Australian Computing History

[ChrisGreaves]

https://ia.acs.org.au/article/2017/acs- ... er-42.html
Scroll to the bottom of that page for a list of published chapters.
Cheers
Chris

[John Gray]

Are you referring (extremely obliquely, I might add) to the "Related Article"
Microsoft will mine bitcoin with your brain?
Otherwise you are going to have to give us a clue about which of the possible 41 chapters you find fascinating / stupid / interesting / worth posting about / ...

[GeoffW]

I was interested in the section about University of Melbourne. I didn't realise that they only started offering computing courses to first year students in 1970 - I started studying in my first year in 1973. Not a pioneer, but I didn't realise I was so close behind.

However, I did start programming at high school in 1970. We had cards where you had to punch perforated holes with a paper clip; the programs were sent by snail mail to a different university in Melbourne to be compiled and run. It could take a few months to get a non trivial program up and running.

[ChrisGreaves]

Otherwise you are going to have to give us a clue about which of the possible 41 chapters you find fascinating / stupid / interesting / worth posting about / ...

I can indeed provide a clue about which of the possible 41 chapters I find fascinating / stupid / interesting / worth posting about, but that might be of small value to specific individuals of this Lounge. After all, who else here held IBM 1620 FORTRAN II punched cards up the the ceiling lights in the Physics Building of the University of Western Australia in the quiet hours of 2am-6am for six months? Who, for that matter, had found invented a need to input 81 characters of data on an 80-column punched card via an IBM keypunch console? And who spent a sleepless night after having destroyed a $250,000⁽²⁾ DEC PDP-6 computer⁽¹⁾, wondering whether he would be required to pay back the replacement cost before, or after, his $87 account at the Uni WA bookshop?

There are enough Australians (and Eastern-Staters) in the Lounge to re-discover part of their heritage amongst the pages, whatever their computing heritage.
Me?
I'm looking forward to all the references to the UK's ICL 1900 series and GEORGE3!

Cheers
Chris
P.S. GeoffW: "I was interested in the section about University of Melbourne. ". See?!!???

⁽¹⁾ Mike Patterson⁽³⁾ told me the next day that it just needed re-booting, a new term demonstrated to me by configuring 36 small binary switches and pressing a green button.
⁽²⁾ All figures given in the then revolutionary $AUS
⁽³⁾ No relation to Les, AFAIK
⁽⁴⁾ And yes, the footnotes are numbered in the sequence of their creation rather than their spatial position in the response.
⁽⁵⁾ And yes, I have used footnotes as postscripts in at least two instances that I can discover.
C

[ChrisGreaves]

didn't realise that they only started offering computing courses to first year students in 1970 - I started studying in my first year in 1973. Not a pioneer, but I didn't realise I was so close behind.

Same boat. When I was introduced to the IBM 1620 on the second floor, i assumed that it had been there forever. Turned out it was bought just before I was allowed to spend hours developing a determinant-calculation program (I wasn't sufficiently mature to complete the matrix-inversion, being somewhat frightened of the moving-target aspect of floating-point arithmetic).

... the programs were sent by snail mail to a different university in Melbourne to be compiled and run. It could take a few months to get a non trivial program up and running.

Same here, but our coding-forms and subsequent job decks went by once-a-day on some days by courier from Wollongong (The Illawarra County Council Building) to AIS in Port Kembla. Turnaround was anything from 24 hours to two weeks (when the AIS payroll program took priority over us programmers-in-training).
I don't know about a few months, though. I was a programmer-in-training (IBM 360/30 Assembler, FTN, COBOL) for nine months, after which I was sent to Newcastle which had no IBM/360/30 so we were trained in 1401 Autocoder (with payroll delays) until the twin CDC-3300s were rolled into the new building, after which I mastered COMPASS then went to work for ICL(UK) in Adelaide, just filling in time until Eileen's Lounge was up and running.

A "different university in Melbourne"? I am impressed. We had to make do with only one University for a million square miles plus several parts of Asia!

Cheers
Chris

[StuartR]

Gosh, I worked for DEC, and I never came across a PDP-6

I just looked it up and this was a precursor to the DEC-10, which I did know.

[ChrisGreaves]

Otherwise you are going to have to give us a clue about which of the possible 41 chapters you find fascinating ...

OK John; here you go:-
"with the first actual sale of a PDP-6 to the University of Western Australia, in May 1964."
This PDP-6 is, in fact, the one I "destroyed" in the early hours of one morning around October 1967.⁽¹⁾

Also note, please, "In August 1966, the Australian Financial Review reported on Australia’s first ‘long distance man-computer conversations’, with the University of WA’s PDP-6 communicating across telephone lines with teleprinters in DEC’s offices in Sydney and Maynard."

Note too please that in this page there is a photo of a computer being untransported by plane (remember Ansett ANA?) whereas in Chapter 16 again computers were now traveling by car (" it was rushed by car from Sydney to Canberra to star at the Australian Computer Conference being held on the day of its arrival.")

Cheers
Chris
⁽¹⁾ "The October Crisis" in "Reminiscences" by Chris P.R.⁽²⁾ Greaves, vol XXXVIII chapter 42.

⁽²⁾ NOT "Public relations"

⁽³⁾ I don't have a 3rd footnote yet; this is just a placeholder.

[John Gray]

It took quite some effort to produce this punched card, and it was a very long time ago...

John 500x217.png

[ChrisGreaves]

It took quite some effort to produce this punched card, and it was a very long time ago...

I'd be much more interested if, instread of pure binary code (from its appearance in the image) it were in Gray code, or Balanced Gray Code (or even Unbalanced Gray code, whatever that looks like)
Cheers
Chris

[BobH]

My first encounter with a computer was in June of 1965. It was an IBM 1401 with CPU, 4 tape drives, memory unit, printer and card reader/punch and lots of cables among 'em. This was before IC's were in use. The memory unit contained 4k - yes kilobytes - of memory in a box about 1 cubic yard in dimension. I first learned to operate it, including how to bootstrap it, Chris. I was then introduced to Autocoder programming, then BAL (Basic Assembler, or was it Assembly) Language, then Fortran, then BASIC, then PL/1, then COBOL. The 1401 was replaced with a 360/30 before I changed jobs and became a full-fledged programmer analyst, then systems analyst, then manager then what we now call an IT director and retired as CIO.

Most of my experience was with IBM equipment and operating systems though I did have the occasional contact with others including DEC, Tandem, and Unisys.

I was operating/programming the bank's computer at night and going to university (uni for you so inclined) during the day. I had a 30 mile one-way commute which took an hour or more out of my already very short day; so I thought I'd try to take a summer course at the local uni (NC State) in PL/1. When I applied I learned that there would be no course because the intended instructor had reneged. In talking with the assistant dean he learned that I already had experience programming (ca. 1968) he offered to hire me to teach the course as it would be very basic. I demurred and continued my commutes.

I took the PL/1 course as a requirement in my business degree at UNC. We had a room with only 6 keypunch machines for the entire student body in all disciplines. It was nearly impossible to get use of one and there was a 30 minute limit. Being the computer manager at work, I had a staff of 8 keypunch operators who punched and verified all my decks. Other students were green with envy and accused me of cheating to the professor. We didn't have a PL/1 compiler on our computer; so I still had to submit them over a counter like everyone else, but I was always able to finish an assignment faster because I had verified card decks. (I still have a box of 5081 stock bearing the UNC logo somewhere.)

[GeoffW]

I am but a mere youngster.

[PJ_in_FL]

I was interested in the section about University of Melbourne. I didn't realise that they only started offering computing courses to first year students in 1970 - I started studying in my first year in 1973. Not a pioneer, but I didn't realise I was so close behind.

However, I did start programming at high school in 1970. We had cards where you had to punch perforated holes with a paper clip; the programs were sent by snail mail to a different university in Melbourne to be compiled and run. It could take a few months to get a non trivial program up and running.

In 1970's my school had electric adding machines that could multiply, if you didn't mind waiting while they pseudo-multiplied by adding the number by the multiple number of times.

I guess we were a little behind the times.

[ChrisGreaves]

I am but a mere youngster.

True, Geoff!
You are as young as I was a mere seven years ago.
Cheers
Chris

[ChrisGreaves]

In 1970's my school had electric adding machines that could multiply, if you didn't mind waiting while they pseudo-multiplied by adding the number by the multiple number of times.

Friden?
Cheers
Chris

[ChrisGreaves]

… contained 4k - yes kilobytes - of memory in a box about 1 cubic yard in dimension.

the IBM 1620 too was a 4K machine with a cubic yard of 16K which took it to 20KB. FWIW I still have my 1983 Radio Shack MC10⁽¹⁾with 4KB and a 16KB plug-in, about the size of a packet of cigarettes, but only a 20-pack, not the newer 25s. Cigarettes I mean, not KB.

…introduced to Autocoder programming, then BAL (Basic Assembler, or was it Assembly) Language, then Fortran, then BASIC, then PL/1, then COBOL.

I am corrigible.

20200907_092212.jpg

Mainframes had machine language which only the System Engineers and weird guys like me and Quentin van Abbe understood. The Set Word Mark instruction was shown as a comma in human literature, but was toggled in (or loaded from paper tape (UK) or punched cards (USA)) as a 1011011, usually written as CA821, where the “c” was the check bit and zeroes were elided. Bootstrap loaders were written in machine language.
The 1401 bootstrap process required the operator (remember them?) to press the Load button on the card reader, which action caused the first card of the deck to be read into the input buffer at location 001 and control was then transferred to location 1. The first seven characters of that first card contained “,008015”, which set word marks at locations 8 and 15, so that the instruction coded at position 8 on the card would be word-marked, and was itself a Set Word Mark instruction, as was the next instruction at card position 15 (and hence memory location 15).
So this first card proceeded like the track-laying machine it was.

The second card held useful instructions rather than track-laying code. These instructions had enough sense to load the next eight(?) cards which were a much stronger program that could load instructions from the following punched cards. So the object deck was a two-card program which loaded an eight-card program which loaded the rest of the object deck (output from Autocoder or FORTRAN).
After that things got less complicated. IBM’s earliest object deck format was one instruction per card, so a 2,000 card source program generated an object deck of about 2,000 cards. The next generation bootstrap loader was capable of loading several instructions from one card, so the 2,000 source card program might be only 700 object deck cards.
In a sense even the bootstrap loading process was bootstrapped to a higher-level capability.

But I digress.

QVA and I were bored, so we set out to develop a one-card bootstrap process, and after two weeks work succeeded. This caused the operators grief, because every one of our programs paused for eleven seconds after they pressed the Load button because we had to clear storage one character at a time (MCS) rather than taking advantage of the CS instruction.
But you probably guessed that already.

20200907_135141.jpg

This example SPS code (note the 4-decimal digit addressing) appears to use symbolic/mnemonic labels for program branches and formatting strings, but not for data locations. Perhaps the programmer was "old school" and didn't trust these new-fangled user-defined data labels? Maybe SPS-1 didn't support data labels?

SPS-1/2 were true symbolic languages in that they allowed coders to use human-mnemonics (, for SetWordMark, / for ClearStorage, A for Add etc) and symbolic labels (you will remember that strange 3 character scheme for encoding 5-digit memory addresses), but Autocoder added more powerful facilities, true macros for one thing (let a tear grace your eye as you think of the IOCS macros), and so I see even the assembly process as bootstrapping itself out of existence, because with Autocoder one could write a pretty good FORTRAN compiler, and of course armed with FORTRAN one could implement Poole & Waite’s STAGE2 macro processor, which we did at ICL-SDC and then invented all sorts of languages and spewed out compilers directly from a Backus-Naur notation. One per day.

But I get ahead of myself.

BAL was the 360 equivalent of SPS, I think, for the 360.
“Assembler” language was powerful enough to support true macro coding. Apart from the 3rd-generation compilers there was nothing more powerful (in terms of lines/day of coding) than Autocoder on the 1401.
The Honeywell 200 supported EasyCoder, but Honeywell was trying to snatch business out of the hands of IBM.
I have put the Daniel D. McCracken and the photocopied payroll program on your bedside table; you can wallow in it the day you arrive.

The 1401 was replaced with a 360/30 before I changed jobs and became a full-fledged programmer analyst, …

That must have happened to all of us. I seemed to be into a new language every six months. Certainly I changed jobs/employers every two years or so. There as always a new machine on the horizon, until I was at ICL where we developed machines built from software. It was at ICL-SDC that Dr Bob Northcote issued an edict “No more writing in assembly-level language”, so we had to bone up on Plasyd which initially looks like assembly language with structured statements; a stepping-stone between 2nd-generation (assembly language) and 3rd-generation structured languages like ALGOL.

I was operating/programming the bank's computer at night and going to university (uni for you so inclined) during the day. I had a 30 mile one-way commute which took an hour or more out of my already very short day;

Brings tears to my eyes. I had stamina in those days. Up at 6, then off for a 90-minute 30-mile bike ride to Midland Junction and back, breakfast, lectures all morning, then in mid-afternoon another bike ride (Nedlands-Perth-Fremantle clockwise or anti-clockwise along Stirling and Canning highways), study, dinner, study and then bed. Across to the Physics building 2am-6am about four nights a week. What was I thinking of? At least we know why I got straight-Cs for my fourth-year mathematics units.
“Uni” because long-distance telephone from W.A. to anywhere was expensive and elision of syllables saved money back in those days.
(This link is for Hans because he won't like the previous link)

… to hire me to teach the course as it would be very basic. I demurred and continued my commutes.

I was awarded a “Dip.Comp & Num Anal” partly on the strength of my having delivered some of the lectures. They were strange days back then.

I still have a box of 5081 stock bearing the UNC logo somewhere

I would ask “Bob, why are you clinging so desperately to the past?” were I not doing so myself!
I see that the rain has stopped, so I am going outside to play in the yard …
Cheers
Chris
⁽¹⁾This link does not show an MC-10, more's the pity.
C

[BobH]

Great post, Chris!

Brings back many memories . . . " Ah the smell of a fresh coding pad in the morning."

[ChrisGreaves]

" Ah the smell of a fresh coding pad in the morning."

"Coding pads? Ee lad, we 'ad ter riot wi t'ketchup bottles on backs o' pizza trays ...."
Cheers, and I'm glad you liked it!
("Liked it? We 'ad ter lick plates clean afore our Nan would mop up t''Sauce wi' ...")
Chris

[HansV]

Again?

[ChrisGreaves]

Again?

You can never have to much of good entertainment.
Entertainment? I remember when ...
Cheers
Chris

[BobH]

Ketchup bottles and pizza trays? We had to prick our index fingers and write on the palms of our hands!

(Sorry - I cannot do a Yorkshire accent.)