‘How it Works’ The Computer

This is a bit of fun really, it certainly isn’t a review of the book because any objective review would say that this book is no real use for understanding how computers work nowadays, but it is an insight into just how much technology has advanced since this book was written in 1971, so we are looking back fifty years. Before the days of computers in the home and decades before mobile phones and despite it being within my lifetime, as I bought this book new, it seems an unbelievably long time ago for technology. I was inspired to read it again after listening to Sir Tim Berners-Lee talking about the 32nd birthday of his invention of the World Wide Web which was celebrated last week. Actually this year (2021) marks the 30th anniversary of the Web being available to everyone rather than just the scientists at CERN which was where he was working at the time so it’s a good time to look back two decades before then to how computers started to be available to even a relatively small business although they were still wildly expensive.

This blog is going to be quite image intensive as I want to include several of the lovely illustrations by B H Robinson because they really tell the story to us nowadays far more than the informative but very technically dated text by David Carey. The first thing that strikes you is the sheer size of the equipment and then as you look further you realise that the two installations shown above don’t have any screens, the user interface is a teleprinter. The first ‘business’ computer I ever used didn’t have a screen either so I sympathise with the operator above, screens did exist but were quite scarce, certainly in the early 1970’s. It is also worth pointing out that the massive amount of cabinets lights and switches in the ‘large computer installation’ made up considerably less computing power than the mobile phone in my pocket.

Back then the sections of a computer were really obvious because they were separate huge cabinets or large pieces of equipment, nowadays everything is in one piece so it is actually easier to envisage how a computer works by looking at these old examples. To start with you need to get a programme and some data into the machine and that was a lot harder than it sounds. The example above uses a card punch followed by a card reader, yes the process was for data entry clerks to type everything into a machine that could produce thousands of pieces of cardboard each with a tiny part of the information and then carry the stack over to another machine (being very, very careful not to drop it because they have to be read in sequence) and then feed them in to load data into the computer.

Just how tedious this job was is shown above, and everything had to be verified because a single hole in the wrong place would make the entire stack useless until it was corrected. The amount of time taken to produce even the simplest programme or data source was unbelievable to those of us today. One way of getting round the danger of dropping several hundred cards, all of which look identical to a human, was to use paper tape instead, at least then it was just on a long reel although these also needed to be handled carefully as they could easily tear.

The main reason I have included the picture above is because it clearly shows a punched card. Each card consisted of eighty columns of numbers and each column could encode one letter or number, this blog entry consists of 7305 characters so would need a minimum of 92 cards to just hold the text; the pictures were not an option on machines like this back then, which I have to keep reminding myself is well within my lifetime as I was nine when this book was published. I say a minimum of 92 cards because I’m pretty certain words couldn’t wrap over cards so there would be blank space at the end of each card where a word wouldn’t fit.

To run a programme again you would need to reload the stack of cards and read them again unless you had a sufficiently large computer centre where you could have magnetic tape storage or even that modern wonder a disc storage device.

In the background you can see the five foot high magnetic tape cabinets, these were pretty quick in the day but nowadays the lag from a request for data and it actually arriving at the CPU (see later) makes them completely redundant, even more so than all the rest of the equipment shown. For example there would be an initial lag whilst the right part of the tape was found for the data needed and the book then explains that the tape could be read at up to 900 characters per second, now that figure is a little misleading as we are talking binary so just 0 or 1, to encode a letter or even a number other than 0 or 1 you need a lot more than one character, in fact you need 8 bits, otherwise known as a byte so reading this blog at 900 characters per second (111 letters per second) would take over a minute. Throughout the book Carey refers to storage in bits, presumably to make the numbers look big and impressive, even working in bytes is hopelessly outdated as we will see shortly.

Ah, disc storage, but just look at the size of the discs, 14 inches (35.5cm) across, six layers in a cumbersome disc pack, but at least you got lots of storage which was very quick to access. Quick yes but in today’s terms quite slow and 7.25 million characters per disc pack. Time for some maths again, lets work out just what that storage is in modern values.

  • 8 bits to the byte, 1024 bytes to the kilobyte, 1024 Kb to the megabyte. (Yes I know nowadays we just use 1000 for ease of calculation but in 1971 it was definitely 1024 as that is the relevant power of 2)
  • 7,250,000 bits = 906,250 bytes = 885 kilobytes = 0.864 megabytes
  • I would need eight complete disc packs to store one photograph taken by my phone, even assuming that such a thing was possible and I don’t even have a particularly up to date phone.

Actually that’s pretty good, the first computer I programmed for a company in the early 1980’s was an Osborne 1 which had 64 kilobytes of memory and two 90 kilobyte disc drives but I still managed to write a working insurance claims handling system for a parcel carrier on it.

Back in 1971 whilst there were computer chips, machines were still filled with transistors soldered onto printed circuit boards alongside the fairly limited integrated circuits available. The memory often however hadn’t moved on from the horribly delicate magnetic core store shown above. This isn’t an analogy as to what is happening you really did have lots of tiny ferrite cores strung onto wires which could be magnetised on or off to signify 1 or 0. As you can imagine the amount of memory was therefore pretty limited although the book claims that it could get up to a million bits (122 Kb).

Output back then would mainly be to a printer or possibly a screen in an advanced setup, I don’t think we need a picture of what they look like.

I’ve greatly enjoyed this trip down memory lane and when I showed the book to a seventeen year old friend she was astonished at the size of the machines and the limits they had. The computer she carries around with her all the time is millions of times more capable than the equipment featured in the book and tens of thousands times cheaper as well when inflation is taken into account.