How things used to be

I can still remember the first time I used a computer. It was in 1974.

As far as modern computing is concerned, that was prehistory! So it's appropriate that the computers of those days were rather like dinosaurs — huge and slow. They were known as mainframe computers and lived in special air-conditioned clean rooms with underfloor cooling and a suction mat in front of the door to remove every speck of dust. They were so expensive that they had to run 24/7 to justify their cost, so machine operators had to work shifts around the clock.

The computer itself, manufactured by ICL, was the size of a large filing cabinet. This contained the central processing unit and core memory, all of 96 kbytes! Other cabinets stood around it, containing magnetic tape and disk drives for data storage. Disks were quite small in those days so most of the data was on tape and only accessible by means of overnight batch jobs.

For data input, there would be a punched card reader and a paper tape reader. The line printer was usually housed in a side room because it was so noisy. As we were a scientific establishment, there was also a graph plotter.

We users never actually saw the computer. All you saw was a terminal plugged into an ethernet socket on the wall. The first terminals were converted teletype machines and my first input into a computer was carried out on one of these. Those of my readers who use Linux may perhaps wonder why a text console is always called "TTY". Well, it's short for "teletype".

Later the teletypes were replaced by Decwriters, which printed 132-column lines on green and white fanfold paper with rows of sprocket holes down the sides. Then came VDUs — Visual Display Units — which were boxes with a screen on the front and a keyboard jutting out of the bottom. Detachable keyboards came much later.

The library's two inputters, to their annoyance, were not allowed to use online terminals of any kind because they were officially classed as typists and not as machine operators. So we bought them a couple of Texas Instruments magnetic tape typewriters. These produced written output on a paper roll (which I used for daily proofreading) and a binary copy on tape. They could also be used as terminals but technically they were typewriters and therefore usable by typists.

There was of course no ethernet socket in the typists' room. However there was one in my office a few doors down the corridor. So a loop of cable was run out of my office window and in at theirs. Every Friday I would hook up one of the tape typewriters to this cable and run the magnetic tapes that had been recorded during the week through to the mainframe. We had to keep up this subterfuge for a year before the grades of typist and punched card operator were finally merged and our typists got their VDUs.

Everything was plain text in those days. There were no mice and no graphics. You typed commands into a shell. The computer would obey the command and then prompt you for the next one. By this time the old ICL mainframe had long been laid to rest and we were all using DEC Vax minicomputers (much like a mainframe but with only one cabinet and they didn't require a clean room). They ran an operating system called VMS and the shell that came with it was pretty sophisticated, allowing me to write a variety of scripts that made life easier for the end-users.

I can still remember vividly my first experience of a PC. The DOS shell struck me as absurdly primitive compared with the powerful mainframe shells I was used to. And the Windows 3 graphical interface presented as an alternative was totally unintuitive. How on earth was one supposed to remember what all those icons stood for? It was like replacing an easy-to-read alphabetic script with a set of Chinese ideographs. And as for trying to control the wildly darting cursor with that ridiculous mouse...

I was pretty sure that it would never catch on. How wrong can you be? That was then; this is now. Home