The first computer I programmed

Created 1 December 2003

The chap on the left is Chris Hall who I haven't seen since 1976. The guy on the right is Peter Tilsley who (together with a very few others) is the reason I am such a proud geek and writing this today. If either of you are reading this, an email would be very welcome.

Pictured the Elliott 803 computer. It's a solid state computer with lots and lots of germanium in it. It's 39 bit (!) with a serial floating point unit and 8k 39 bit words of magnetic core. The boot ROM was actually a bunch of diodes taking up the first 4 words of memory. Actually, I think it was 3 1/2 words, but that's another story. This was where they'd written the instruction set to make the boot loader as tiny as possible.

I first programmed it in 1976 in Algol 60, when I was aged 11 at Loughborough Grammar School in England. I'd write my program down on paper first, then line up to type it into a five hole paper tape teleprinter. I'd then take it over the the Paper Tape Station (or PTS), first loading up the compiler from paper tape, and then committing my handiwork to the compiler. If I was lucky, there'd be no syntax errors.

If there were any errors, it was often a case of re-keying it, or if you were lucky, you'd find a free teleprinter (far left in picture) with a reader attached so you could manually edit the tape onto a new tape. The smart ones amongst us new how to read the tape directly and it was often possible to fix a syntax error by either punching new holes with a little device, or covering up existing ones with 'chad' or 'anti holes' or 'holes' using sticky tape. There was an ongoing argument as to whether the bits that came out of the paper to make the holes should be called 'holes' or 'anti holes'.

Once you'd used sticky tape on the paper tape, it was always luck of the draw if the tape would now go through the PTS reader without snagging. If it did snag, the speed of the paper tape would inevitably mean it would often rip as it abruptly stopped, leading to more tape repairs.

Your output would come out in the form of five hole paper tape which you then fed into a Creed teleprinter reader to show your answers.

With luck on your side, you could have the first ten numbers of the Fibonacci sequence in a couple of weeks.

The real nerds got their hands on the Solartron analogue computer (back left in photo). This was quite neat, although to be honest you really needed to be into calculus to use it. You programmed it with leads and cables, not keyboards. You could do a pretty mean moon landing game on it. I was finally up to joining that uber nerd gang at the age of about fifteen or so.

Almost all of the maintenance was done by the kids, aged between 11 and 18. This included getting the oscilloscope and soldering iron out. There was no real teacher involvement. We were left to our own devices. Until the day I got caught in the computer room when I should have  been in a gym class. What they didn't realise was that I hadn't seen a bat and a ball for at least decade.

All of us are now working in the computer industry, and most of us are also Radio Amateurs. I find it ironic that many of the guys who took the micky out of the nerds at school have ended up earning a crust in the industry themselves.

Mail Howard, G6LVB