|The story of Gary Kildall's Great Missed
Opportunity has passed into folklore. But what is the truth behind this
cause celebre? Gordon Eubanks, Symantec supremo, told Clive
Gary Kildall wrote CP/M, the first mainstream
desktop operating system. He invented the concept of a Basic Input Output
System (BIOS), the core logic which marries hardware to the operating system.
He was a founding father of desktop computing, yet history mainly recalls
his greatest mistake. He was the man who gave away the IT industry; the man
who gave Bill Gates the world.
The story goes that two suits from IBM had arranged to meet him at home on
a certain day in 1980. Kildall was off flying his plane, and had left
his wife Dorothy to do the talking. She balked at signing an agreement
to not disclose anything they told her, and showed them the door.
Nonplussed, the suits then approached a fledgling company called Microsoft
about the small matter of developing an operating system for the first IBM
Such is the legend, already enshrined in alt.folklore.computer. Only it wasn't
quite like that, according to one man who was around at the time. Gordon
Eubanks founded Symantec, one of the biggest software companies to have
grown fat by plugging the gaps left by Microsoft. He knew Kildall
from way back in the early seventies.
They were very different characters. Eubanks was drafted into the
Navy during the Vietnam War, and stayed on to get sponsored for graduate
school. He formed his first company while still a student and has been an
aggressive, even predatory, businessman ever since.
Kildall, a specialist in compilers, was one of his tutors and a brilliant
programmer, but by all accounts was out for a good and easy life. He wrote
CP/M (Control Program for Microprocessors) in 1973, almost as a by-the-way,
to help him develop software for the 8-bit 8080, one of Intel's first
Eubanks couldn't understand him. "I remember having lunch with him
one day and he said to me, 'I don't know what to do with the CP/M.' So I
said, 'You had better make it a business.' And he said, 'I am not sure if
people will buy it.' I replied, 'Oh Gary, come on...'." It is now
more than two decades later and Eubanks still shakes his head in astonishment.
Within five years of having had that conversation, hundreds of thousands
of machines, using Z80 or 8080 processors, had been sold with CP/M as their
operating system. Kildall had formed a company called Intergalactic
Digital Research, which he later shortened to Digital Research, and became
For a time, Eubanks was in direct competition with Bill Gates,
selling rival versions of Basic to run on the CP/M machines. He was still
in the Navy, his mother was running his company from her home in California,
and he decided it was time to get out. "Then Gary offered to buy the
company at a really high price. I think he paid ten times revenue for it
in Digital Research stock, which ended up being worth a lot."
Eubanks joined Digital Research but left after two years. "It became
clear to me that Digital Research did not have the will to win and they were
losing opportunities. So I went off and did my own thing."
The problem was that events had been too easy on Kildall. "He felt
everything was in his court, and he could do whatever he wanted. This was
where Gary and I had a hard time... where we did not get on very well."
Disaster had nearly occurred some time earlier when hard disks were introduced
and CP/M would only support floppies. Hardware manufacturers, tired of trying
to get an upgrade out of Kildall, almost reached the point of developing
a rival operating system. "All of a sudden Gary realised that business
was starting to dry up because the floppy-disk systems were not selling.
People wanted hard disks and high-density disks."
Kildall finally ordered a crash program to write a CP/M upgrade. "When
something like that happens... it's like when someone has a heart attack,
they get a whole new view on life and start to work out... But Gary
never realised how close he came to losing his business, and he did not change."
The same thing happened all over again when Kildall was slow to bring
out a CP/M upgrade to run Intel's new 16-bit 8088 and 8086 chips. Tim
Patterson, an 8088 boardmaker at Seattle Computer Products, got so tired
of waiting that he wrote his own operating system, called QDOS. "Tim
got frustrated, as did a lot of people, about Gary's attitude to this
kind of thing," Eubanks recalls.
IBM had been slow, too. It was still stuck in the age where a computer filled
a room and could be used to milk its owners of millions. IBM did not want
to know about desktop computers and didn't want anyone else to know either.
By the end of the seventies, the microcomputer business had become too big
to ignore. IBM decided it had to get in on the act. It could not afford the
time to develop its own model from scratch, so the decision was taken to
build a machine from off-the-shelf hardware components and bought-in software.
Kildall's Digital Research was the obvious place to go for an operating
system, hence the famous visit to the Kildall home. Eubanks
says: "I've told this story to lots of people and they just won't get it.
All they want to get is that IBM showed up and Gary was off flying
his aeroplane. The problem is that this is very wrong."
For one thing, Kildall never dealt directly with hardware manufacturers.
He left that to his wife Dorothy. "Gary was very laid-back.
He didn't care that much. Dorothy ran the business and he ran the
technical side and they did not get on." And who could have known that the
IBM PC was going to be important? "IBM was just one of dozens of companies
who were in the [microcomputer] business."
Dorothy was talking to some people from Hewlett-Packard, Digital
Research's biggest customer at the time, when the IBM representatives showed
up on the doorstep. She was in the throes of preparing to go on holiday the
next day. "That was what really caused the problem," says Eubanks.
That, and the contrasting characters of Gates and Kildall.
"The real issue wasn't that Gary refused to talk to IBM. The real
issue was that Microsoft had a much better vision for the business.
Gary was very laid-back. He did not care that much. And Bill
was extremely focused and driven."
Gates did not even have an operating system at that stage. After IBM
called, he promptly bought Patterson's QDOS for $50,000. It was little
short of a CP/M clone, but it was to become MSDOS and run nine out of ten
of the world's desktop computers.
News of the deal spread quickly. Patterson rang Eubanks, warning
him to port his Basic to the new operating system. "I said, 'Jeez,
Tim, why is that?' And he said, 'I can't tell you, but a big Seattle
company has just licensed it, and licensed it on to a hardware company that's
bigger than anyone you can think of.' I said, 'Let me get this right. You
are telling me that IBM licensed it from Microsoft.' Tim said, 'I
didn't say that but you should definitely support it'."
Digital Research pioneered pre-emptive multitasking, and its GEM graphical
operating system was more successful than early versions of Windows. But
the company never regained the pre-eminence it had in the seventies, and
was bought by Novell in 1991. Kildall died in 1994 at the age of 52,
from head injuries received during a night out in Monterey, California.
These days, Eubanks regularly pauses in London to brief journalists
about the latest products from Symantec, a company he bought in 1982 from
the proceeds of his early business ventures. Gates' move in buying up QDOS
seems to have provided something of a model, because Symantec has grown by
a series of similar strategic acquisitions, including Central Point Software,
Peter Norton Computing and, most recently, Delrina.
As Eubanks puts it, "[Symantec's] strategy is to focus on businesses
with good growth prospects and the opportunity to become market leader...
We use acquisitions to accelerate entry into key markets." About his early
success, he says: "I was lucky. I was in the right places at the right times."
His last word on Kildall is: "Gary could have owned this business
if he had made the right strategic decisions."