COPElAND | 159
scope. Nor did Colossus incorporate Turing’s all-important stored-program concept. The com-
puter was programmed by hand, by means of large panels of plugs and switches. The female
operators used these to (quite literally) re-wire parts of the computer each time it was required
to follow a new set of instructions. This laborious method of programming seems unbearably
primitive from today’s perspective when we take Turing’s glorious stored-program world for
granted. ENIAC, too, was programmed by re-routing cables and setting switches. It might take
ENIAC’s operators as long as three weeks to set up and debug a program.^23
The rest of the world did not know about Colossus, but its impact on Turing and Newman
was colossal indeed: Colossus was the connection between Turing’s groundbreaking article of
1936 and his and Newman’s post-war projects to build a universal Turing machine in hardware.
Turing’s opportunity to realize his dream of a stored-program all-purpose electronic com-
puter came in October 1945 when he joined the National Physical Laboratory, a government
research establishment situated in the London suburb of Teddington. Newman’s opportunity
came almost simultaneously, with his appointment in September 1945 to the Fielden Chair of
Mathematics at the University of Manchester.
Turing’s ambitious Automatic Computing Engine, or ACE, is described in Chapters 21 and
22, and the more modest Manchester ‘Baby’ computer in Chapter 20. It would have been fitting
if Turing and the National Physical Laboratory had won the race to build the first universal
Turing machine in hardware. But life is not always fair: as Chapter 20 relates, Newman’s group
at Manchester romped home.