John Ryan

I worked with Independent Newspapers in Middle Abbey Street, Dublin in the 1960s.

I joined the company in 1960 and was based in the works office where a staff of three provided clerical support for newspaper production. The payroll system was critical. To calculate the payroll, the data for each employee was input on a calculating machine and the results were printed on a card. The data was not stored on any medium. To calculate the current week, the previous week’s figures had to be input along with current week data. It was a cumbersome and labour intensive operation.

In 1968, an ICL 1901 mainframe was installed for general computing. The first system to go live was the company payroll. The improvement for the payroll departments was remarkable. The payroll was printed on continuous stationery and pay envelopes were printed. All wages for those on the weekly payroll were paid in cash. Those making up the cash now had what was known as the coin analysis, which gave the exact breakdown of notes and coins required.

A new data processing department was created with a purpose-built air-conditioned room to house the 1901. It was located at the front of Independent House and had special access from the street to take delivery of the 1901 hardware. Normal entry to the room was through an air-lock section which had a special mat to remove dust from shoes. A temperature and humidity monitor in the centre of the room printed readings on a chart. Beside this unit was an emergency button to shut down the electricity supply to the 1901. A raised floor allowed for all cabling and a lowered ceiling for the air conditioning.

There was a viewing area for visitors to Independent House who could see the 1901 in action and no doubt marvel at the magnetic tapes twitching away. This was best seen when the #XSMC sort program was running and using all four tape decks. Papertape punch machines were in an adjoining room, also visible from the viewing area.

The data processing manager, Paul McCarthy, and the senior programmer, Derek Kaye, were recruited from the UK, where they had worked with ICL systems. The rest of the staff were recruited in-house: programmers Harry McCarthy and Dessie O’Brien, computer operators Dennis Moore and Pat Harney, and card punch operators Rita Doyle, Mary Flynn and Rita Hogge.

A vacancy arose on the transfer of Pat Harney from operations to programming in 1969. I applied, passed the aptitude test, was interviewed and got the job. I did a training course in ICL House in Adelaide Road. After some months, I was transferred to the programmers and Liam Hayes was appointed as an operator.

View of the Independent Newspapers computer room in 1968, featuring operator Dennis Moore. A temperature and humidity monitor and an emergency button to shut down the electricity supply are both mounted on the pillar beside the operator’s desk.
(Photo source: Irish Independent 20 September 1968. Photographer unknown.)

The 1901 consisted of a processor running a single program, four magnetic tape decks, a papertape reader, a printer and a teletype terminal. The operating system, then known as the executive program, was loaded into the processor each morning from a reel of papertape. A number of manual switches were set on the processor panel and the papertape reader allocated. The large reel of papertape would be read. This procedure would be successful most times but would fail on occasions. There could be a misread of papertape, the papertape could snag or tear in the reader. The papertape would then have to be rewound and reinserted into the reader. Sometimes the problem would be with the processor giving a parity error, indicated by a sequence of lights on the processor. The unit would be powered off and the loading re-tried. A solution used on occasions involved the operator running a pen along a row of circuit boards in the processor cabinet, an action that became known as “tickling the bricks”. This would tighten or re-seat boards.

At the end of each day, the 1901 was powered off using switches on each unit and a master switch.

All data files and computer programs were held on magnetic tape. Certain tapes were stored outside the computer room in a fireproof cabinet. Specific tapes were allocated for masterfiles.

Tape details were entered on a Kardex system. A number of generations of files were kept and tapes would be overwritten. Using the program #XQMY to prepare a tape for re-use was known as scratching the tape. A number of tapes would be available for general use and these were known as scratch tapes. Each tape had a plastic ‘write permit ring’ which would be present for writing or removed to secure the tape.

The 300 lines a minute ICL-badged printer produced lines of 120 characters selected from a rotating barrel. The required characters were struck by hammers to produce the line in one rotation of the barrel. When printing on pre-printed stationery (invoices, statements etc), the position of each printed line was controlled by a papertape loop located in the back of the printer. This loop was prepared by the operator using a hand punch, usually referred to as the dibbler.

The teletype terminal was used to communicate with the processor. Instructions were entered to load a program e.g. FIND #PR01 #TAPE, to activate a program (e.g. GO 20) and for other parameters required by a program. All input and output from the teletype was printed on a roll of paper (available with carbon copy if required) and formed the record of each program run along with a jobsheet.

The jobsheets were maintained by the operators based on data supplied by the programmers and contained all information required to complete a program run. Each basic jobsheet was created using a typewriter and then photocopied. A photocopy was then used for each job with the variable data (e.g. tape identification numbers) handwritten. Subsequent to each run, jobsheets and logs would be checked to ensure that a run was completed correctly.

The initial applications to run on the 1900 were payroll, newsagents accounts and advertising accounts. Each application consisted of a number of programs, each program limited by the amount of storage available in the 1901 memory. The unit of storage was a 24 binary digit word with a total of 16,384 words available. The lines of code written by the programmer were punched onto papertape, copied onto magnetic tape in source version and then compiled into the version which was loaded into the memory at run time.

Assistance and guidance was always available from ICL and the first program in the payroll suite was written by John Roben from ICL.

Programmers specialised in certain applications. I wrote and maintained programs mainly for payroll and advertising debtors. The latter application processed accounts for advertisers in the three main national newspapers (Irish Independent, Evening Herald, Sunday Independent) and provincial newspapers (Drogheda Independent, Wexford People, Kerryman).

My first few programs were written in Cobol and all subsequent ones in PLAN. Although PLAN was a low level assembler language, I preferred it to Cobol.

The 1901 hardware required regular attention from the ICL engineers, who included Pat Dooley, Mick Venables, Denis Ryall, Jimmy Mannix, Harry Sargent and John Doran. A special room had been included within the main computer room to provide a working area for the engineers and to store machine parts. Scheduled maintenance of the hardware took place every second Wednesday when a half day was allocated. Special diagnostic programs would be run and parts replaced as required.

There was always a good relationship between the Independent data processing department and the ICL engineers. In those days before call centres you could make a phone call and talk to them directly.

There were occasions when machine breakdown would cause critical jobs to be taken out to ICL computer installations in other companies. This was fairly straightforward to organise as far as resources were concerned. As all files were on magnetic tape, they could be taken out to another site along with papertape and stationery. It was a matter of finding free time on a suitable machine. A number of companies could provide that time, including Helys in Santry, H. Williams in Dundrum, May Roberts on Grand Canal Quay, Player Wills on South Circular Road, C&C in Ballyfermot and Guinness’s at James’s Gate.

Other staff who worked on the 1901 in the early 1970s were systems analyst Joe Garvey, who was recruited from Becton Dickinson, and programmer Linda Laffan from Lever Brothers.

The 1901 was in use until 1977 when it was replaced by an ICL 2903.

I remained in the department working on the 2903 and subsequent ICL systems – ME29 and Series 39 – until the year 2000. The advances in the technology were remarkable, but the 1901 always brought back fond memories.

Last edit: March 2018

© John Ryan 2018