March 2010 Archives

The one and only man who kept pace with a computer for almost 30 minutes

| No Comments
| More

One of the joys of researching the history of computing is tracking down people who worked in the industry in the early days that are happy to share their memories. This week I was lucky to talk to such a gentleman who led a team of "human-computers" in 1952 in the run-up to the development of the Harwell-WITCH computer now being restored at TNMOC.

The early 1950s was a pioneering time in the UK and the computing frontier towns of the time were Manchester, London, Cambridge and a village near Oxford. Of course, the Oxfordshire village of Harwell did happen to have the UK's atomic energy research establishment on its doorstep!

Computing was an altogether more personal experience then: just a small team of young people equipped with voluminous mathematical tables and hand-cranked calculators.

At Harwell they were working week in and week out on calculations to support the design of Britain's first atomic power stations - calculations that could take days, if not weeks, to complete.

That is until the electronic division completed its first electronic computer and presented it to the human-computer team. It wasn't the fastest machine, but it was relentless and regularly worked well into the night.

The team of mathematicians weren't convinced of course, and took a great deal of persuading to adopt the new machine. My contact had set himself in a race against the computer. He managed to keep pace for nearly 30 minutes - until his arm ached from turning the calculator crank handle!

Given a few prompts, our pioneer also remembered just what the first program he wrote for the machine did, and the fact that it worked first time.

Perhaps we can all remember our first time, or at least when it worked!

Finally, while some people are preparing for the London Marathon, at TNMOC we are now going through a rigorous selection procedure for volunteers who will attempt another race when the Harwell/WITCH computer is restored and running. Those with strong arms are eagerly sought.

On Digital Archaeology

| No Comments
| More

One of the key speakers at the Vintage Computing Festival we are holding in June is Christine Finn, author of "Artifacts: an archaeologist's year in Silicon Valley".

At times I do feel like a traditional archaeologist with all their challenges, delights and disappointments. We don't often get to select our digging sites, but often get called in to be emergency archaeologists to rescue computers and systems from an imminent wrecking ball or the local authority tip.

We often hear tales from visitors about machines that are no longer used and that will be lost if we don't do something about it soon. Sadly, that machine is often beyond salvaging, and frequently it is a system that we would have liked to have had. The lost machine then magically acquires an almost mythical status - it would have been rare, complete, fully operational and with all the original manuals and software!

This past weekend saw the museum rescue gang heading off to a small office in Aylesbury to collect an ICL System 25 'together with disks, printer and all the manuals'. We didn't know quite what to expect. When we arrived the reason for the miraculous survival of this machine was apparent - it was at the far end of an office block, up a steep and complicated staircase! It was to require all our best (and improving) excavation techniques.

The System 25 was a very popular machine in the UK for small companies and in particular for point-of-sale systems. It was an unusual computer in that it operated in decimal rather than binary. It was based on the US designed Singer System 10, that had originally been designed by Friden. Singer acquired Friden, and then in 1976, ICL acquired Singer. The machine we were rescuing was a later model System 25+ consisting of the processor cabinet, a disk cabinet containing two EDS80 80MB removable disk drives, and a 132 column line printer.

The pair of 80MB disk drives were dismantled and carefully man-handled down the stairs, the removable covers from the (actually quite small) processor cabinet we removed and despite its lack of wheels carefully got down to the ground floor.

Which left the line printer that needed to be moved! In the 1980s line printers were built to last, no flimsy plastic or aluminium covers, but heavy sheet steel which also covered a large transformer and a heavy motor. It was impossible to take it apart, so five of us struggled with it down the stairs - one step at a time, with multiple stops to catch our breath! One hour later it was outside as good as new.

The huge collection of manuals is particularly interesting. It allows us to reconstruct the use of computers in this company since the purchase of their first System 10 in September 1980. The site log reports the engineers installing their first system, subsequent upgrades to an ICL System 25, and finally the installation of their System 25+. Their last ICL machine seems to have been the Unix based DRS3000.

I can only imagine the look of dismay when ICL delivered the machine and were told it was to be installed upstairs, but this simple fact seems to have saved the system for TNMOC! I suppose archaeologists find exactly the same thing: the more accessible, moveable artifacts have long since disappeared.

Our next steps are to examine and document the hardware we rescued, plan the operation to make safe copies of the disks, and of course to continue the research into the back story of this very popular British computer. Once we have secure copies of the system and application software, we plan to construct a System 25 emulator that anyone can run, and of course the system will be on display at the museum in our Large Systems gallery.

A Golden Age of Computing?

| 1 Comment
| More

I was recently asked to consider what might have been the Golden Age of Computing, and I think I have two answers.

With my serious computer history hat on, I opted for the 1960s.

We started the decade with a chaotic scene of computer manufacturers offering multiple and mutually incompatible products. Machines were supplied as 'bare-metal' and the end user began the job of writing his (and it was almost always "his") own application software in the full knowledge it would all need to be re-written when the machine was upgraded. Programming computers was still very technical and difficult, and relied heavily on an in-depth understanding of the particular machine's design.

But then computing came of age. Hardware was transistorised, reliability increased, and standard programming languages were created. Many like FORTRAN, COBOL and BASIC are still with us today. Computers stopped being just boys' toys!

Women like 'Steve' Shirley took a lead -- although the fact that she used the name "Steve" reminds us that she realised life was easier if potential customers initially thought you were another chap!

Our industry, Electronic Data Processing, was all conceived during the decade of flower power and free love.

By the end of the decade, computing had been transformed: applications could be bought off the shelf, mass-produced mini-computers were beginning to appear and computer purchasers could expect an immediate benefit.

A more personal answer is that I think we might each have our own golden age of computing, one that has a common theme. It was probably when we were in our mid-twenties, no longer a novice in the industry, but not so far up the corporate ladder that serious customers shouted at us and we didn't have to break bad news to the board! We had new trainees to boss around and we knew everything about the system - we were masters of the universe project!

So, taking my computer history hat off, long live the early 1980s! Just booting CP/M is still a thrill - so the Personal Computing Gallery at TNMOC is therefore one of my favourite haunts and a secret pleasure!

Subscribe to blog feed

Categories

 

-- Advertisement --

 

About Archives

This page contains links to all the archived content.

Find recent content on the main index.

 

-- Advertisement --