I wrote:
"My computer probably has more disk storage than existed in the whole
world in 1960."
I meant total data storage in 1960. There were only a handful of 5 MB
IBM RAMAC hard disks back then.
Mike Melich and I made some back of the envelope estimates of total
world data storage circa 1960 the other day and came to this conclusion.
Very roughly, the largest single collection of data circa 1960 was
probably the U.S. 1960 census data. This was the first census to be
stored electronically. It was stored on 7,297 UNIVAC II-A tapes (1957
model), which I believe had 6 MB of capacity . . . but I am not sure.
Maybe 3. Anyway, 6 would be 43 GB total. It is likely that in the
entire world, there was not more than 100 times this amount of data
storage in other facilities such as NASA, the insurance companies,
major industrial companies, airlines and so on. That's 4 TB, which is
just about as much as I have, including backup disks.
I am making another estimate based on a 1968 computer textbook, which
references the number of computer installations at that time (~90,000).
In the early 1960s when I was 8 or 10 years old, I visited the data
processing faculties at the Census Bureau, where my mother worked. It
made an indelible impression. This was the largest and most advanced
computer facility in the world at that time. I toured the warehouse
facility where thousands of tapes were stored up to the ceiling with
a woman who gathered tapes into a shopping cart. I watched dozens of
tape drives, CPUs and printers at work. The programmers then sat me
down and gave me a large printout, explained what a rounding-off
error was, and set me to work looking for one. I remember being
impressed that the machine was complicated enough could make errors
of that sort.
I do not recall finding an error. I never imagined that I would spend
a large part of my life looking for similar computer errors and bugs.
As Mike McKubre said about calorimetry, "when I considered doing cold
fusion in 1989, if I known that I would spend the next 20 years
worrying about calorimetry, I would have run screaming from the
room." Years ago I recall reading the autobiography of a computer
pioneer who described being flummoxed by errors in bugs in the
earliest computer code. Encoding algorithms was much harder than
anyone imagined it would be. He wrote (as I recall): "it gradually
dawned upon us that we would spend most of the rest of our
professional careers looking for program bugs."
- Jed
