User:James Donnelly Duff

Story about James D (Jim) Duff, an early Computer Major Software Creator

This document is the fact that I claim to be one of the earliest COMPUTER MAJOR SOFTWARE CREATORS in Australia, and possibly in the world. The year was 1966 when I moved to the Australian Capital City, Canberra, and started writing a major part of the software.

Readers can jump ahead to item 6 as the main part of this story.

The first 5 item listings created in the start of this story are related to the early parts of my working life, as I like to be able to mention how I started my working days.

There are also several coloured items to show the MAJOR WORDS.

1.Early working days

After leaving secondary school at the age of 16, my first two years were with a bank in Geelong and Melbourne, Victoria, Australia.

I was then asked by a family friend Bill King, who was working in the Commonwealth Department of Social Services (DSS) in Geelong, to join him working in that department. Working in that second employee, my weekly wages came from 10 pounds per week to 20 pounds per week - wow.

I joined the (DSS) on 15th March 1960 and stayed there until January 1966 in Departmental jobs within Victoria, Australia, and then from February 1966 I was introduced into Automatic Data Processing (ADP) in Canberra, ACT Information Technology (IT) as it is known today.

Left the department on - to 20th April 1978 – and went to London to work for nearly two years.

During my early days at the Geelong Regional Office where I started in 1960, I worked in the Pensions and the Unemployment and Sickness Benefits (USB) areas until 1962, when the Department decentralised the Family Allowances Branch from the State Head Offices to Regional Offices. Whilst I wished to do that work at the Geelong office, a more experienced person from Melbourne, Malcolm Elliott, was given the job, so I took a similar position in the next regional office to be decentralised in Wangaratta, Victoria.

Attached is a copy of the front page of the Wangaratta daily newspaper, the Chronicle Despatch, dated September 5, 1962, showing a photograph of the following people:  Mr A J [Bert] Bartlett, Registrar at Wangaratta, Mr L W [Laurie] Loveless, Mr P J [Pat] Cosgriff, Mr Jim Duff, and Misses Linda Hawley and Margaret Price.

This is part of my working history, and the reason for mentioning this is that there is a reference to Brian Withell soon in this story.

2. Early learning weeks

In 1964 I moved from Wangaratta to Melbourne and did an EXIT (EXaminer In Training) Course, and this was the year the Beatles came to Australia.

In 1965, I did the 14th ADP Systems Analysis Course in Melbourne, run by the Public Service Office in La Trobe Street, that provided the Program Training subjects. We learned to program in machine language that we would take to the RMIT (Royal Melbourne Institute of Technology) in Swanston Street, on an Elliott 803 Computer that they had recently bought.

Other members from DSS on that 1965 course I entered were Tony Slattery, Ted Burton and John Brewer from the Sydney Office.

3. Early software working years

Early in 1966, I moved from Melbourne to Canberra, ACT to join the first lot of software writers.

The Department of Social Services was an early pioneer in the implementation of computer systems in Australia, and the first to use the programming language called COBOL (COmmon Business Oriented Language).

4. Early mentioned names of workers doing early data analysis

Providing some members’ positions from memory, Kevin Kimball was in charge of the ADP Branch, Tony Battanta was in charge of the Pensions ADP Group, and Brian Withell was the Pensions Project Manager. I had met Brian three years earlier, when he and another Systems Analyst visited the Wangaratta Regional Office to do Systems Requirement Specifications for the processing of Child Endowment (an ADP system I would work on many years later).

In addition to the higher level directors and project managers etc, I was one of four members of the

Pensions ADP programming group that implemented the major part of the first computer system that went live in 1967. This took over from the punched card systems that had been in operation for many years. Prior to going live, a trial system had been written by Bill Lewisson and Jim Humphreys, in COBOL, on a Honeywell H200 computer, comprising the NSW group of Widows Pensioners.

5. Early storing of information prior to being used by the computers

The systems in those days were based on sequential batch processing where the records were stored on magnetic tapes. A series of programs would read a tape, or set of tapes, created by a previous program, do their own functions and then write the updated data onto an output tape, or tapes, which would then be passed on to the next program in the process. Run diagrams showing the program sequences and passage of data between them are attached, and the following paragraphs are a precis of the programmers involved.

6. The first four major software programmers

Peter Carmody wrote the edit programs, PEN02E and PEN04E. The first set of data was contained in batches of punched cards that were initially checked for validity and written onto magnetic tapes. The ‘clean’ data from the two edit programs was then passed to the Update program, PEN06U, written by Ian McDonald. I AM James D (Jim) Duff, the writer of this document, and back then, I wrote PEN07B, which was a ‘Breakout’ program, essentially the second half of the main update program that split out the processed data into Payments, Statistics and Reporting data files. The following year, Ian and I combined PEN06U and PEN07B into one program. The Statistics data was then processed by the PEN08A program written by Brian Christensen.  Then, the reporting data was processed by PEN11F written by ME.

7. Some other named SOFTWARE Programmers as well

The Payments data created on tape by the Pensions System was then passed over to the Payments System that sorted it and printed the cheques. Some payments, however, were converted to a different tape format (5 columns instead of the 8 columns used by the DSS system tapes) and sent to the Commonwealth Bank where the data was used for direct transfers. Some names from the DSS ADP Payments Group are Len Ryan, Graham Kerr and Mike Whitelaw.

8. Brian writing an early general software program

It could be said that Brian Christensen wrote the first spreadsheet program, years before PCs hit the scene. It was PENSPRES (Pensions Special Reporting System), a facility to enter your own table of input values, and have the answers come out the other end. Of course, the input was punched cards, and the output hard copy, not like the visual data entry that came about years later.

9. First update programs

After the Pensions and Payments systems went live in 1967, I wrote the first update programs to increase the Pension Rates in 1968 (LEGAMEND) and 1969 (PACT69), following the annual legislative amendments. Ian McDonald had to update the PEN06U program to contain the new Pension rates, and take into account the updated War Pension rates as part of the means test assessments. I have photographs of myself and Ian taken at that time in 1969. We also appeared in an article in the April 1970 issue of Careers Honi Soit, produced by the Students' Representative Council of Sydney University, and distributed to schools by the Commonwealth Employment Service.

10. How hand written software papers were early written

Back in those days, program code was hand written on paper grids representing the 80-column punched cards, that would then be passed to the female operator who created the punched cards. It was a standard practice to write in pencil on every second line so that any amendments could be made with an eraser, plus using the alternative empty lines. However, Ian McDonald and I (Jim Duff) were so self-assured (and self centred, no doubt), that we used to write our code with fountain pens – on each line! I still have my Lamy Fountain Pen that I bought that year, but I believe Ian lost his in-Cyclone Tracy four years later. As a result of our close working relationship, the pair of us were known as MacDuff.

In addition to writing software, I also wrote the Pensions User Manual, of which I still have a hard copy in my library.

Some Technical Background Information

In 1965, the Department generated a Specification for Automatic Data Processing Equipment, and  in June 1966 the General Electric Company won the tender with its specification for a GE 615 computer for Sydney, New South Wales. After doing training courses in 1965 and other initial planning work during 1966, the Department had to cancel GE's tender, as it was unable to provide a second computer as a backup to the original one to be installed in Sydney. In early 1967, a reissued tender was won by IBM who specified a System 360 Model 40 to be installed in Sydney that year, with the ability to provide a backup computer, and install subsequent equipment in Melbourne a year later.

Technical details of the IBM computer model can be found at http://en.wikipedia.org/wiki/IBM_System/360   but, for the purpose of explaining the generation of pre-printed forms, it must be noted that the printers in those days had a chain of capital letters in one font only, with a maximum number of 130 characters per line;  ten characters per inch, and six lines per inch. The usual continuous printer paper was 15 inches wide by 11 inches long.

In order to make 'pre-printed' output forms, one had to design a blank form and leave room for the printer to fill in the data accordingly. This is what I had to do for three forms, two of which are enclosed as samples. The first one was the Record Printout, Form SA99, for which I had to use a special small font typewriter, and lay out the item labels as shown. In order to test out the program, I then cut out the areas where the data would be printed, and use the form as an overlaid view on plain paper printouts.

This initial design was a 'one-up' form that used only part of the printer's width. The following year, I designed and programmed a 'two-up' form that would be more efficient by using up the full 130 characters per line.

The second similarly developed form was the Pension Advice, Form SA100, that was issued for each new grant or pension amendment transaction entered into the system. A sample of this form is also enclosed. The tricky part of this form was the concatenation of a series of phrases, based on a string of phrase index numbers, to make up the relevant sentences. The next step was then to know where to split the sentence to fit on each line - now known as word wrap.

There is no sample available for the third form, the Form R, Form SA101, that was issued as a Review form each year to check the current state of the means test for each pensioner.

Several workers came to Canberra to learn how to manage the software running in the new computers. In 1969 a lady called Catherine Ross McCormack was a lady who I met, and two years later we got married.

In late 1970, when the Department upgraded to a small number of User Terminals comprising a keyboard and monitor, I was involved in the design of the Pensions Online Enquiry System (POLES). This enabled customer service officers to look up details of changes made to a pensioner's data from the computer rather than search for the paper files containing their details. Enclosed is a copy of the specification covering page with my logo of the system.

A few years later, “Amazing Grace” / “Grandma COBOL” / Dr Grace Hopper came to visit Australia from 18th May to the 7th of June 1972, and gave presentations in several capital cities including Canberra.

Later that evening, Brian Withell, Tony Battanta, Peter Carmody and I (James (Jim) Duff) had the pleasure of having a very interesting chat, and a few beers, with Grace at the Canberra Rex Hotel. Dr Hopper served in the US Navy for forty-three years, rising to the high rank of rear admiral. She was - The First Woman to Program the First Computer in the United States.

Another 'online' system was created by me while I was in Adelaide, where I spent two years (1974 - 1975) to assist in the installation and operation of that State's computer system. In order to help the customer service officers, I generated a regular list from the Pensions Master file tapes that were then converted to microfiche. These were known as "Duff Data".

When I resigned from the Department in April 1978, Bill Lewisson, my superior officer, made the following comment about me in his farewell speech "There is an old saying that you can tell a pioneer by the arrows in his back".

So, in summary, I am proud to have been involved in the early days of creating Computing SOFTWARE in Australia, for the Commonwealth Department of Social Services, later named as the Commonwealth Department of Social Security – and also being one of the early Software creators around the world.

James D (Jim) Duff

May 2019

Attachments

(Page)

(6) Wangaratta1962

Front page of the Wangaratta daily newspaper, the Chronicle Despatch dated September 5, 1962

(7) SA99-OriginalDesignLayout - 1966

Copy of the typewritten cut-out form SA99, Pension Record Printout, used as the design and testing

document.

(8) SA99-Original Sample of Pension Record -1967

Sample of the original one-up Form SA99

(9) SA99-Second Version of Pension Record Printout

Sample of the following year's two-up form

(10) SA100-Sample

Sample of a Pension Advice - 1967

(11) Ian-1969-Inside

Photograph of Ian at work in 1969 - see his fountain pen on the computer printout

(11) Jim-1969-Inside

Wonder what I was thinking of with that look

(12) Jim-and-Ian-1969-Inside

A photograph of us together at work

(12) Jim-and-Ian-1969-Outside

A photograph of us in the Pitt Street plaza at the back of Australia Square building

(13) Mike Whitelaw who passed the fortnightly pensions to the customers banks etc.

Early Picture of part of the I.B.M

Two male workers at printers. Next picture at the right side of this one.

(14) Front page of a report book

Copy placed in the University of Melbourne Library – place name circle added at top right.

(15) JimDuff-and-IanMcDonald-at-Computer-1970

Photograph in the article in the April 1970 issue of Careers Honi Soit

(16) Jim Duff sitting in second picture. Other workers just outside the room.

(17) POLES Logo

Front page of Pensions On Line Enquiry System

(18) POLES Interaction Diagram

A diagram showing a broad chart of the program elements involved

(19, 20, 21) Pensions Run Diagrams

Diagram showing the flow of data between the several sequential programs

(22) Nice document about me – James D Duff. “To whom it may concern. 1 May 1978”

September 5, 1962

From the front page of the Wangaratta Chronicle Despatch daily newspaper.

An early public picture of me (James D Duff) in the Department of Social Services

These days in computers, you can view pictures and expand them to be read.

1966 Original Design of Form SA99 – Pensions Record Printout

This form was created using a small font typewriter, and was designed for continuous pre-printed forms. The printers could only output upper case characters at 10 characters per inch horizontally, and six lines per inch vertically. This working sample had rectangles cut out where the data would be printed, and was used as an overlay to check the positioning of the data on testing printouts. A real form with live data is shown in a later page.

September 1967

Live sample of an original Form SA99

Second version of the Form SA99 – Pensions Record Printout

This was a two-up version of the Record Printout, and utilised the full 130-character width of the upper-case printers.

As well as going through a machine to strip off the side edges containing the sprocket holes, it was also split down the centre.

Form SA100 – Pension Advice

This was another form that comprised a pre-printed layout, and was filled with selected data in the reporting program PEN11F. This form comprised a top layer of paper, plus a layer of carbon and a second paper layer that was kept as the office copy.

Ian McDonald, top, and Jim Duff, below, at work during the updates to pension rates and software accordingly in 1969 at the Sydney Office

Ian McDonald, right, and Jim Duff, left, at work during the updates to pension rates and software accordingly in 1969 at the Sydney Office

Ian McDonald, right, and Jim Duff, left, in the Pitt Street plaza, close to the Department’s Sydney office.

Another early member is Mike Whitelaw. He was in a small group who took the major software payments each fortnight to deliver them to the banks for the pensioners to have them received.

Early picture of part of the Sydney I.B.M. computer

The next picture is part of the right-hand side of this picture.

Another early picture of part of the computer area showing the reels.

This picture shows the name of the computer – IBM. As well as the names of us early Software providers, left to right, James D (Jim) Duff and Ian McDonald.

Grace Hopper visitor to us in Australia.

Jim Duff sitting in the second picture.

Pensions On Line Enquiry System – POLES

Design document front cover

Pensions On Line Enquiry System – POLES

Interaction Diagram

Pensions Run Diagram – Page 1 of 3

Pensions Run Diagram – Page 2 of 3

Pensions Run Diagram – Page 3 of 3

Nice document about me from my director, as I had left the Department of Social Services, later renamed as the Department of Social Security.