Showing posts with label Alan Turing. Show all posts
Showing posts with label Alan Turing. Show all posts

Monday, April 13, 2009

The Ballad of the Troubleshooter

In the end of March I wrote an article about Manchester Mark I. Some days ago the article got plugged by Slashdot geeky newsite and as a results thousands of readers poured on my blog to read it. This was nice but it brought with it a surprise even better. I got comments from two of the people who actually worked on the Mark I machine.
In December 1953 Owen Ephraim was a newly appointed Ferranti maintenance engineer sent to maintain the Mark 1 at Manchester University. His job was to to find and repair whichever of the 20,000 valves or 100,000 components had failed. He worked with Alan Turing and as the MTBF (Mean Time Between Failures) was as low as 1 hour, Turing was delighted to get a few minutes running time after re-entering his software on tape or by hand, whenever he handed the 'working' machine back to him after repair. Owen tells that Turing usually had a useful suggestion as to which 'module' was at fault and then left him to identify and replace the component by sight or touch. (bright valves are OK, cold components were probably gone). The big danger was to have a second fault develop while looking for the first. In that case all evidence became doubtful.
He continues:
"Turing was quiet and focused on his work. We did not chat nor have any contact that was not related to the work in hand. He was 17 years older than myself and I had no idea of his previous work.
A second 'engineer' joined me in early 1954 and I was instructed to teach him. He was called Frank ? as I remember, good technically but a bit remote socially. On one occasion I found him in the computer room when I arrived on the Monday morning. He claimed to have spent the nights in the broom cupboard as he did not wish to travel home. I did not report his misdemeanour and it was not repeated. He left abruptly soon after Easter as I remember and I now suspect he was an MI5 plant to watch Turing.
I was not particularly surprised when Turing did not turn up on Tuesday the 8th as he was absent from time to time without warning. I did not hear about his suicide until the Thursday or Friday. I 'cleared my desk' and returned to 'normal duties' at the main 'works' in another part of Manchester, a few days later.
8 months later I was sent to Italy to install the Mark I in Rome
"

And in Rome was also the second guest of my blog who also commented the Manchester article and shared his memories of that glorious time. In 1955 Prof. Roberto Vacca was maintaining a Mark I machine located at the National Research Council in Rome, Italy. He worked out a theorem in number theory (distribution of 0's and 1's in powers of integers) and Owen Ephraim was his mentor. In 1957 Roberto wrote a ballad while he was freaking out watching an oscillograph. There was an issue with the machine: every 10-30 hours it miscalculated certain sums when the number to be summed contained a 00100 configuration which in duotrigesimal (base 32) reads @. So he wrote a program (nicknamed tete-à-tete) which repeatedly summed @@@@ = 00100 00100 00100 00100 and figured out the issue.
THE BALLAD OF THE TROUBLESHOOTER
Note from the author:
"The ballad is bawdy, but it was written under stress -- many hours of staring at an oscilloscope in turns without even batting our eyelids to watch a rare fault occurrence. So the vulgarity may be forgiven - after 52 years. I guess there is a statute of limitations."
Sweat is tickling down my face,
The taste thereof is salt:
I'm running on a race
With an intermittent fault.
I don't think I can cope
With the hunt for the bloody clod
To stare at the bloody 'scope
Is a job for a bloody sod.
I do begin to hate
The bright traces on the screen;
It shouldn't be my fate,
But the fate of a bloody queen.
The boss he doesn't guess
What a difficult job it is.
To clear this bloody mess
Takes bigger ball than his.
To clear this bloody muck
Takes more than half a mo'
And blast and curse and fuck
Who says it isn't so!
Roberto Vacca, 1957
You can reach Roberto at his website http://www.robertovacca.com/inglese.html. He is also a writer of essays and novels. You can read about his books and buy printable copies here.

Friday, March 27, 2009

Manchester Mark I - Loveletters

I know I'm quite late (or early) for a St. Valentine article but I just got late and really can't wait until the 2010 cupid.
I just wanna share with you this curious piece of history of the information technology.


The Manchester Mark I, also nicknamed "The Baby" is known to be one of the earliest electronic computers and, as the name suggests, it was developed by the University of Manchester, UK, starting from August 1948.
The Mark I is also famous because from its academic environment it spawned the world's first commercially available general-purpose electronic computer, the Ferranti Mark I.
Programming one of these illustrious ancestors was tricky. They had no defined assembly language and the instructions had to be submitted in binary code. On one intermediary version of the machine, the programmer had to manipulate switches to input programs. The output device consisted in a CRT (cathode ray tube) "screen" displaying a series of dots.
From August 1953 to May 1954 strange love letters began to appear on the the notice board of Manchester University's Computer Department. Something similar to the following example.

HONEY LOVE
YOU ARE MY DEAR PASSION: MY ADORABLE FERVOUR: MY ARDENT INFATUATION: MY ARDENT DEVOTION. MY PASSIONATE LUST BREATHLESSLY HOPES FOR YOUR LIKING. 
YOURS BURNINGLY
M. U. C.



All these letters were signed M.U.C. and I bet they were quite a chat facilitator for the campus people which did spot them, day after day, on the notice board. Remember that we were in the fifties and certain expressions of love and passion were maybe far too frank for the taste of the well bred people of the time. I can imagine the surprise when those people discovered that those proofs of feelings came from the "mind" of a machine. As a matter of fact "M.U.C." stood for "Manchester University Computer", and the love letters originated from a programme designed and written by Christopher Strachey (1916-1975), one of the researcher on the Manchester machine. Exploiting the built-in random generator of the Ferranti Mark I, the first commercial computer in history, he managed to generate texts with a purpose.
Years before, on 1949, the Manchester Mark 1 -an early academic-only computer- had successfully run a programme to search for Mersenne primes. The announcement of the success of this task caught the interest of the British press which widely reported the event using the sentence "electronic brain". This description on its turn generated a discontent in the Manchester University's Department of Neurosurgery. Sir Geoffrey Jefferson, of neurosurgery at Manchester University, expressed on the following terms when relating about the subject:
Not until a machine can write a sonnet or compose a concerto because of thoughts and emotions felt, and not by the chance fall of symbols, could we agree that machine equals brain-that is, not only write it but know that it had written it. No machine could feel pleasure at its success, grief when its valves fuse, be warmed by flattery, be made miserable by its mistakes, be charmed by sex, be angry or miserable when it cannot get what it wants.
I don't know if Strachey took inspiration to this debate for writing his programme, but true is that his intent was to show that the computer could be set up to produce artifacts which could arouse emotions into those who interacted with its output. Maybe it's also worth remembering that Manchester University's Computer Department was also the work place of Alan Turing (1912 - 1954), the father of modern computer science and one of the fathers of artificial intelligence. In 1950 Turing created the so called Turing Test with the purpose of being a reference upon which testing the ability of a machine to "disguise" itself as a human.
By the way, fast forwarding the years until nowadays, Mr. David Link created a Manchester Mark I web-emulator which runs the loveletter program by Christopher Strachey. You can access the page here.



Here's a list of resources on the subject.
Pictures and media are Copyright of David Link and The University of Manchester. They are here reproduced under the terms of the Fair Use.