moinsen / Software historyDE ↗
← 1951

Year overview

1951

Context and research status

Recherchestand: teilweise · 19. November 2026

Maurice Wilkes, David Wheeler and Stanley Gill published The Preparation of Programs for an Electronic Digital Computer in 1951. Its subtitle refers to EDSAC and libraries of subroutines. ↗

Change and influences

Computers were provided for different tasks: the Census Bureau accepted a UNIVAC I, while J. Lyons & Company used LEO I for bakery calculations. Sources differ on some dates of acceptance and operation. UNIVAC · LEO

Reusable subroutines were intended to make program preparation easier. Event

Languages and techniques

Wilkes described microprogramming as a systematic way to design a computer’s control unit. Event

Ferranti’s special-purpose Nimrod displayed its decisions in the strategy game Nim with lamps and could be slowed down for demonstration. Event

The Computer History Museum dates regular operation of MESM near Kyiv to 1951. Event

Geoffrey Hill programmed Australia’s CSIR Mark 1 to play music around 1950–1951. Repeated pulses at the speaker produced tones whose pitch depended on their timing. Event

The third edition of the exponential tables corrected an incorrect value from the 1939 and 1947 editions. Event

The EDSAC book discussed libraries of subroutines: reusable routines could save work and return to the calling program after execution. Event

Key players

According to the Computer History Museum, John Bennett proposed the Nim machine; Raymond Stuart-Williams described its construction. Event

Wilkes, Wheeler and Gill are named as the book’s authors; Eric Mutch had made a major contribution to the earlier Cambridge report. 1950 report

Market and everyday use

At the Festival of Britain, visitors could play against Nimrod and watch a computer at work. Event

Ferranti delivered a Mark I to the University of Manchester; in the US, the Census Bureau acquired a UNIVAC I. A 1952 agency report dates a contract to 1948 and acceptance to 30 March 1951; later agency accounts differ. Ferranti · UNIVAC

LEO I ran bakery calculations for J. Lyons & Company in November 1951. The company later also used the computer for payroll. Event

Further evidence remains open.

Events

Figures and facts

A later LEO chronology records a correct bakery calculation on 5 September, but not yet regular operation. Event

A museum article dates the completion of a full bakery run to 30 November based on a contemporary report. A chronology gives 7 December for the first routine run; the relationship between the accounts remains open. Comparison

According to David Caminer, LEO valued the output of about twelve Lyons bakeries; sources give late November or early December for the start of routine processing. Date comparison