The Kaypro I prototype, by Stan Brin,
author of the famous SmartKey utility for the Kaypro series:
There was a Kaypro I, but only one of them. I
received the Kaypro I for one afternoon to demonstrate at a trade show.
The Kaypro I was the prototype. It had Single-sided, single density
floppies, similar to the Osborne, and were mounted vertically, one on each
side of the monitor -- which was mounted in the center, not the on the
left, as in the Kaypro II. It was colored GREEN.
The first Kaypros came with a really bad WP called "Select" and
no MBasic, but at first that wasn't a sales hurdle. The Kaypro's large
screen and rugged metal case were major selling points.
Once WordStar became available in 82, it was EXTREMELY popular, especially
among students and writers - I sold hundreds to Hollywood screenwriters
and many of the great scripts of the 80's were written on KayPros.
But a program was required - SmartKey, my program.
I was the designer and publisher of SmartKey, but I didn't write the code,
or come up with its name.
The author, Admiral Nick Hammond of the Royal Australian Navy, wrote an
early version while studying in the US.
I discovered that SmartKey could be the basis of an important commercial
program, mainly because it was the only way to write a screenplay at the
time. It was the first TSR, or Terminate but Stay Resident
In 1985, we came up with a dedicated version of SmartKey for the KayPro
with a real popup window.
The KayPro era lasted less than two years. It was outsold by the IBM-PC,
which was WAY more expensive and couldn't do any more, but was much better
marketed and had the promise of extra - but pricey - memory.
Farrell, former employee
of Kaypro/Non-Linear Systems clarifies:
I am writing to clarify what seems to be some confusion regarding Kaypro
Corporation's relation to Non-Linear Systems.
During the 1980s, I worked in the purchasing department for
Kaypro/Non-Linear Systems. Kaypro Corporation was created as a division of
Non-Linear Systems, and both products, Kaypro's line of computers, and
Non-Linear System's line of digital readout volt meters, were manufactured
at the Solana Beach, California facility.
I have to agree with my former colleague Rick Bartlett. The Kaypros that
actually worked when they came out of the box were found to be very
durable, as was their original intention when Andrew Kay envisioned a
portable computer that could be used at the job site by his daughter and
son-in-law in their architectual/building business.
Kaypro had a high out of the box failure rate at the dealer level due to
poor quality controls at the manufacturing plant. I can remember siting in
my office and watching Grampa Frank Kay accidentally turn the sprinklers
on a rack full of brand new Disk Drives waiting to be installed on the
line. Although, the majority of units that got into the hands of customers
worked well as most Kaypro Dealers tested units themselves before sale to
The original self-contained, aluminum
"bread box" design was very innovative at the time, as was the
bundling of a variety of software that came with it. I had one of the
original desktop models known as the "Robie" at home, and used
it for many years.
Kaypro/Non-Linear Systems was a fun
and interesting place to work in the early days, and is definitely worthy
of a place of honor in the history of personal computers.
Darrel Pittman mentions:
The motherboard of the original Kaypro II was
essentially a knockoff of the Xerox 820-II, and the Xerox "BigBoard II".
The Kaypro II could run much of the same software.
Lee Patterson reports:
As a Kaypro II owner, I can tell you:
1. It arguably had the BEST-engineered keyboard
ever produced for a personal computer, made by Keytronics. [Notwithstanding
lack of function-keys; those weren't implemented yet]. It had a smooth
tactile feedback, good key travel, and the keys were sculpted (indented)
both left-to-right and top-to-bottom, across each key. Therefore each
key formed a shallow dish to hold your finger [deeper dishes on "F"
and "J"], and tended to resist position-creep as you reached
for the top row, etc. This is similar to older IBM Selectric typewriters
(with the interchangeable font-globes). For some reason, the keys on virtually
all modern PC keyboards are merely sculpted in the left-to-right dimension,
forming troughs instead of bowls. This sounds relatively insignificant,
but makes a difference to a fast touch-typist. And it was built sturdy,
if somewhat heavy -- almost mil-spec.
2. The Kaypro II had no fan (it was cooled
by convection) and was thus silent, except for the smooth keyboard clicks
and occasional floppy disk access.
3. The keyboard hooked up to the main (CPU)
unit via a standard telephone handset cord, making replacement a snap
(no pun intended). Likewise if you needed a longer keyboard cord, many
were thus available at the local phone store. Even though the bottom of
the keyboard was too large and flat to make it a useful 'laptop' style,
the longer-cord capability made it easier to set up the keyboard on the
desktop, and have the cord routed indirectly to a shelf or such above
the desk (remember, computer-desks hadn't really been marketed yet), leaving
more desk space clear for books and supplies. To transport, you could
just unsnap the keyboard cord, and toss it inside the upright CPU (CRT
at top), before closing it by snapping on the keyboard against the CRT
face. There were little feet on the back of the CPU so it could stand
upright. The feet flared out at bottom, so they doubled as posts around
which to wrap the attached AC power cord. Kaypro really did think of everything,
and engineered it well.
4. The software included BASIC by a little company called "Microsoft";
last I heard they're still around ;-) It could be programmed, as well
as run public-domain games like "Adventure!", one of the CLASSIC
one-person role-playing games. (Starts something like, "You are standing
in a dry streambed by a rusty metal grate that is locked..." You
would [after retrieving the key from a nearby building] get into traps
like "You are trapped in a maze of small twisty passages, going off
in every direction"). Some people even worked out a 'map' of the
locations, based on exact phrasing of such 'maze' passages. A lot of fun,
before decent graphics
5. The 5x8 character-matrix did include
true 'descenders', if I recall correctly. Without them, it was harder
to read lower-case type on screen. Such features as that, plus the 80-character
lines and bundled software (new idea) like (professional-level) Perfect
Writer etc., made it a smart choice for serious writers, which is why
we chose it. We already had a Commodore 64, which was a great machine,
but not for writing. (The Commodore had 40-char. lines, no descenders
[?], mushy keys, non-standard layout, no number pad, fuzzier display on
NTSC video to TV, proprietary printer port [only went to thermal printer,
mainly good for listing code] instead of IEEE-488 parallel interface on
the Kaypro going to, say, an Epson MX-80.)
6. Perfect Writer was not WYSIWYG - there
wasn't the ability to draw characters on-screen with essentially graphic-functions.
It was an ASCII-based layout (I can't remember the name for that standard).
So if you wanted to have a boldface phrase "This is really boldface"
with the word "really" in bold italics, you would have to type
it as, "@B[This is @I[really] boldface]". So of course, your
on-screen layout could be very different than your printed output. Likewise,
the onscreen display was fixed 80-character lines, so if you had selected
a proportional-spaced font on your printer (we had a JUKI daisy-wheel
printer that did some of that), your line-lengths would not be the same
either. I think the rationale was that, in a world where modem protocols
(and printer drivers, for that matter) were still being developed, it
was an attempt at standardization and simplification of transmission;
though others may have better information on the subject.
7. There were some aftermarket upgrades
that came out near the end of Kaypro's transition to MS-DOS. I remember
we had our floppy disk drives upgraded to quad-density, double sided drives.
960K (?) each. Likewise, there was a faster replacement processor (I forget
which one, but I think it was 2X the speed) which I had installed, that
could be invoked by a standard switch on the back panel. (Not all the
software would run at the faster speed). Also, there was an anti-glare
screen cover (our Kaypro II had a standard high-gloss CRT face); it was
a very fine cloth mesh which was surrounded by a plastic frame, and the
cover slid over the CRT face trim on the console unit. It worked pretty
8. Speaking of the display, I found it much
easier on the eyes -- even restful -- to have Kaypro's dark screen with
green letters, than a white screen with black letters (which I realize
looks somewhat like a piece of paper -- more important for today's desktop-publishing
capabilities). I attribute this to Kaypro's reduced illumination requirements,
compared to today's white-screens with full-screen, full-spectrum, full-time
illumination of a low-persistence phosphor, flickering at [whatever the
refresh rate is]; which I suspect may have something to do with workplace
discomfort (cf. various studies of the effects of working under flourescent
lights --- same principle). But I could be wrong. (Of course, with today's
desired full-color multimedia functionality, it's a moot point.)
9. As you have said, the machine was incredibly
rugged and well-engineered. That kind of durability is generally not designed
into consumer equipment today. I truly miss the Kaypro II's friendly and
durable presentation, and would use it again if it ran today's software.
In 1985, Arthur C. Clarke published a sequel
to 2001 : 2010 Odyssey Two. He worked with Peter Hyams in the movie
version of 2010. Their work was done using a Kaypro computer and a modem,
for Arthur was in Sri Lanka and Peter Hyams in Los Angeles. Their
communications turned into the book The Odyssey File - The Making of 2010.
SBasic bug, by Jacques Guy:
I don't think that the Kaypro came with
Microsoft Basic, though. SBasic (a compiled dialect of Basic) yes. But
not Microsoft Basic. I have an anecdote about SBasic. I was writing a
program to compute the mean and standard deviation of a data set. It
gave obviously wrong values for some data. A bug? Must be. I spent a
whole night looking for the bug. Eventually, in despair, I tried
calculating sqrt(10^20). The answer? 10^17. Ahem!