The numbers game: computing and computers

From Earthwise
Jump to navigation Jump to search
From: Wilson, H.E. Down to earth - one hundred and fifty years of the British Geological Survey. Edinburgh:Scottish Academic Press, 1985.
The text is derived from an 'orphan' work. BGS are committed to respecting the intellectual property rights of others. After extensive effort we are unable to trace the copyright holder of this work. Despite this, we would like to make this very important work on the history of BGS, written by a former member of staff of the Survey, available for researchers. If you are a rights holder and are concerned that you have found this work for which you have not granted permission please contact us with proof you are the rights holder.
XVII The numbers game: computing and computers. Chapter-head sketch by Robery Geary.

The numbers game: computing and computers[edit]

The history of the application of electronic computers to geological projects in the Survey goes back to 1965 when W A Read, then in the Lowlands Field Unit, was inspired by a visit to Edinburgh by two American sedimentologists, Krumbein and Merriam, who were developing the technique of trend-surface analyses. Read first published a paper on the use of computers in such analyses in 1966 and in the following years he and J H Dean produced a series of papers on the statistical treatment of the considerable data available in the records of the Edinburgh office. Read subsequently became one of the few geologists to be granted 'sabbatical' leave when he went to the Kansas Geological Survey for six months in 1970-71.

When Dunham became Director in 1967 he was met by a proposal from Read that the system of computer storage of geological data, already in use in the oil industry, should be applied to the IGS archive. In the following year the Director's Report records the work being done in the Edinburgh office in conjunction with the Universities of St Andrews and Reading.

At Dunham's first meeting of senior officers the subject of computing was raised and D A Gray, newly promoted to Head of Hydrogeology Unit ,remarked that the tentative work being done in various parts of the Institute should be co-ordinated. He promptly found himself chairing a Computer Committee and, as he says, it took him seventeen years to get rid of the responsibility for the subject.

In 1969, Victor Loudon, from Reading University, was recruited to the post of Computer Liaison Officer and the Computer Unit appeared as part of the Hydrogeology Department, under which aegis it continued till 1974. By the time Loudon was appointed the use of main-frame machines, operated by the Science Research Council or by commercial organisations and accessed by remote terminals, was developing rapidly in a dozen units throughout the Institute and the first in-house computer had been installed by the Geochemical Division at Grays Inn Road. Much of this early activity was the result of individual initiative and Loudon had great difficulty in persuading the activists to channel their activities through his new unit.

By the early 1970s computers were in use throughout the Geochemical, Geophysics and Mineral Resources Divisions, the Hydrological and Palaeontological Departments and elsewhere, and extensive use was being made of commercial, Science Research Council and Edinburgh Regional Computing Centre machines, as well as the few in-house installations. In addition the NERC Experimental Cartography Unit in South Kensington was working closely with the Drawing Office and the Radioactive and Metalliferous Minerals Unit on the automation of map production.

In 1974 the Computer Unit became a separate entity and Loudon and his small staff were given the thankless task of co-ordinating all this activity. The main concentration of users had developed in Edinburgh, where IGS computers were installed in the Regional Computer Centre until Murchison House was able to accept them. In London the Exhibition Road, Grays Inn Road and Princes Gate offices all had terminals linked to the Atlas Laboratory of the Science Research Council. In conjunction with staff from the Atlas Laboratory a new data-handling software system was developed, suitable for a wide range of geological applications. Known as G-EXEC it was to be further developed during the rest of the decade.

The mid-seventies saw a build-up of Computer Unit numbers to a maximum of about twenty, though in this effervescent discipline there was a rapid turnover of staff. The new Keyworth office was provided with a representative, and a link with the network, in 1977. Using external facilities, notably the Rutherford Laboratory and the Univac Computer in the Institute of Hydrology at Wallingford, Loudon and his staff were making progress, albeit on a shoe-string budget, when in 1977 the Unit was reviewed by a Visiting Group, which also looked at computing facilities in other NERC stations. For the next two years there was a virtual stalemate while the Group and NERC tried to evolve a solution to the problems they had found. In spite of the evidence, becoming ever more clear, that the future trends in computing would be towards smaller but more powerful machines — a desk-top computer in the eighties has the power of a room-full of grey cabinets in the 1960s — it was decided that there should be a Centralised Computing Service for the whole of NERC, based in Swindon, and depending mainly on the in-house main-frame computers at Wallingford (Institute of Hydrology) and Bidston, Cheshire, (National Institute of Oceanography). As a result the IGS Computer Unit ceased to exist in October 1979 and the staff were transferred to the NERC Computing Service.

If the existing systems had been absorbed gradually into the new organisation it is probable that staff would have adapted to new methods, but the tone of the new NERC Computing Service (NCS) was set by the hectoring attitude common in the Swindon HQ and peremptory directions and prohibitions caused the rapid loss of most of the experienced IGS computer staff and a virtual breakdown of the system in the early 1980s. The obsession of the new service with independence from outside equipment — including the SRC Atlas — and concentration on the upgrading of the specialised Honeywell computer in Bidston, to act as a remote access main-frame for the whole organisation, was the cause of a great deal of ill-feeling, including the threat of the Department of Energy to withdraw much of its contract work from IGS.

The story of the NERC Computing Service and its hardware is outside the remit of this account, but it may be recorded that its early years were less than satisfactory for computer users in IGS.

The computer saga was to be paralled a few years later by the struggle to adopt word-processors in the Institute. The first of these was obtained, using 'customer' funds, in 1980 to enable publication deadlines to be met in the Mineral Assessment Programme for the Department of the Environment. The spectacular success of this machine and the desire to develop the use of these systems soon led the Swindon cohorts to demand control of all word-processing equipment.