Research comments: The Chronicle charts the development of the first two types of LEO computer up until the decision by Lyons to buy a LEO III. It covers LEO I from the first suggestion that the Lyons board allow Thomas Raymond Thompson (TRT) and Oliver Standingford to go to the USA to study "the possibilities of electronic calculation".

The go-ahead for the development of LEO II is given on 6th July 1954 (page 21).

• 1947-1949
  • Ernest Lenaerts start work at Cambridge (on EDSAC) 1st December 1947 (page 1).
  • John Pinkerton starts work at Lyons as an Electronic Engineer on 17th January 1949, despite the fact that EDSAC does not yet work. He starts by studying clerical procedures at Lyons before going to Cambridge to study EDSAC (page 2). By 8th April 1949 he needs an assistant and a draughtsman to work with him to progress plans for the calclulator (page 3); Ernest Kaye is employed as the assistant on 15th August 1949 (page 4).
  • The first experimental circuits are prepared at Cadby Hall on 8th March 1949 (page 3).
  • The team develop ideas on the conversion of data from decimal to binary on 15th March 1949 (page 3).
  • EDSAC successfully runs its first job on 9th May 1949 (page 3).
  • By 23rd May 1949, Standard Telephones and Cables are invited to help develop the calculator's auxiliary input/output equipment (page 3).
  • On 11th July 1949, Wayne Kerr Laboratories agree to produce electronic circuits for the calculator and by 26th July they are given the first specifications for prototypes (page 4).
  • John Simmons suggests the name 'LEO' on 14th September 1949 (page 4).
  • The Coventry Gauge and Tool Co. are commissioned to produced LEO's mercury delay tubes on 5th October 1949 (page 4) and the first ones are delivered to Lyons on 22nd March 1950 (page 6).
• 1950
  • Directors, managers and supervisors are given a demonstration of LEO's capabilities (such as exists at that point) on 7/8th March 1950 (page 6).
  • The first plans are made for a photo-electric reader to be developed by Lyons for data input in March 1950 (page 6).
  • Thomas Raymond Thompson (TRT) moves to the LEO project on 5th April 1950 (page 7).
  • Less than a month later, David Caminer also moves over to the project to take charge of programming (1st May 1949, page 7).
  • The first problems with Standard Telephones and Cables' development work are revealed on 30th May 1950 and by October they seem significant (page 7).
  • Tony Barnes joins the LEO team on 20th November 1950 (page 7). He takes over responsibility for the operational side of LEO on 29th July 1953 (page 17).
• 1951
  • Plans were made in January 1951 for the Bakery Sales Valuation Job (L3) to be LEO's first job (page 9).
  • On 15th February 1951, Princess Elizabeth (who became Queen Elizabeth just a year later) visited Cadby Hall and sees LEO in action (page 9).
  • LEO is demonstrated performing clerical work to Lyons directors on 17th April 1951 (page 9).
  • LEO staff are treated to a dinner on 16th May 1951 at Lyons' Trocadero to celebrate their successes (page 9).
  • LEO is demonstrated performing clerical work to Lyons managers and staff reps on 17th/21st/22nd May 1951 (page 9).
  • The Standard Telephones and Cables incomplete auxiliary equipment is demonstrated to Sam and Anthony Salmon and others on 26th June 1951 (page 10).
  • LEO performs its first successful clerical job, Bakery Sales Valuations (L3), accurately on 5th September 1951, although at this point the calculator cannot perform the task regularly (page 10).
  • On 5th October 1951, due to continual delays the LEO team decide to develop alternative input/output systems, running a new development programme in parallel with Standard Telephones and Cables' work (page 10).
  • The first non-Lyons job (ballistic computations) is offered to LEO by Brigadier Hinds of the government's Ministry of Supply (the Ordnance Board) on 16th November 1951 (page 11). On 14th January 1952 Lyons were given the contract for undertaking daily ballistic computations (page 12).
  • The first of the routine runs by LEO of the Bakery Sales job takes place on 7th December 1951 using data for the week ending 30th November 1951. The job is done by computer from that point on (page 11).
• 1952
  • By January 1952, LEO is undertaking calculations for the Meteorological (Met.) Office (page 12).
  • On 10-12th January 1952, LEO was given a reliability test producing tax tables for the year. It ran for over 59 hours, of which over 51 hours was useful work giving an operational efficiency of 87% (page 12).
  • Data is read from and recorded onto magnetic tape during a wages program run on 8th may 1952 (page 13).
  • The LEO team meet with British Tabulating Machines (BTM) on developing input/output equipment for LEO on 13th May 1952 and on 21st July 1952, they agree that they will supply Hollerith machines (page 13).  On 17th November, BTM agree to supply tabulators, card feeders and card punches for use with LEO (page 15), with the first tabulator being supplied on 6th May 1953 (page 16). The tabulator prints payroll results for the first time on 29th may 1952 (page 17).
  • The Lyons Teashop Orders job (L2) moves forward in June 1952 (page 13).
  • Lyons' own photo-electric tape reader starts routine use in June 1952 (page 13).
  • Lyons agree to produce Annuity Tables using LEO for the Institute of Actuaries on 29th July 1952.
  • Lyons agree to use LEO for guided missile calculations for De Havilland Propellers on 18th September 1952 (page 14).
  • By October 1952 the team realise that modifications will need to be made to LEO to work with input/output equipment other than Standard Telephones' (page 14) and work to prepare the Bakery Valuations and Payroll programs to work using the alternative equipment begins on 17th November 1952. The alternative equipment is codenamed 'consolidation facilities' (page 15).
  • Standard Telephones' output equipment manages to produce payslips on pre-printed forms on 31st December 1952 (page 15) and the first successful trial of the Lyons' Payroll program (L1) takes place on January 1953 (page 16).
• 1953
  • On 24th March 1953, the Inland Revenue approach Lyons about using to LEO to calculate PAYE tables (page 16).
  • The first time that data for the Bakery Valuations job (L3) is input using punched cards and decimal/binary data conversion is automatically performed is on 5th August 1953.
  • On 19th August 1953, Ernest Lenaerts proposes the future use of a new shorter tube mercury delay line store. The pulses for these tubes would be one quarter of a microsecond in width, reducing the access time to the store to a quarter of the previous speed (page 17).
  • The decision to abandon the problematic Standard Telephones input/output equipment was finally made on 25th August 1953.
• 1954
  • On 6th April 1954, LEO I was intended to be used to calculate the new tax tables as part of the government's budget. In the event however, no changes were made to tax rates so LEO's services were not required (page 20).
  • The general approach to building LEO II is agreed on 14th January 1954 (page 19) and detailed proposals for the second machine are made available by John Pinkerton on 4th May 1954 (page 20).
  • On 27th May 1954 a dinner is held at Lyons' Cumberland Hotel to celebrate the inauguration of "the first high speed fully-automatic electronic calculator to be designed and built for general commercial work" on Christmas Eve 1953 (page 20).
  • The Lyons board give the go ahead for the construction of LEO II. At the same meeting the board agree to build other mark IIs for sale or hire to other companies through the establishment of a subsidiary company LEO Computers Ltd (page 21).
  • The Lyons Tea Blending job (L4) is first run on 2nd October 1954 (page 21).
  • Trials of the Teashops job (L2) begin on 10th October 1954 (page 21) and regular runs begin on 30th October 1954 involving 42 teashops (page 22). By 21st November 1954 LEO runs the program for all Lyons' teashops (page 22)
  • Airmec are commissioned to produce storage units for LEO II (page 22).
• 1955
  • Staff involved in developing the LEO II prototype move from Cadby Hall to Olaf Street, Willesden on 17th January 1955 (page  24)
  • Sargroves are commissioned to develop a photo-optical document reader for LEO II (page 24).
  • Jim V. Smith joins LEO Computers as a programmer on 31 January 1955 (page 24).

Provenance :
From David Caminer's papers