16 May
2007
16 May
'07
3:13 a.m.
Yes slight aberration of the brain. The building was called "The Atlas
Computer Laboratory" but the system was ICL.
One of my first jobs as a junior engineer (circa 1972) was to refurbish
4K core memory stores for an Elliot system.
They were about three feet long with a Mullard Core stack in a box in
the middle and plug in cards on either side.
They used early transistors of the OC71 era (Yup -ve supply). I soon
learned all about read, write and sense amplifiers and how the cores
actually worked.
As a final test I would put the store back in the system (Elliot 4100?)
load a very complex FORTRAN program from paper tape ( it worked out
handicaps for large racing yachts based on their dimensions) then a data
tape. The answer came out on one of two IBM golfball printers (no
keyboards).
That's what I call classic computing as opposed to a classic computer!
Rod Smallwood
-----Original Message-----
From: cctech-bounces at classiccmp.org
[mailto:cctech-bounces at classiccmp.org] On Behalf Of Roger Holmes
Sent: 15 May 2007 19:44
To: cctalk at classiccmp.org
Subject: Re: Harwell (was RE: A local computer history group for my area
. ..)
From: "Rod Smallwood" <RodSmallwood at
mail.ediconsulting.co.uk>
Well now let me see now I left School for College in 1964. My major
subjects were Industrial Electronics and Computing.
We were taken to Harwell Research centre to see 'The' Computer. It was
an ICL 1900 series system. It took up three floors of
a substantial
building. Input on the top floor (80col Cards and papertape).
Processing
and storage on the middle floor and output on the ground floor. Had to
be that way or the vibration from the line printers
would have shaken
the building to pieces.
I remember it was called Atlas and even had an operating system called
George III. So that's just over 40 years ago. In
computer terms
certainly vintage if not veteran.
Hi Rod,
I think you have mixed up two different systems. The Ferranti Atlas
(later ICT/ICL Atlas) was in a custom built building, I understand the
CPU occupied two floors, and yes there was one at Harwell I've been
told.
I would expect that Harwell also had at least one ICT/ICL 1900, and that
would indeed run the George 3 multi access system.
I did my computer science degree at Queen Mary College (University of
London) and we had a ICL 1905E, upgraded to a 1904S which ran George
2 batch processing and the Maximop multi access system. Later I used a
1906 (S?) at the Royal Aircraft Establishment and it used either George
3 or George 4, and I was relieved to find that all the Maximop commands
worked on it.
When I was in the sixth form we visited the computer centre at the
NatWest tower - huge floors - IIRC, one floor had nothing but IBM CPUs
(probably 370s - this was 1969) and other static electronics, one floor
had all the storage - magnetic tapes, discs, and juke box like machines
which picked up strips of magnetic tape. Another floor had the hard
peripherals - card readers, line printers and a massive document reader
with a curved 'retina' on top about 20 feet across where all the optical
sensors were. Another floor had all the cheque reading machines, the
operators were moaning about people stapling their cheques and the
staples getting caught in the works. At that time a random 2% of the
cheques were also processed manually for quality assurance purposes, and
presumably to prevent fraud. That day was the first I heard of the
fraction of a penny (or cent) scam, where all interest payments were
rounded down and all the fractions of a penny credited to the
programmer's account.
The same day I visited the University of London's central computer
centre (as opposed to the individual college's private ones). They had a
CDC 7600 and one operator was using a terminal to send messages to chat
up a girl operating one of the remote stations which fed the 7600 with
punched card data and printed the results. A very early form of Internet
chat room you might say.
Roger
Owner of a ICT 1301 built 1962