|
 In
my Web Line column in Issue 4, I referred to an Internet site
called The Dead People Server. One of its entries that I chose to
quote concerned Grace Hopper, famous in her own field of computing, but
little known outside it.
As
a matter of fact, I can think of three notable Americans this century with
the Hopper surname, each in a different profession. Apart from Grace,
there have been the more familiar Hollywood villain Dennis (b. 1936) and
the great painter Edward (1882-1967).
In
this essay, however, I want to celebrate the unique achievements of Dr.
Grace Hopper, mainly because in a former life I myself worked in
computing, where I was able to benefit directly from her pioneering
efforts.
Grace
Brewster Murray was born 9 December, 1906 in New York City. A B.A.
by 1928, she earned an M.A. in 1930 (the year she married Vincent Foster
Hopper) and a Ph.D. in maths in 1934. For several years she was on the
faculty of her old college, Vassar, as an associate maths professor.
A
woman with a doctorate in maths was rare enough in the mid 1930s, but all
that was merely a prelude to her moment of destiny, when in 1943, during
World War II, she joined the United States Naval Reserve, commissioned as
a lieutenant. Because of her maths ability she was assigned to the Bureau
of Ordinance Computation Project at Harvard University, where she became
the third programmer of the first large-scale computer (the Harvard Mark
I).
From
an early age, Grace had been good with gadgets, and would disassemble
alarm clocks just for fun, so when she saw the Mark I, all she could think
about was taking it apart and figuring it out:
That
was an impressive beast. She was fifty-one feet long, eight feet high, and
five feet deep. And she had 72 words of storage and could perform three
additions a second.
Forget
its lack of speed for a minute and focus on its memory capacity.
Seventy-two words of storage. That's around 500 bytes in today's
terminology. Your home PC has at least 16 million bytes.
It
was while working at Harvard that Grace Hopper is said to have coined the
term "bug" for a computer fault. The bug she discovered was a
moth which had caused a hardware error and which she duly pasted in the
project logbook.
In
1946, she was returned to inactive duty, but continued to serve in the
Naval Reserve. Three years later she joined what in the future became the
Sperry Corporation to work on the first commercial computer. She had
already come to believe that programming did not have to be a difficult
task, and one of her first innovations was to introduce the idea of
reusable program code. Till then programmers constantly had to repunch
certain commands for every program they wrote. Hopper encouraged her team
to create the commands once and keep them in shared code libraries. This
reduced errors as well as programmer stress.
Trouble
was, computers only read binary (zeroes and ones). Herein lay her next
great challenge. Hopper was convinced that programs could be written in
English and then translated into binary code by another piece of software
called a compiler, and set about proving it. By 1952 she had an
operational compiler that recognised 20 English statements, to be used for
typical business work (such as payroll and billing), and published the
first paper on the subject.
This,
my PC friends, was three years before Bill Gates was born.
But
at first Hopper met little more than scepticism.
Nobody
believed that. I had a running compiler and nobody would touch it. They
told me computers could only do arithmetic.
She
persevered, however, and was a crucial part of the team in Sperry's Univac
Division which in 1958 finally released Flow-Matic, the world's first
marketable English-language business compiler.
Flow-Matic
was the progenitor of another and more famous language, COBOL, which
appeared the following year, after the concerted efforts of a group of
U.S. government users and computer manufacturers led by Hopper herself.
The acronym COBOL stands for COmmon Business-Oriented Language, and was
designed so that applications could be more expeditiously developed for
the rapidly burgeoning commercial computing market, no matter what
computer they were intended to run on. More importantly, it allowed
applications to be transferred from one brand of machine to another, with
minimal reprogramming.
You
know, despite the hype surrounding the usefulness of the desktop computer,
there are still many more programs written in COBOL in use today than in
any other language. Indeed, it's been estimated that worldwide there are
about 300 billion lines of computer code, fully 240 billion lines of which
are in COBOL. In fact, it's in those COBOL programs that the notorious
"millennium bug" is more likely to lurk than in your home
computer. Not that it's the fault of COBOL. No, the decision to store all
dates without the century (with "70", for example, implying
"1970") was the result of limited capacity on tapes and disks
attached to mainframe computers in the 1960s and 70s
Hopper
worked on at Sperry until 1967, when the Navy recalled her to oversee a
project to standardise its computer programs and their languages. She
remained in the Navy until her retirement in 1971, aged 65.
But
within a year she was persuaded out of retirement to full-time active duty
once more. In 1983 she was appointed to the rank of Commodore, the title
of that grade being changed to Rear Admiral in 1985.
She
retired for the second and final time in 1986. At 79, she was the Navy’s
oldest serving active-duty officer. Despite her grand age, she was
determined to remain active and useful, and immediately became a senior
consultant at the Digital Equipment Corporation.
Intellect
and persistence and are two attributes that made her a great leader. She
was determined not to let anyone obstruct her vision of creating a much
wider audience for computing. Not surprising, then, that during her
advance up the Naval ladder, she had to induce a lot of people to change
their habits. Every day she heard somebody say, "but that's how we've
always done it". Hopper believed that change was good, and needed.
I'm
going to shoot somebody for saying that someday. In the computer industry,
with changes coming as fast as they do, you just can't afford to have
people saying that.
To
prove that things did not always have to be done a certain way, she had a
clock on her office wall that ran counter-clockwise.
During
her long and productive life, Grace Hopper received many awards from the
worlds of both academia and commerce. One of the most striking occurred
when she was named the first computer science "Man of the Year"
by the Data Processing Management Association in 1969.
In
1991 she was awarded the National Medal of Technology, the first woman to
receive America's highest technology award as an individual. It recognised
her as a computer path-finder, who spent a half century helping keep the
U. S. on the high tech leading edge.
She
died on New Year’s Day, 1992.
A
biography by C W Billings was published in 1989: Grace Hopper: Navy
Admiral and Computer Pioneer.
After
over forty years of trailblazing work, Grace Hopper considered her
greatest accomplishment to be, not, as you might expect, her indispensable
contribution to COBOL, so much as all the young people she had trained.
She was an inspirational professor and a much sought-after speaker, in
some years travelling the world addressing more than 200 conferences and
meetings.
I
was fortunate enough to be in one of those audiences. It was towards the
end of 1968 or early in 1969. I had recently become a COBOL programmer
myself and joined the Australian Computer Society, under whose auspices
her Australian visit was arranged. I remember I arrived a bit late at the
AMP Theatrette at Circular Quay that evening, and by that time there was
standing room only. It was the Society's most successful guest speaker
meeting thus far — hundreds had come to see and hear and meet this lady
in her sixties who was already a legend to us.
She
spoke first about the importance of reusable code, and then about the
history of COBOL itself, which is what most of us were there to hear
about. In the late 1960s she was well into systematic verification of
COBOL compilers, for which, she said, she had a developed standard bunch
of punched cards to test any new releases from manufacturers. (Yes, we
were still all using punched cards then.)
And
she wanted us to know of her love for the Navy, too. Having extolled its
virtues, she asked,
You
all know the Pentagon, I assume? But does anyone here know what's at the
centre of the Pentagon building? No? Well, now you do, and you can tell
all your friends. It's a café.
In
her talks Grace Hopper often used analogies and examples which, like
herself, have become the stuff myths are made of. One of her party tricks,
which she now demonstrated for us Sydneysiders, was as follows. She bent
down to the voluminous bag she had with her and extricated a piece of wire
about a foot long. This she stretched out before us.
You
know what this is? It's a nanosecond. It's 11.78496 inches long, this
wire, and that's how far electricity travels in one-billionth of a second.
|
