marketing the Pivot, a ten-pound, $2,495 MS-DOS portable with one disk
drive and a well-reviewed clone of WordStar called NewWord. And a
revitalized Osborne—on the verge of escaping the Chapter XI bankruptcy
proceedings—would soon sell its own version of the machine.
Data General, the mini-maker written up in the best-seller _Soul of a
New Machine_, was another threat to Kaypro. In September 1984 it
unveiled the Data General/One, which offered IBM compatibility, a
built-in 737K floppy disk drive, and a 25-line flat screen. A software
developer joked that the letters on the screen were so faint that the
machine should be sold with a coal miner’s hat to read it. But
eventually the flat screen would be just as readable as the Kaypro’s
cathode ray tube. With the inevitable price drops and refinements,
machines like the Pivot and the Data General/One would wreak havoc on
Kaypro sales unless Kay retaliated with the right flat-screen portable
of his own. IBM, AT&T, and Compaq, too, were designing their own
flat-screen machines. And already Hewlett-Packard had been selling a
$2,995 flat-screen computer with the popular Lotus 1-2-3 spreadsheet
program in Read Only Memory—permanent memory.
Another firm—Epson, a branch of Seiko, the giant Japanese conglomerate
known for watches—was shipping a $995 computer with a compressed version
of WordStar in ROM. The screen displayed only 8 lines of information at
a time. But within months a 16-liner from NEC (formerly called the
Nippon Electric Company) hit the market at the same price, complete with
WordStar and built-in gadgetry for talking to other machines on the
phone.
WordStar was now the only word-processing program offered free with the
“luggable” Kaypros. If Andy Kay’s lap-sized micro couldn’t run it or a
clone like New Word, he might be paving the way for some of his
customers to buy rivals’ flat screens that did.
Most buyers used Kaypros to write or type. And how many would relish a
switch to a new word-processing program?
And what about piping WordStar-composed letters and other documents from
the luggables to the lap-size Kaypros? Customers could do so, but then
they might have to insert new commands to tell their printers to
underline or produce other special effects.
Beyond everything else, Andy Kay had better offer good software to fight
rivals selling flat-screen machines similar to his. And WordStar over
the years had generally been the most popular word processor for micros.
4 ❑ WordStar: The Creators
Arthur Clarke is _the_ Arthur Clarke, the science-fiction writer who
gave us _2001_ and that beastly computer named HAL. Seymour Rubinstein
is a California businessman with only fleeting mentions outside the
magazines of the computer trade. He is literate but not literary. When I
talked to him, he had not read Clarke’s latest novel. Yet Seymour
Rubinstein and a colleague played more than a small role in the the
writing of _2010_; for they created WordStar, the word-processing
program that Clarke uses.
WordStar is to micro software what _Citizen Kane_ is to movies: it is
old—by computer standards, anyway—but it’s one of the best of its kind.
If, when writing, Clarke wants to insert a phrase several sentences
back, for instance, he can just zip the cursor to the proper place, then
type in the addition. He needn’t enter a special “Editing” mode. Some
WordStar rivals may break up your thoughts by making you change modes.
Moreover, with simple keystrokes, Clarke can insert, drop, and move
whole paragraphs.
“I can make corrections without hesitation that I wouldn’t have done
before,” Clarke said enthusiastically about WordStar and his Archives
micro. “It’s at least doubled my production with a quarter of the
effort.” His normal output for his books and articles was still around a
thousand words a day, but WordStar had made his writing more fun, and
now he was churning out letter after letter to “my neglected friends.”
“I said I’d retired,” he told me over the phone from his home in Sri
Lanka, “but now I’m working on three or four things simultaneously. I
haven’t touched a typewriter since I got this computer a year or so
ago.”[16] You can easily understand why he hasn’t. No longer, for
instance, does a noisy bell tell Clarke that he’s nearing the right
margin—he just keeps on typing and a feature called =word wrap=
automatically takes him on to the next line.
Footnote 16:
Clarke responded by phone in early 1983 to questions I’d mailed about
his use of WordStar.
Granted, WordStar is traditional software. It doesn’t use =icons=, for
instance, those cute little pictures of wastebaskets or file folders
that some snazzy new programs will flash across your screen to tell you
what you’re doing. And as of this writing, anyway, the original WordStar
didn’t offer =split screens= to show more than one electronic file at
once. WordStar 2000, a related but not identical program, does. First
marketed in late 1984, it may supplant plain old WordStar eventually,
but the original program will always have a special place in the hearts
of the cognoscenti. It’s an indisputable classic. More than a million
people have used WordStar since it appeared in 1979. Rubinstein and Rob
Barnaby, the brilliant programmer who did the actual coding for
WordStar, are legends.
As computer technology matures, the machines themselves will be mere
commodities like televisions or Walkman imitations. It’s the programs
that run on them which will make the difference.
WordStar is at least adequate, and mostly superb, in all but the last of
these areas:
1. Absence of bugs. The software maker should have gotten all the bugs
out so that the program will run reliably on your computer. You can’t
write a complex program that will be absolutely bug-free. But you can
make it bug-free 99 percent of the time.
2. General ease of use. A program should be easy enough to learn _and_
use. Being logical helps; ideally, you can use combinations of
familiar commands to coax the program into doing many things that you
haven’t done before.
3. Good documentation. The manual should be clear and logically
organized. Either the simplest material should come earliest, or else
there should be a beginner’s version; also, the manual should boast a
good, comprehensive index.
4. Usefulness to beginners and old pros alike. You can adjust the best
programs to suit your own level of skill with them.
5. Speed. It lets you do your job fast, especially when you use it with
a hard disk.
6. Power. Related to speed. The program can quickly accomplish
complicated tasks like substituting one word for another in a contract
thousands of words long. Imagine the boon to lawyers plugging new
names into standard boilerplate.
7. Fewer chances for botch-ups. Good programs limit the chances for
careless errors in the first place.
8. The Jewish-uncle effect. Ideally, your software will slow you down or
flash a warning when you’re about to ruin your work with a few taps on
the wrong keys.
9. Damage limitations (if you _do_ goof).
10. After-the-goof feedback. After you’ve botched up—and we all do
sooner or later—the program will tell you how you did so.
11. Ability to customize. You or at least a software expert can
customize good programs for _your_ use on _your_ machine.
12. Availability of “accessory” programs to make your original software
still more useful.
13. Support. Ideally, the software seller will stand behind his product
if you have problems. MicroPro’s record here is far from the best;
this may change.
Again, WordStar isn’t perfect (see Backup III, “The Lucky 13,” for a
more detailed description of it and criteria for choosing good
software), but it comes closer than most other popular programs.
■ ■ ■
Here are a few terms useful to people shopping for a word processor:
1. A =cursor= is just the marker on your screen—a blinking line,
maybe—that shows where a letter will appear when you type. _Cursor
keys_ move the cursor up, down, left, or right.
2. A =file= is an electronic version of a letter, report, or other
document or collection of data.
3. A =control key= is what you start holding down to turn a letter or
series of letters into commands for your computer. Used with the
control key, the letters normally won’t show up in the document
you’re typing. Some programs, by the way, may use the control key
and an =escape= key, which could be a way to get from one part of
the program to another. Or the escape key may take on functions
similar to the control key. In addition, there’s an =alternate= key
that you may also use to help boss your machine around.
4. To =scroll= just means to move from place to place in your
electronic document. It’s like rolling or unrolling a Torah or other
scroll, except here there’s no ink or paper, just patterns of
electrons in your computer memory and dots on your screen. On most
computer monitors you see only 24 lines of typing at a time.
5. A =menu= lists commands on your screen. It can tell you how to
locate a document on your disk or what to do while you’re at work
there.
6. A =block move= is the ability to move material from one part of
your document to another.
7. A =global search= tracks down specified words in your document.
8. A =search and replace= substitutes one word (or group of words) for
another.
■ ■ ■
WordStar’s creators, Seymour Rubinstein and Rob Barnaby, are like Orson
Welles and Herman Mankiewicz of _Citizen Kane_. The movie would have
been hack melodrama without the brilliance of both the producer-director
and his writer; and yet critics still argue over _Kane_’s exact origins.
So it is with WordStar. When _Time_ said Rubinstein had written the
program, a friend of Barnaby’s whipped off an angry letter to the
editor. I won’t take sides here. Rubinstein, however, far from playing
down Barnaby’s WordStar role to me, unhesitatingly passed on his phone
number when I asked. Barnaby is equally willing to acknowledge the
importance of his collaborator. Even the letter writer emphasizes the
usefulness of Rubinstein’s salesmanship and his perception of the
software market back in the late 1970s when WordStar was born.
A short, stocky, bespectacled man in his fifties, Rubinstein grew up in
New York City, the son of a pinball- and novelty-machine owner who died
while Seymour was still in grade school. His mother worked as a clerk.
“My first job,” Rubinstein said, “was a helper on a fruit truck when I
was twelve years old.” Two years later, however, his neighbors were
“letting me ruin their radios,” and by his mid-twenties, he seemed
settled into the routine of a TV repairman.
But Rubinstein grew restless after six months’ reserve military duty in
the 1950s; he attended night school at Brooklyn College and took up
technical writing.