One of the aims of Pararchive is to bring together community groups, members of the public and cultural institutions to collaboratively engage with our shared cultural and historical heritage in order to ensure its ongoing relevance. This partnership is enabling project members to interact with the diverse assets and collections of our project institutional partners with a view to adding value to such artefacts while at the same time promoting community-led research. An illustrative example of how we practise what we preach is Chris – a member of the Arduino Manchester community group who is interested in researching the history of physical computing in Manchester. Courtesy of the Museum of Science and Industry (MOSI), Manchester which is part of the Science Museum Group and with the kind assistance of Sarah – Assistant Curator at MOSI, Chris visited the physical computing collection and shares his experience below:   

Manchester Small-Scale Experimental Machine (Nickname: “Baby”)

Chris at MOSI-25.03.2014So you may have heard of this one – the world’s first stored program computer. At MOSI you can find a very faithful, working replica in the main foyer. I was able to speak to the volunteers that day and get a little more detailed information about the computer. It was built at the University of Manchester in 1948 and mostly based on the transistor technology of the time, thermionic valves. As a result of these valves and the other power supplies within the machine, it uses an enormous 3-4KW of power.

The parts to build the replica came from a mix of the garages of hobbyists, and government surplus parts that could be easily bought. As an example, the keyboard (beneath the main monitor) is made up of 5 x 8-button panels that were also used in spitfires – a clear indication of the materials that were easily available in 1948.

Chris at MOSI-25.03.2014-2

This machine has a number of important innovations built in. One was its modular design, which allowed the engineers to replace and maintain parts more easily – an essential bit of good practice as computing machines became more complex. Another innovation (that I found particularly surprising) was the way the memory worked. The CRT displays, for the most part, are the memory store of the computer. There are four displays in use: one used to store the program; and two store intermediate states as the program runs. The fourth is a monitor that can be switched to display the state of the other three. I was lucky enough to get a diagram of the modules in use.

Chris at MOSI 25.03.2014-3There is no long term storage solution built in – meaning that the machine would lose everything and need to be reprogrammed if it was switched off. A feature of the replica is an interface with a modern computer, which allows it to be programmed quickly and easily, without having to resort to entering machine code at the keyboard. The program running at the time of my visit was a simple scrolling animation of the word “BABY”.

The Hartree Differential Analyser

This computer is more of a single purpose machine, albeit a very charming one. I’m told that the first prototype of this machine was made with meccano, and to test its prowess was used to solve the wave equation for a hydrogen atom. From this successful demonstration the machine I saw was funded.
Chris at MOSI-25.03.2014-4
This machine is capable of doing calculus! A complex arrangement of rods, gears, differentials and torque amplifiers allow this machine to integrate or differentiate an input function, and draw the output onto paper. I was able to see it in use – the volunteer on hand first used it to draw a sine wave (I have a paper copy but can’t seem to find it), then after some “reprogramming”, which involved physically altering the mechanism.

Chris at MOSI-25.03.2014-5

He was able to demonstrate drawing a circle as reflected below. You may notice that the start and end points don’t quite line up – I was told that drawing a circle was a common ‘hello world’ program for differential analysers – the amount by which the start and end points differ is a quick, visible measure of the accuracy of the machine. Missing or broken parts mean that this machine isn’t quite as accurate as it once was.Chris at MOSI-25.03.2014-6 

By the sound of Chris’ account, it seems he enjoyed his visit. If you too would like to learn more about the “Baby” – the world’s first stored program computer, and about the story of early computing in Manchester and also wish to see a Pegasus computer which is a little later than the Baby computer in the Computing Gallery which is located on the 1st floor of the Electricity Gallery in the 1830 Warehouse, feel free to get in touch with Sarah to arrange an appropriate time.