by Lee Keet (as told to Luanne Johnson)
Copyright ©1998, Luanne Johnson, All Rights Reserved
In 1967, I was working for IBM in sales and decided that IBM’s track record with
the System 360 provided a opportunity to go out on my own. I started originally with IBM
as a systems engineer and was assigned to a project to develop a bill-of-materials
processor for manufacturing companies. This was part of IBM’s strategy to create
software that would help to sell the hardware in various vertical markets and which would
be given away free to the customers.
After working as a systems engineer for awhile, I decided to get into sales because
there was more money to be made there. I had to fight my way into a sales territory
because I was looked at as a heavy hitting engineer and not a guy who could ever sell
anything. They initially gave me a sales territory that didn’t have anything to do
with my experience with manufacturing systems but once I had proved myself as a salesman,
they gave me back the manufacturing territory and I made it my expertise to sell the
manufacturing software products along with the hardware to large manufacturing accounts.
And that’s how I got really interested in software as an entity.
The 360 was such a success from a hardware sales standpoint and such a disaster from an
installation standpoint that another IBMer and I decided that we really needed to go into
the software business to help customers get their 360s installed. We started turnkey
systems, inc. in April, 1967, and our original goal was custom development contracts, not
software packages. We were reasonably successful doing large custom development work and
were one of the first developers of online systems in the Northeast.
IBM brought out their first display terminal, the 2260, which had an obvious
application for online order entry and online activity. But we discovered that good
software in between the application program and the hardware was non-existent. IBM had
some programs that had been developed by systems engineers in the field for various IBM
customers, included one called CICS, but we looked at it and thought, boy, that’s a
piece of junk. So we said, why do not we develop our own software package to interface
IBM systems to 2260 environments?
This was in 1969. Initially the idea was to write a product that would help us deliver
the custom programming contracts for our customers but we ended up getting hooked on it.
There’s no other way to put it. We saw that what we were selling to Company A led to
enhancements that could be used for Company A and induce a sale to Company B. I guess if I
went back I would say that we couldn’t get out of it because we were always running
to meet our commitments. With the optimism of youth, we were always selling the system
with a few more features than it actually had at the time and then rushing to upgrade it
in time to make the delivery.
In the early days, it was a daily process. Because the people that were developing the
product were the same people who were writing the documentation and designing the
advertising campaigns and the marketing brochures. The same people who were getting on the
phones and saying can I come see you and tell you about this wonderful product and then
going out and selling and installing it.
One of the guys who worked for us in the early days was Joe Farrelly who later was Vice
President of Research and Development at ADR. I sent him off to Atlantic National Bank in
Florida to install our product, which was called Taskmaster, under a contract that
included several features that hadn’t been developed yet. So he would install the
system during the day and develop the new features at night. He’d be on the phone to
me and the other guys in Oxford, Connecticut, telling us that we need this or the customer
needs that and we would develop along side of him and rush him code, sometimes reading it
to him over the phone. I do not think the man slept for five weeks.
But that’s the way that a software product got developed in those days. We sold
one of the early versions of Taskmaster to American Tobacco. We had announced it as a
multi-tasking system but in actual fact we were using a much simpler technology we had
invented that subsequently became known as pseudoconversationality. Which was actually
fine for American Tobacco’s purposes because true multi-tasking isn’t necessary
unless you’re running 20 transactions a second through the thing and they
weren’t even running two transactions a second through it.
But one day I got a call from the manager and he said we did a test and it’s not
truly multi-tasking. I said, oh, didn’t you get the memo on that? And he said, no, I
didn’t. So I wrote up and sent him a backdated technical memo that said multi-tasking
in this version had been disabled because of a serious design flaw and that it would be
repaired in the next release of the product. Then I assembled my entire crew and we went
up to Oxford, Connecticut, where we rented computer time from Uniroyal. And we invented
the first multi-tasking system in the history of this business and shipped it to American
Tobacco among others.
And that’s the way you get hooked on this commitment to vaporware and fighting to
keep up. I’m very pleased to say that I do not think we ever disappointed a
customer and we never announced anything that we didn’t deliver. But those were hairy
days.
 top of page
 return
to the list of anecdotes
|
