[spectre] (fwd) David Link on Very Early Algorithms, Science Museum London, 12 March 2009

[Andreas Broeckmann]

Date: Tue, 10 Mar 2009 11:06:23 +0100
From: "Dr. David Link" <david at khm.de>

:::

The Archaeology of Very Early Algorithms, 1948-58 - Christopher
Strachey's Love Letter Generator
Date: 	12th March 2009
Time: 	14.30 Room is open from 14.00 - meet up with society members
Location: 	Fellows Library of the Science Museum, Exhibition Road,
London, SW7 2DD


About the seminar

This talk will describe early steps in the programming of the first
computers, illustrated by some interesting examples.

David Link will present his reconstruction of the "loveletters"
algorithm (1952) by Christopher Strachey together with his
implementation of an emulator of the Ferranti Mark 1 and the project
currently underway to build a functional replica of the machine. He will
provide an overview of the software developed in this formative period
of British computing, and discuss examples of the software he managed to
locate.

In June 1948, a team led by Frederic Williams and Tom Kilburn built the
first fully electronic, digital, stored-program and universal computer
worldwide at the Electrical Engineering Department of Manchester
University: the Manchester "baby" prototype. It was based on the first
reliable means of volatile storage, the Williams tube. Max Newman and
Alan Turing from the Mathematics Department played an important and
still disputed role in the conception of the machine.

  From 1949 to 1950, the computer was extended and modified on a daily
basis, to become the "Manchester Mark 1". In 1951, the firm of Ferranti
industrially produced an `engineered' production version of it, the
"Ferranti Mark 1", of which two were sold, one to Manchester University
and one to Toronto University.

The machine was programmed in four-letter operation codes from the 5-bit
Baudot alphabet, common in teleprinters and wartime cryptography. Soon,
a number of conventions were created to make programming less
cumbersome, using "Schemes" and "Autocodes".

It is estimated that already by 1953, some 50 scientists were employing
the machine in their research. The main disciplines they were coming
from were x-ray crystallography, molecular chemistry, meteorology,
engineering, and nuclear physics.

When the Ferranti Mark 1 in Manchester was replaced in 1958, it was
disassembled, and this is why only very few parts of the machine remain.
Unfortunately, the situation is similar for the software developed for
it. Despite intensive efforts, I could only unearth an alarmingly small
fraction of these pioneering programmes. A part of British
techno-cultural heritage of international importance is at stake.

About the Speakers

Dr David Link took his PhD in philosophy in 2005 at Humboldt-University,
Berlin, with a thesis on text generating algorithms in the early years
of computer development ("Poetry Machines / Machine Poetry"). He works
as a theorist, artist and programmer. His current research focusses on
the convergence of mathematics and engineering in the early 20th
century. For further information see his website, and the article there
titled "There must be an angel. On the beginnings of the arithmetics of
rays", in: Variantology 2 (2007).

Background

The Digital 60 website has more about the Ferranti Mark 1 computer, and
about programming on the Ferranti Mark 1.

http://www.digital60.org

A brief profile of Christopher Strachey is in a recent issue of
Resurrection.
-- 
David Link
Alpha 60 technologies
/Systems/ Development
T: +49 173 244 81 57
M: david at khm.de
W: alpha60.de