Early days in ERCC - 1967-1972

The re-discovery in the EUCS store of a set of minutes from 1967-1972 gives an interesting glimpse of the very different world of the early days of the university computing service. In late 1967 ERCC had a KDF9 mainframe. Essentially it was usable only by those able to program in Algol 60 or Atlas Autocode and users had to write virtually every program from scratch. The entire set of ready-to-use applications software was a handful of mathematical and statistical routines, plus an open-ended survey analysis program which needed a custom-written Atlas Autocode procedure for each execution.

There were of course no networks, no e-mail, no desk-top keyboards and only the most primitive word-processing, since lower case printers were rare and variable-pitch printers non-existent in the university. Most users, including ERCC staff, considered themselves lucky if their program was run more than twice a day.

Meanwhile in the US there was beginning to be considerable growth in academic software circulating around universities and government research institutions. Because at the time hardware was perceived as expensive and software as cheap, much of this software was distributed at no, or only nominal, charge. Huge increases in productivity were potentially available by bringing in and promoting the use of ready-made software for common computations but there was very considerable reluctance amongst users to trust the results of programs not written by themselves.

A complicating factor was that much of the American software was written in Fortran, a language not promoted in Edinburgh at the time. Further, current thinking was that ERCC should produce its own documentation for all library programs, in a standard format, as well as a catalogue for the library; this was potentially a tremendous burden.

The Program Library Sub-Committee was set up to address these problems. Their minutes show the progress achieved .

Access to Fortran was obtained at the National Engineering Laboratory, East Kilbride, and a daily courier van was used to transport cards and listings. (Early communication systems took many forms. Rothamsted Experimental Station used UCL machines thanks to an old age pensioner commuting to London with a suitcase full of punched cards). Later, an IBM 360/50 was installed at ERCC: this offered a much richer variety of compilers and applications software.

Program Library Sub-Committee minutes follow. These are most conveniently seen if the browser is configured to open a pdf file automatically.

Dec 1967

The committee is set up. There is hesitation about the extent to which academic departments will collaborate.

Jan (1) 1968

Start of documenting programs and searching out new ones.

Jan (2) 1968

Overview of current program library.

Feb 1968

Identifying gaps in the basic library.

Apr 1968

Looking at System 4, IBM and ICI software. Consideration of Fortran, Algol, PL/1.

Jun 1968

ICI and Atlas Lab representatives attend. Discussion on documentation standards.

Oct 1968

ICI offer their software for £30,000. ERCC begin to import US software. Departments are asked what help they need.

Dec 1968

Documentation development continues. ICI software is criticized. ERCC tries to get NCC to produce library catalogues and documentation standards.

Feb 1969

ICI software offer is rejected. Demand from departments is summarized. The IBM 360 is being installed at ERCC.

Apr 1969

IBM and SHARE software is explored. There are now 491 programs in the catalogue.

Nov 1969

ICI is willing to negotiate. A questionnaire is sent to all 400(!) users.

Dec 1969

There is an appeal to users to identify gaps in the library. A working party on survey analysis is set up. Future meetings will discuss particular areas of applications.

Jan 1970

Report from the survey analysis WP. SPSS would be a good idea. However one member would prefer the old 'write it yourself' approach.

Feb 1970

Overview of Physical Science experience. A 'simple' library should be established to bring in current non-users.

Apr 1970

SPSS is now implemented.

June 1970

Overview of Engineering Science experience. ERCC has £2,000 in its software fund.

Dec 1970

Phase II of the committee with new membership. Remit: to consider less orthodox applications, to review the library and to coordinate new work. [Not so interesting from now on but minutes retained for completeness.]

Feb 1971

Developing documentation. First Social Science SPSS course.

Mar 1971

More on documentation, including a list of bought-in manuals.

Apr 1971

Location of reference sets of documentation.

May 1971

Seminars on specialized field of application. Presentation to Computer Board on their visit.

[Aug 1971]

The Program Library Unit is formally established within ERCC with the remit of promoting the use of library software throughout all UK universities. PLU became a separate university department in August 1974 and re-merged with ERCC to form EUCS in 1987.

Jun 1972

Overview of PLU activities. Reference to ICL New Range (2900). By default the final meeting of committee.