The Editor Grovels

The first issue of Resurrection for which I have been responsible, dropped onto the doormats of CCS members early in August. Or, at least, it dropped onto the doormats of most members in August. Sadly, something appears to have gone wrong with the distribution, with some members receiving two copies and others receiving none at all! It is doubly embarrassing to report that our gallant chairman’s mat remained unblemished by copies of Resurrection, as indeed, did my own.

So my first task is to apologise to all those members whose copies went astray and to ask anybody who has yet to receive Resurrection 43 to get in touch with Kevin Murrell, our long suffering secretary - email, letter, carrier pigeon - your choice. Contact details are on page 31.

In Resurrection 44, we profile Donald Davies, the man who brought packet switching to the world and so opened up the possibility of the Internet. It is difficult to exaggerate the importance of innovation on this scale, yet the work was done, not in some vast telecommunications corporation, but at the National Physical Laboratory in Teddington, a mere stone’s throw from Resurrection’s busy editorial office.

Our feature article, in this edition, is our chairman’s erudite account of the development of Titan, also known as the Ferranti Atlas 2. Here is Titan on a moderately busy day -

News Round-Up

The financial plight of Bletchley Park has attracted the attention of the heavyweight press. In July The Times published a letter from some 97 distinguished academics, pointing out the significance of Bletchley Park. Not to be outdone The Independent initiated a campaign of its own the following month, publishing no less that five articles in one day, including an editorial, in support of Bletchley Park. Both papers have published follow up letters and articles.

As a result of all this publicity, visitor numbers have shot up, with an increase of 30% from 2007 and an extraordinary attendance of over 3,000 during the August Public Holiday.

Meanwhile, the e-petition mentioned in Resurrection 43 (petitions.pm.gov.uk/BletchleyPark) has risen to fifth place in the Downing Street list of 5,000 e-petitions, with no fewer than 15,533 signatures as Resurrection went to press. This is important because the top five petitions are displayed on the system front page, thus, with luck, attracting yet more signatures.

Finally, welcome light relief in the form of a BBC Radio 4 comedy series, ‘Hut 33’, which was recently repeated. Starring Robert Bathurst, it featured the unlikely adventures of a group of wartime codebreakers characterised as upper class twits. Well, it made me laugh! No doubt, it will be repeated again on BBC 7 in due course.

101010101

Donations to the National Museum of Computing have risen sharply recently with several substantial private donations and a magnificent $100,000 donation from cryptography company PGP and from IBM who, in September, led a kick-start initiative to encourage further donations from the technical community across the globe. See www.pgp.com/stationx.

An Elliot 905 has been acquired, and has been installed at Bletchley Park, near the Elliot 803B.

101010101

We regret to report the passing, in August, of Sir Edwin Nixon, Chief Executive and then Chairman of IBM UK from 1965 to 1990.

101010101

The Science Museum in London has put the first version of ERNIE, the machine built in the 1950s to select winning Premium Bond numbers, on display in the museum in Kensington (www.sciencemuseum.org.uk/visitmuseum/galleries/ernie.aspx). Retired in 1972, ERNIE 1 has since been in store at Wroughton. Like Colossus, ERNIE 1 was designed by the late Tommy Flowers and built at Dollis Hill.

101010101

News from our American cousins. The IT History Society (http://ithistory.org) has recently achieved a membership of more than 400. The IT History Society, formerly known as the Charles Babbage Foundation, was created with the goal of “enhancing and expanding works concerning the history of Information Technology, and demonstrating the value of IT history to the understanding and improvement of our world”.

101010101

The Heinz Nixdorf MuseumsForum (http://en.hnf.de/default.asp) claims to be the largest computer museum in the world. Located in the former administration centre of Nixdorf Computer AG, in Paderborn, northern Germany, it boasts some 5,000 objects and an exhibition area of 18,000 square metres.

Society Activity

Pegasus Working Party
Len Hewitt and Peter Holland

Pegasus continues to run well after its long rest. We have had minor problems with the alternator and power run up, but in general, Pegasus is working well. Following the completion of the special Babbage Engine, the temporary screens have been removed (after five years) and ERNIE 1 is now on display (see above). This has opened up the area and generated a lot more interest on “In Steam” days, which is always welcome.

We have still not been able to find a Creed engineer who could handle our Type 54 teleprinters. If there are any out there please get in touch.

Our “In Steam” days are every other Wednesday from 5th November, switched on from 11:00 to 15:00.

Contact Len Hewitt at .

Bombe Rebuild Project
John Harper

In previous reports, I have mentioned our Checking Machine. Although this is a fairly simple device, we increasingly realise how important this was in successfully finding original Enigma settings. As many will have read previously, we rebuilt one of these machines to go alongside our Bombe. Electrically, it is very similar to a German Enigma machine in that, when one presses a key, a different light is illuminated, the current having passed through three drums, a fixed reflector and back through the drums. There is no need for a stecker board (see Resurrection issue 6) in the way that it is used because it works on the drum core settings. It does not, however, have any drum progression mechanism. This is done by hand (and has to be) for the intended process.

The machine is used to check whether the results a Bombe produces when it stops truly represent those expected from the menu. For example, one of our test jobs comes up with four stops, only one of which is correct. The incorrect ones are due to the fact that the menu is not fully comprehensive. There is a trade-off between making a menu larger (and so losing overall machine capacity) and encountering a number of incorrect stops. As the Bombe can be restarted whilst a stop is being checked on the Checking Machine, little overall time is lost.

The way in which the machine is used is that all the positions on the menu are checked, one at a time. At each step, the letter illuminated is noted and after the drums have been moved to the next position on the menu this letter is input, the result noted and so on, until a complete loop has been made around the menu. If the last letter lit is the same as that input at the beginning of the operation, then this is likely to be a good stop. If it isn’t, then this stop is of no use and another has to be found.

We now have a three wheel German Enigma machine on display in our area which (almost) completes our full set of hardware used to break Enigma ciphers at Bletchley Park during WWII.

I say “almost”, because, as reported previously, we have now identified the special Typex used in the Bletchley Park machine room. This was used by the lady and others of whom I spoke in my Spring 2008 report. It used a method called ‘clonking’ to find the rest of the settings used by the German operator after a good stop had been verified on the Checking Machine. This machine appears to be based on Typex subassemblies, but not in any form previously identified. The base casting is lower, as if the bottom two inches have been removed, and no cover is fitted at the front. The keyboard is virtually the same as a standard Typex, except that the mechanism that inhibits more than one key being depressed has been removed. The ‘scrambler’ unit appears to be completely standard. We would like to reproduce this machine. So if any reader knows of a source of Typex sub-assemblies, we would be very pleased to hear from you.

Recently Their Royal Highnesses the Prince of Wales and the Duchess of Cornwall visited Bletchley Park. We were one of the selected stopping points and they spent about 10 minutes with us. By prior arrangement, Prince Charles started our machine and appeared to be very pleased with what he saw working. In fact, he looked much more closely at our efforts than he had planned. All in all, the visit to our Rebuild and to the whole of Bletchley Park was considered a great success.

Sponsor Sought for Bombe Rebuild Enhancement

The Bombe Rebuild has now been complete for over a year. During this time, numerous WWII jobs have been re-run successfully. However, we have not been able to attempt some of the decrypts in the Bletchley Park Archive. These require the full set of eight drums rather than the normal five that mirrored the wheels used in German army and air force three wheel Enigmas. We would very much like to enhance our Bombe with the three extra drum types. We have a substantial proportion of the necessary components left over from the original production, but there are a few unique parts still needed. We estimate that we can make these for £2000. We are therefore seeking a sponsor who could be identified with this valuable enhancement.

Our website is still at www.jharper.demon.co.uk/bombe1.htm.

Machine Emulation
Dave Holdsworth

Dave is starting to undertake a review of the collection of machine emulators and other software which the Society holds at sw.ccs.bcs.org. He hopes to report on its origins and will, in particular, examine some of the older material that we hold.

Our Computer Heritage Pilot Project
Simon Lavington

Some progress has been made with listing all the English Electric DEUCE and KDF9 computers to have been built. Since no manufacturer′s definitive delivery documents have come to light, compiling this information relies on the discovery of subsidiary source documents and upon the memories of those who worked on these machines. If any Resurrection reader is familiar with either of these English Electric computers, please take a look at: www.ourcomputerheritage.org/wp and follow the links to: ′N1X1: List of English Electric DEUCE deliveries′ and ′N4X1: List of English Electric KDF9 deliveries′ to see whether our story accords with your memories. If you have any comments or, better still, original documents that could improve upon the accuracy of these two lists, then please let me know at .

There are areas of the Pilot Study where we still need help. Particular computers that lack active volunteers willing to compile technical information include: the Ferranti Mark 1 and Mark 1 Star; the English Electric DEUCE, KDN2, KDF7, KDF6 and KDP10 computers; the BTM HEC and 1200 computers. If any reader is familiar with one of these machines and is able to devote time to compiling technical information for eventual uploading to the Our Computer Heritage website, then we′d be keen to hear from you.

Pioneer Profiles - Donald Davies

Donald Davies was one of the most outstanding British computer pioneers to have led the field in the post-war decades. Davies, who spent most of his career at the National Physical Laboratory (NPL), was a multi-talented engineer, administrator, and policy-maker. All of his talents combined when he invented packet switching in the 1960s, today a foundation technology of the Internet.

Donald Watts Davies was born, in modest circumstances, in Treorchy in the Rhondda Valley. Following the death of his father, the family moved to his mother′s birth town of Portsmouth. Davies entered Imperial College at the age of 19, graduating with a first class honours degree in physics in 1943. He spent the remainder of the war years on the atomic weapons ″tube alloys″ project at Birmingham University, as an assistant to Klaus Fuchs. At the end of the war, Davies returned to his scientific studies at Imperial College, obtaining a first class honours degree in mathematics in 1947. He also won the Lubbock Memorial Prize as the outstanding mathematician of his year.

In 1947, Davies obtained a position at the NPL, where Alan Turing was designing the ACE computer. The ACE project was over-ambitious and it foundered for two years due to bureaucratic obstacles, eventually prompting Turing’s departure. Davies emerged as probably the only person at the NPL with the right blend of electronic, mathematical, and administrative capabilities to get the machine built. Instead of Turing′s ambitious ACE project, he settled for the Pilot ACE, a small experimental model. The machine first worked in May 1950. It proved sufficiently powerful that a commercial spin-off, the DEUCE, was manufactured by English Electric and became one of the best selling British computers of the 1950s. Although Davies had taken on these heavy responsibilities (he was not yet 30), he still found time to show his inventive and playful side. For example, he built a noughts-and-crosses playing machine which was demonstrated at a Royal Society soirée in 1949.

Demonstrating his noughts and crosses machine at the Royal Society in May 1949

After the construction of the Pilot ACE, Davies became involved in computer applications such as a road traffic simulation and machine translation. In 1963, he became technical manager of the Advanced Computer Techniques Project, a government initiative to keep the British computer industry at the forefront of research. In 1965, he was seconded to the new Ministry of Technology in Harold Wilson’s Labour Government which, famously, had been elected on the lure of its “white heat of technology” manifesto.

In 1966, Davies returned to the NPL to become Superintendent of its computing activity, which had lost its way during the previous decade. Renamed the Division of Computer Science, Davies reinvigorated computer research and gave it a more practical focus. It was in this context that the work on data communications and packet switching was done. Davies had become interested in data communications following a visit to the Massachusetts Institute of Technology in 1965, where he had seen one of the first time-sharing computer systems in which a single mainframe computer was shared among many interactive users. Davies recognized that a major problem with the remote use of time- sharing systems was the “bursty” nature of the data communications traffic. A user sitting at a terminal occupied an entire telephone line, but spent most of the time thinking, so that the phone line was only about two percent utilised. This made access to time-sharing computers via long-distance telephone lines prohibitively expensive. His concept was to apply the principle of time-sharing to the data communications line as well as the computer. The result was a technique he called packet switching, by which a single line was shared between many users who sent their data in individual packets. A small, experimental network was established at the NPL in 1970, and a great deal of theoretical work on network simulation and congestion was undertaken in the 1970s.

Davies had an ambitious plan for a network of packet-switching centres that would create a national infrastructure for computer communications. However, it was a decade before the lethargic, pre-privatisation Post Office telecommunications division created even an experimental packet-switching service. In the United States, things moved much faster. In the Department of Defense′s Advanced Research Projects Agency (ARPA), Larry Roberts was struggling with the same problem as Davies. As soon as he heard of packet switching, he built it into his experimental computer network. This network, the Arpanet, was the prototype for the Internet.

In 1979 Davies stepped down as Superintendent, to return to research on his favourite topic of data communications. By now, computer communications had become an everyday reality. In financial institutions, in particular, a whole new set of problems of data security and encryption had surfaced in which Davies immersed himself. He wrote a major book on computer network security. After his retirement in 1984, at the age of 60, he became a leading consultant on data security to banks. In 1987, he became a visiting professor at Royal Holloway and Bedford New College.

Davies received numerous awards and honours, largely for his work in data communications. He was one of the first distinguished Fellows of the British Computer Society in 1975. He was awarded a CBE in 1983, the von Neumann medal in 1986, and was elected a Fellow of the Royal Society the following year.

Davies always had a deep interest in computer history, and the leisure of retirement combined with his professional expertise in data security enabled him to become a recognised authority on wartime cryptographic machinery.

Donald Davies passed away in 2000.

CCS Web Site Information

The Society has its own Web site, which is located at www.computerconservationsociety.org. It contains news items and details of forthcoming events and also electronic copies of all past issues of Resurrection, in both HTML and PDF formats, which can be downloaded for printing. We also have an FTP site at ftp.cs.man.ac.uk/pub/CCS-Archive, where there is other material for downloading including simulators for historic machines. Please note that this latter URL is case-sensitive.

Titan - the Poor Man’s Atlas?

Our esteemed chairman tells the story of the Cambridge development of the Ferranti Atlas - Titan.

Perhaps I should first explain Titan’s relation to Atlas. Titan was the name of the prototype Atlas 2, being a cut-down version of Atlas, properly known as Atlas 1. But I must also, briefly, tell the history of computing in Cambridge which led to the development of Titan.

EDSAC

EDSAC was the first Cambridge machine. It was the world’s first stored- program computer to go into regular service after the Manchester SSEM (which didn’t go into service). Of course, Manchester was one year ahead of us - we admit that. EDSAC was built between 1946 and 1949 and immediately went into a regular user service. Maurice Wilkes’ policy for the machine was that, as soon as it started working, users across the University should be encouraged to use it for their research. Maurice Wilkes sent round a note to key professors, saying: “I have built an automatic calculating machine; please come and use it”. And he even let research students use it - which, in many institutions, was unheard of in case the precious machine was wasted.

The basic statistics of EDSAC were as follows: it had a speed of 300 instructions per second; it had two kilobytes of memory in mercury delay lines; there was no online file storage, apart from an experimental magnetic tape system; and, at its peak, it had about 50 users. In effect, EDSAC replaced mechanical hand calculating machines. So it was 1,500 times more powerful than the technology it displaced. By this measure, it was the largest step forward in computation power in each of the technology generations. EDSAC was taken out of service in 1958.

EDSAC 2

EDSAC 2 was the second Cambridge machine. It was started in 1952, and was in service from 1958. The machine had packaged circuitry with replaceable plug-in units, the largest of which was several feet long. It was made up of miniature vacuum tubes. In the mid 1950s, the ferrite core became a feasible option and this technology was adopted. The machine also had a reasonably reliable file store made from two twin magnetic-tape drives purchased from Decca.

EDSAC 2’s job queue was its users, who were able to test their programs in two or three periods each day. They queued up by the input desk. Each user was allowed three minutes to read in their programs and data, run a brief test and collect output from a paper-tape punch or printer. The job control scheduler was a line of other users who provided the necessary peer pressure to keep to time.

EDSAC 2 was the world’s first microprogrammed computer and had built-in floating-point arithmetic. The instruction rate was 10,000/sec (only 40 times that of the EDSAC), with 88Kbytes of random-access memory. The peak user population was around 200 users.

The close down of EDSAC 2 in 1965 started a peculiar Cambridge tradition. We would never ‘open’ new machines, because everyone knew that new machines were usually suspect, tended not to work very well and what software they had was often unreliable and late. In short, users feared and mistrusted new machines. But, by the time one was closed down, users had developed an affection for the machine, and closure was a cause of emotion. Users crowded into the computer room, and watched Maurice Wilkes feed in the final job (on black punched tape, of course). EDSAC 2 duly played the Last Post. Grown men were seen to weep.

Titan

In 1960, half-way through the life of EDSAC 2, Cambridge started thinking about its third machine. Although we had enjoyed building new machines from scratch, we were conscious that our users wanted something more ready-made, so we sought to buy one.

And, of course, it had to be something that was worth buying, a machine significantly more powerful than EDSAC 2. At about that time, the government authority for funding universities was making grants to enable them to acquire their first computers. A grant of £250,000 was indicated for Cambridge. At that time many universities were buying the English Electric KDF9 which cost about £250,000. KDF9 was a fine machine, but EDSAC 2 was already the power of a KDF9, and there would be no point in us having one. After some searching, Cambridge found two machines that were sufficiently more powerful than EDSAC 2 to be worth considering and could take a growing user load.

One was the Ferranti Atlas and the other the IBM 7090. We could afford neither. There had been intensive talks with Ferranti and IBM, but a funding solution eluded us. Then suddenly, out of the blue, Maurice received a letter from Peter Hall who was then Ferranti’s computer manager in Manchester. Here are some quotations from that letter:

“Your computer problem is occupying our minds very much at the moment.........Cambridge and Manchester Universities are unique in that they are both machine designers and builders. Can we not exploit this at Cambridge? If we, Ferranti, sold you at “works cost” (he never defined “works cost”) un-commissioned standard parts of the present Atlas and a slow store, could you do the necessary connecting together, design any special bits, modifications as necessary and commissioning? In other words, we will sell you large chunks of standard hardware and you do the rest.

“In return for letting you have this hardware at works cost, you would let us have all designs and information on the work you do relating to it; e.g. in connecting in a slow store, programming etc. At a very rough preliminary guess we would, under this arrangement, let you have the following bits of hardware ... “

Peter then listed the parts to be provided, at ‘works cost’, including the CPU, minimal control for peripherals and magnetic tape units, slow (6μsec) memory, power supply, card reader, printer and control desks.

“I must emphasise that these items will be un-commissioned, and you would not just fit them together. We would, of course, send you all information to commission them yourself. You might like to send someone up here during commissioning of earlier machines, and we could probably send a man to work with you in Cambridge.

“This is just a thought. It’s very sketchy, but what do you think ? It seems to me that, if you have the necessary effort, it will help both you and us, because this would then be the basis for a second model of Atlas for the Ferranti market.”

Cambridge hadn’t many options available but, with the experience from the two EDSACs behind us, we were game for the challenge. We were being offered half a machine at a price we could afford and had the opportunity to work with Ferranti to design and prototype the rest of it.

There was discussion on what to call the new machine. In Cambridge there were some who preferred EDSAC 3, while Ferranti had a tradition of using names from mythology. “Titan” was suggested and started to be used. Then there was another letter from Peter Hall:

“We’ve been thinking about the name for this machine, and we think we want to call it Atlas 2”

Maurice Wilkes diplomatically accepted the suggestion, although Titan remained the name for the prototype Atlas 2.

Hardware

The initial Titan configuration was:

Arithmetic and Logical Unit (CPU) - identical to Atlas 1
Base-limit registers (instead of Atlas 1 page-address registers)
16k (48-bit) words of memory
A re-designed Peripheral Co-ordinator
Scatter read/write between memory and magnetic tape
Six one inch pre-addressed tape drives (there was no drum store)

Atlas 1 had a very elaborate one-level store - with page-address registers, drums and sophisticated software to transfer information between one and the other seamlessly: a very large virtual memory. So, system programs and user programs could spread themselves widely in the memory address space but at a substantial cost in hardware and in software complexity. Titan simply could not afford this.

But nevertheless, the requirement was to design an efficient multiprogramming system, so we had to make do with the poor man’s paging system: two registers, one determining where the current program started, and one saying how long it was.

The hardware peripheral controller had to be re-designed because, again, Atlas 1 was complex with many kinds of peripheral device, whereas all we needed was to control paper-tape readers, paper-tape punches, printers and magnetic tapes.

Magnetic tapes were the same as on Atlas 1: one inch wide tape containing pre-addressed 512 word blocks of data. Tape control on Titan was innovative: when reading a block, the controller could scatter the data in eight 64 word blocklets to eight separate memory addresses; similarly the hardware could collect eight separate blocklets from memory and store them in a single block on tape. This meant that memory management for data streaming and even user memory allocation could be handled in quite small units, important when you only have 16k words to play with.

David Wheeler was the design authority for Titan/Atlas 2. He had full control over the hardware team, consisting of both Cambridge and Ferranti staff; such was the trust between Ferranti and us. At the same time, Roger Needham, who was then just finishing his research thesis, helped David by writing design automation software on EDSAC 2, which laid out the printed circuit boards. Two Ferranti engineers were seconded to us for installation and commissioning. They seemed to camp out in the Eagle - years earlier, the pub of Crick and Watson fame.

A User's View of Titan

To house the machine, the University let us have an old-fashioned, tiered lecture theatre. The room had to be gutted and the floor levelled, but we kept the very top tier as a viewing gallery; important because we already realised that a machine of this type could not be operated by the users themselves, and that they were unlikely to love a machine they couldn’t see.

I’m told there is a published history of Ferranti which claims that Titan was built because Cambridge wouldn’t agree to have any machine that originated from Manchester. Not true! Titan was a machine fundamentally designed by Manchester. We were very happy to have it.

System Software

The Titan operating system was necessarily different from the Atlas 1 supervisor which was built around the hardware one-level store that we didn’t have. The Atlas 1 system programmers could just scatter programs and data all over memory and let the one-level store do the memory management; whereas on Titan, to keep the machine busy, we had to be very careful how we buffered data. One must remember that, in those days, operating systems were written not to make it easier for the user to use, but to optimise the efficiency of the machine - to make sure that the CPU was kept busy and not kept waiting for input and output. In fact operating systems in those days tended to make computers less not more easy to use, and there are probably few who remember the dark ages of the Fortran Monitor System. Nowadays ... - but one sometimes wonders!

There was no option for the Titan operating system to be other than a total re- design, although we had the benefit of the process technology devised at Manchester. This gave rise to what was essentially a separate joint project with Ferranti; a team at Cambridge working with a Ferranti team located at Lily Hill in Bracknell. The Cambridge team consisted initially of David Barron, Barry Landy and I. The Ferranti team was led by Chris Spooner.

It was Chris who brilliantly designed the input-output buffering system, the so- called Magnetic Tape Well. Like the Atlas 1 system, the Titan supervisor was designed as a multi-programming operating system, with the well acting in place of the one-level store and drum.

The basic hardware was delivered to Cambridge in 1963, and enough of it was commissioned to enable us to run programs in October that year. The supervisor and other system software was scheduled to become operational by late 1964. Like everyone else before and since, we were no exception to the rule that big software projects are always late.

It was at about that time we heard that Manchester was also running late. But Manchester had been rather astute: a temporary operating system and a temporary Mercury Autocode compiler had been quickly developed which were giving some user service admittedly in a non-optimal manner. David Barron, Chris Spooner and I travelled north to see for ourselves and were duly impressed. Any service is better than none, and theirs was holding off what would otherwise have been extreme user pressure.

We came back fired with enthusiasm and realised that this would provide an all- important breathing space. Two persons were commissioned to build a temporary supervisor and a temporary compiler. The temporary supervisor was written by Peter Swinnerton-Dyer; it had single serial job processing, no multi-tasking, and all input-output buffering was held in memory. This was inefficient, but it worked. At the same time, Maurice Wilkes himself wrote an EDSAC 2 Autocode compiler using his innovative list-processing system known as WISP. Later, Peter Swinnerton-Dyer re-wrote the compiler using more conventional techniques.

As the so-called main supervisor project dragged on, Peter said he thought he could write something better than the temporary supervisor, and started designing an input-output buffering system using magnetic tape. This worried those of us on the main supervisor team, realising that if we didn’t get a move on, we would have system software written entirely by one person. An admirable incentive!

The temporary supervisor was operational from 1964 and the main supervisor was re-scheduled for late 1965. It didn’t, in fact, see service for a good year after that, but that is another part of the story.

Time Sharing

As explained earlier, the main supervisor was a joint project with Ferranti. Like Atlas 1, it was a multi-programming job system: no terminals, no interaction, straightforward job input through paper tape readers or card readers into a magnetic tape buffer, spooled off, scheduled and run. Because the system was multi-programming, short jobs could overtake long jobs. Quick jobs could be scheduled with high priority. But everything was run strictly offline. Whereas the EDSAC 1 and 2 were machines that the users actually ran themselves, it was considered impractical to let users anywhere near Titan. Instead, they would hang their paper tapes on a peg on the wall, and their results were hung on the same peg somewhat later; apart from the opportunity to schedule the relative priority of jobs, this was little better than the batch operating systems of those days.

While the main supervisor was still under development in 1965, Maurice Wilkes set off for one of his regular visits to MIT in the US, where he saw the CTSS - the Compatible Time-Sharing System - on an IBM 7090. Maurice was totally enthused by what he saw and returned saying time sharing was the future, and this is what we must do. Given that we were several years into a very different kind of operating system, the news was hard to take.

After a good deal of encouragement from Maurice, and some hard thinking, we agreed that if MIT could do it, so could we. What was more, we could and would do it better! There were, of course, a few problems to overcome in re-directing the project. Firstly, ICT (who, by this time, had taken over Ferranti’s large computer interests) was working on a version of the supervisor that was to go to the Atomic Weapons Research Establishment at Aldermaston. The requirements were already diverging. So we virtually had two separate teams developing two different systems. It was mutually agreed Cambridge and ICT would go their separate ways.

Then, we made some hardware changes to Titan. David Wheeler designed a second pair of base and limit registers to provide what could be described a crude and simple form of memory segmentation. Thus a program could be securely located in two parts: the program itself and its data. If the program was interactive, we could get away with a single copy in memory with multiple copies of the data, one copy for each interacting user. For example a text editor, so important in a time sharing system, might be contained in a 512 word block, with the data for each user located in separate 64 word blocklets. This was essential because there was no way we could implement the so-called memory swapping of other early time-sharing systems - there was no fast data channel to a drum store.

But there had to be some form of sizable backing store with efficient random access. So, with the invaluable assistance of Basil de Ferranti we acquired a disc; this was donated by ICT in return for access by them to the completed system - something they never claimed. So, a Data Products disc store was procured which initially provided a user file store.

The third hardware development was to build a multiplexor to link local and remote teletype terminals into the main machine. Designed by David Wheeler, it had a few odd features. Characters read from a terminal were presented to the machine in reverse order and inverted; we quickly learned the technique of look- up tables, which was easier than persuading David to re-engineer his design.

But there was still one problem we had to solve, caused by the inability to implement efficient program swapping. Our new disc was adequate for user files, but not fast enough to swap active programs and their data in and out of memory. We had to turn our multi-programming system into a multi-access system. This was achieved by nothing more than sleight of hand.

Utility programs, such as text editors or file management functions, were written in tight code with minimal in-memory data, so that many users could be served by them using very little memory - just one copy of the program with as many blocklets of data as active users. A user program, however, could not access the terminal and could only read and write data to the file store. But as soon as the program was waiting for user input, any collected output was immediately spooled to the terminal. So a user program could undertake tasks on- line with immediate input-output, but could not actually do single character interaction. We therefore provided what might be termed interactive job-step running, which was a substantial advance on the old off-line regime and was what most people needed. We managed to support about 24 simultaneous users on what was, by now, a 48k word machine. The sleight of hand was very successful.

A file store for user programs and data was largely the work of Sandy Fraser. Sandy designed a sophisticated file system with very flexible user access controls. He developed mechanisms to enable the owner of a file to control who could write, who could read or who could only execute a file. Then it could control not only which user could do what, but which program could do what. There was a sophisticated backup and restore regime where files were automatically copied to magnetic tape and recalled on failure.

There was an advanced system for controlling the use of the very limited file space available. As well as an assigned upper limit, each user was allocated a quota which was added to his account daily and from which his actual use was subtracted each day. Users could create and modify files only while their accounts remained positive, thus giving flexibility and an incentive to save.

Finally, which we didn’t rate as particularly important at the time, there was the security of passwords. In those days, one tended to store passwords in a system file which you did your best to hide from enthusiastic inquisitive users. But, if someone cracked security and gained access to that file, everything was immediately compromised. Roger Needham hit on the idea of scrambling passwords with a one-way encryption algorithm. The system only needed to compare one scrambled password with a newly introduced scrambled password to check security, and the scrambled passwords were of no use to a hacker. The technique is now applied in most password handling systems.

Thus, it turned out that a system implemented in a well-designed software methodology could be adapted from an off-line multiprogramming facility into a time-sharing on-line system, providing a general-purpose university computing service.

Applications

Titan provided the University’s computing service for a good many years, and it was used for applications in research and teaching; it was used by scientists, by non-scientists and it was used by students. New areas of application were pioneered on Titan. The one to which I wish to refer here is Computer Aided Design.

A CAD research team was established in the laboratory. A PDP7 with a high- performance display was connected to Titan. These facilities enabled the CAD team to take CAD from a research topic in computing to become an application tool in engineering research. Later, when a second Atlas 2 was installed in a government-created institute in Cambridge, CAD techniques were systematically introduced into industry.

The basic statistics of the final Titan installation were an instruction rate of 0.25 MIPS, 0.75 Mbytes of memory and 128 Mbytes of file storage, and getting near to 1,000 users. Most of those users worked on-line, at least for their program development, controlled by some sophisticated resource allocation and control techniques. It was still a far cry from today’s personal computers, but by the standards of the day, it was a powerful machine serving a lot of users.

Closure

Titan lived on until 1973 when it was closed down and replaced by an IBM 370/165. We continued the tradition of close-down ceremonies. By then, Titan was in another building. We assembled the users in a nearby 500-seat lecture theatre which was filled to capacity. Live CCTV pictures of Titan were transmitted into the theatre, and a dramatic scene followed. Talks and tributes were first given. Then the final Close Down command was shown being typed on the operators’ console, followed by the chief systems programmer shutting down the operating system. The hardware technicians went round the computer room turning off all the units, and tension grew in the lecture theatre. Many were not sure how to get through the next few minutes. Finally, the Chief Engineer appeared, and approached the main power switch; he put his big hand on the switch and the camera zoomed in. The switch was turned and it went round and round and round (a student had loosened the switch the night before). The tension in the lecture theatre was suddenly broken, and everyone had a good laugh. That’s the way to say good-bye to a faithful old computer!

Summing up

So, what did we actually achieve? Firstly, we got a big enough third computer for our funds. We also helped Ferranti develop a new version of Atlas. The operating system was pioneering in many ways and provided practical and efficient time sharing facilities for a large user population. To support nearly 1,000 users on a single machine was crazy when you think about today’s technology.

The late Roger Needham, at the 50th anniversary of the EDSAC, ended a presentation about Titan in his inimitable fashion:

“It was fun. If you are in our trade, nothing gives you a charge like having put together a system which nobody else can match.”

Chris Spooner read Mathematics at Trinity College, Cambridge in the early 1950s and is reported to have achieved the top first in his final year. Sadly, he died while this article was in preparation. This article is dedicated to him.

Editor’s note: This is an edited transcript of a talk given by the author at the Science Museum on 20 September 2007. Contact David Hartley at .

The Ferranti Sirius at Monash University

Barbara Ainsworth tells the story of the Ferranti Sirius (or rather, several Sirii) at Monash University in Melbourne, Australia. Incredibly, two machines survive. One is now a centrepiece in their Museum of Computing History.

The Monash Museum of Computing History (MMoCH) was established in 2001 to preserve the early computing equipment distributed around the different campus sites that now form Monash University, Melbourne, Australia. Amongst these early pieces was a Ferranti Sirius computer which was stored under a stairwell. This computer has now been transferred to the MMoCH and is on permanent exhibition as part of the museum’s display at Caulfield campus. The provenance of this computer is an on-going research project of the MMoCH.

Initial research was focused on the Ferranti Sirius still on campus, but references and personal recollections from early computer users proved to be very confusing. Further investigations revealed that there were actually four different Sirius computers on campus at different times during the 1960s. In 1962, these four computers were located at different sites; two Sirius computers at the Melbourne Computer Centre operated by Ferranti Ltd, one at Monash University Clayton Campus and one at a commercial research operation run by ICI Australia & New Zealand (ICIANZ), at Ascot Vale.

Monash University entered the computing area in the early 1960s but there were already local developments in computing in Australia during the late 1940s and 1950s. These were overshadowed when the major foreign companies expanded their computer sales operations into the Australian market in the 1950s, establishing outlets in Sydney and Melbourne. By 1962, there were about 80 computers installed or on order around Australia. About a third of the computer installations were located in private commercial operations. Eight universities had installations including Monash University. A small number of government departments had their own computer. As an alternative, the computer companies operated service bureaux. IBM dominated sales at this time.

The Ferranti Melbourne Computer Centre was established in 1960 with Barry de Ferranti as Sales Manager. Richard Cross and Richard Levingston moved from England to work at the Melbourne Ferranti office. Robin Goodchild was the bureau programmer and started with Ferranti in 1962. The Ferranti Melbourne office operated from “Stanhill”, Queens Road, Melbourne. This busy Computer Centre housed a 1,000 word Sirius and later, a 7,000 word Sirius. Barry de Ferranti recalled that these machines were in great demand.

“The team I assembled to establish the Melbourne Computer Centre in Stanhill, in 1960, helped promote technical computing in Victoria. Before long, Melbourne-based government departments, manufacturers, universities and utilities were loading our little Sirius beyond expectation; even a motor journal ran a survey, with extraordinary response, causing us to work shifts to cope.”

Ferranti produced a range of computers in the early 1960s, each named after mythical figures, including Perseus, Pegasus, Sirius, Orion, Mercury and Atlas. At Ferranti in the late 1950s, Gordon Scarrott developed an unusual transistor circuit, the Neuron, for the Orion 1 computer. To test the Neuron circuit, a test bed computer called NEWT was constructed. NEWT took on a life of its own and was re-engineered and sold as “Sirius”. It was announced to the public in a press release on 19 May 1959. It offered Sirius as a transistorised, desk-sized, electronic, digital computer. The release claimed that it would be the smallest and most economically priced computer in the European market.

Sirius was manufactured at the Ferranti factory in West Gorton from 1960 to 1963. J.F. Wilson records that the company produced around 22 installations. Research shows 15 being sold to real customers and about five used by Ferranti as demonstration models and in bureau services. Only seven were exported. Strangely, four of these seven were sent to Melbourne, Australia.

Sirius was not the most powerful of computers. Its serial architecture kept the cost and size down, but meant it was not particularly fast. It had only 1,000 words of store which could be increased to a maximum of 10,000 words. Ferranti Ltd sold the basic CPU with 1,000 word capacity. Memory could be expanded by adding on additional memory cabinets with 3,000 words of store. Input and output was entirely by punched paper tape. However, it had some very attractive features. It was one of the earliest computers to use transistors rather than valves (vacuum tubes). So it was relatively small (small enough to stand behind an office desk), had low power requirements (it ran off a standard 230 volt 13 amp socket) and had no need for special air conditioning. A full system consumed about 2kW. It had a decimal display and a facility to slow the processor for demonstration and educational purposes. It was a good, general purpose and relatively inexpensive machine for its time, ideally suited to educational establishments.

Monash University decided to purchase a computer in 1961. It was a new educational institution, the University receiving its charter in 1958 and enrolling its first students in 1961. Today, Monash University is composed of a number of campus sites distributed around Victoria. The first campus under this university title was located at Clayton.

Monash wished to create a computer centre to provide administrative support as well as academic research facilities, as part of its new Clayton campus.

The Computer Facilities Committee was established under Professor Westfold, Chairman of Mathematics, to investigate the different models available and compare their prices and capabilities. The Computer Facilities Committee had trimmed their report to consideration of two different computer models by 29 September 1961. The two frontrunners were the IBM 1620 and the Ferranti Sirius.

In December 1961, Professor Westfold delivered the final recommendation to purchase a Ferranti Sirius to the Vice-Chancellor, Professor Matheson. Professor Matheson officially ordered the Ferranti Sirius on 20 December 1961.

The original tender document from Ferranti offered Monash University:

- One basic Ferranti Sirius computer with 4,000 word storage,
- two TR5 tape readers,
- one teletype punch,
- one desk-mounted, simplified set of tape editing equipment.

This offer was put at a cost of A£33,137. The tender gave a delivery period of nine months from the confirmation of order.

Something Borrowed

As part of the contract, Ferranti Ltd offered to provide the loan of their new 7,000 word Ferranti Sirius, which was already being built in Manchester, until the delivery of the University’s own computer. The Ferranti Bureau would effectively be located at the University, but the Monash Computer Centre staff would have direct access to the Sirius. The University would provide space for the temporary loan of the 7,000 word Sirius, with two sets of Ferranti/Creed model 75 tape editing equipment. Ferranti would supply a fully trained programmer during office hours. The University needed to employ a computer operator (referred to as ‘she’ in the proposal) to be trained by the Ferranti programmer. The University would be able to book two hours / day free of charge and up to three hours extra if the machine was not booked.

This shared arrangement was apparently quite successful. Ferranti Ltd completed the production of a new Sirius for Monash University and it was shipped to Australia in late 1962. Brian Parker was sent from England to help with the installation. It was operating, after acceptance tests, by November. The Ferranti Bureau 7,000 word machine then was returned to the Melbourne bureau. Cliff Bellamy transferred from Ferranti to Monash University in early 1963. He subsequently spent more than 30 years supervising and developing computer facilities at Monash.

Our Own Machine

The Monash Sirius about 1963

In its quarterly report for January to March 1963, the Computer Centre gave some details on the operations of the Sirius. They were operating the 4,000 word Sirius with two paper tape readers and one tape punch. There were two Creed teleprinters. Students were using the computer for elementary data processing and technical computations. The Computer Centre was also actively encouraging staff to learn Autocode and machine code programming. The largest single user of the computer, in terms of hours, was the Administration Department, with 91 hours in the first quarter. Other major users were the Physics Department, Engineering Faculty, Chemistry Department and the Computer Centre. However the Maths Department only used three hours. The Sirius was in high demand. Bellamy reported, in May 1963, that he expected the machine to be running two eight hour shifts daily to meet user demand. Alan Cowley worked on the machine in 1963 and he converted it to a real-time, interrupt-driven machine for a specific project in the Chemical Engineering Department.

The Computer Centre at Clayton campus was expanded with the purchase of a CDC 3200 in April 1964. By 1967 the Sirius was probably only used about two hours a day on an irregular basis as most work was run on the CDC 3200.

ICI Australia & New Zealand is Generous

In 1967, the University was offered the donation of another 7,000 word Ferranti Sirius from ICIANZ, who had also purchased a Ferranti Sirius in 1962. Their Ferranti machine had been installed at ICIANZ Central Research Laboratory at Ascot Vale in late February 1962, having been purchased for A£40,000. Brian Parker commented that this machine had reliability problems. The processor backplanes had been resoldered to remove possible dry joints.

The Sirius was ICIANZ’s first computer. The company newspaper ICIANZ CIRCLE holds several articles noting the arrival of the new computer. There is also a quite detailed article on how the Sirius actually processed information, including a strip of coded tape reproduced down the side of the page of the article. Recently run on a simulator in the UK, this tape is indeed the code relevant to the hypothetical problem described in the text!

The company continued to update their computing facilities. When they purchased an IBM System 360 in 1966, it was decided that this was adequate for all their computing needs. In 1967 ICIANZ offered to give their Sirius to Monash University. The University accepted the donation of the ICIANZ Ferranti Sirius but, in the Annual Report for 1967, the Sirius machines were described as “Other Equipment”.

The original 4,000 word machine is described with the statement, “...is maintained by, but has little other active support from, the Computer Centre.” The ICIANZ machine is dismissed as well with the sentence, “The original intention was to use it as spare parts but later it was installed in the Chemistry Dept., for special duties.”

The Sirius installations were finally decommissioned in 1972. The original 4,000 word Sirius was relocated to open display in a building at the Clayton campus until it was given to the Monash Museum of Computing History and placed on display as part of its permanent exhibition at Caulfield campus in 2005. The ICIANZ 7,000 word Sirius, located in the Chemistry Dept., was donated to Museum Victoria in 1975.

Yet Another (Rather Special) Sirius

Meanwhile, a few miles away, the Chisholm Institute in Caulfield had developed its own computing school during the 1960s. A second (1,000 word) Sirius computer at the Ferranti Melbourne Computer Centre was relocated to Caulfield on lease. It was then purchased in about 1964. The 1,000 word Sirius was originally built as a prototype model. In 1961, it was exhibited, as a demonstration model, by Ferranti, in the British Pavilion, at the Italian Centenary Exhibition in Turin. It was then sent back to be used at the Newman Street bureau in London. Later it was sent to the Melbourne Ferranti Bureau and subsequently transferred to Caulfield Technical College. The College placed it in the Department of Electronic Data Processing. On open days, staff would program the computer to make “music”. The processor had an audio device which could be programmed to make different tones. The computer could play several pieces. Peter Juliff remembers the computer playing a theme from Bach’s 4th Brandenburg Concerto and then “Cockles and Mussels”. The Sirius was replaced by new equipment in the late 1960s and disappears from the record.

After several changes of name and various mergers, Chisholm Institute became part of Monash University. The MMoCH is located at the Caulfield campus.

Looking Back

The decision to purchase a Ferranti Sirius in 1961 started a long development of computer technology and education at Monash University. The Ferranti Sirius was a small production computer, but it was an economical installation that suited the needs of both academic and administrative operations in the early 1960s. The large number of installations in Melbourne, nearly 20% of all Sirius computers produced, reflects the sales abilities of the local Ferranti office. This small computer had a significant influence on Monash University with all four installations being associated with the University at some point during their working life in Australia.

Sirius on Display at the Museum

Two of these installations have, fortunately, been preserved. The 4,000 word Sirius is now on permanent display at the Monash Museum of Computing History, in a dedicated showcase, supported by a range of input/output devices. The display includes a short film “Instant Arithmetic”, produced in 1963 by S.E. Fargher, which illustrates the process of calculating a problem on a Ferranti Sirius for a non-academic audience. The larger 7,000 word computer, also supported by a selection of input/output equipment, is in storage at Museum Victoria. It is unfortunate that the 1,000 word Sirius, displayed originally at Turin, has been lost. There is also no record of the fate of the 7,000 word Sirius at the Ferranti Melbourne Computer Centre.

Cliff Bellamy, Director of the Computer Centre, noted in 1965 that “Monash is probably the most computer-conscious university in Australia”. He attributed this to the low average age of the staff and their acceptance of new approaches to problem solving. He could have also cited the acquisition of the Ferranti Sirius in 1962. The reputation and output from the Computer Centre started on this early computer. Although the Sirius was small, it demonstrated the potential of computing in the work of the University. The installation of Sirius, in 1962, started the strong tradition of innovative computing at Monash University.

The MMoCH has received invaluable assistance from Steve Poulton and the late Brian Parker, with many technical and historical details on the development, operation and installation of the Ferranti Sirius at Monash University. We have also received advice from John Feist, Chris Burton and a large number of staff and students who had experience with using the Sirius installations at Monash University - Clayton campus and Caulfield campus, as well as ICIANZ’s computer. Former employees of Ferranti Ltd in Melbourne have also contributed their knowledge to this research project. Staff at Monash University Archives provided access to the University’s early files on computing.

Editor’s note: Barbara Ashworth is the curator of the Monash Museum of Computing History. A longer version of this article, which includes an interesting overview of early computing development in Australia, can be found at www.infotech.monash.edu.au/about/projects/museum/papers/first-computer-at-monash-university-v4.pdf.

You Can Become a Computer Programmer!

In the early 1970s, the IT industry was expanding at an astonishing rate. There was a severe shortage of programmers and other IT staff. That caused salaries to rise very quickly (Oh happy days!). In London, programming training schools appeared, like a rash, all over the West End. Some of them were good and some less good, but they all held out the promise of a lucrative career in IT to the unsuspecting public.

One such organisation advertised heavily on London Underground trains. They used a variant of the old IBM aptitude test. Five number series were shown each followed by a “?”. The strap line was “If you can do three of these puzzles by the station after next, you can become a computer programmer.”. Obviously, being a computer programmer was the aspiration of every right-thinking person. Every few weeks the series were changed. Naturally, those of us already in the business couldn’t resist the challenge.

Well, the training school wasn’t in the business of putting people off, so the puzzles weren’t very hard. Or, at least, three of the series weren’t very hard. Anybody who was half-way competent could solve three of them by the time the back of the train had left the station. Generally, the fourth took a little longer and the fifth was the difficult one, but could indeed, usually be solved by the station after next.

Then, one day, a new puzzle appeared. And the difficult one was more difficult than usual. Not only could it not be solved by the station after next, the end of the line was encountered before I had made any headway. In fact, it took me three whole days.

Now, given the rather superior collective intellect of the Computer Conservation Society, I suspect that many of you will do better than I did. So let’s see:

The first few people to email the correct next number (preferably with an explanation) to the editor will be acknowledged in the next issue of Resurrection.

Good luck!

Letters to the Editor

From Harold Hankins

Ron Foulkes’ article on the development of the MetroVick 950 and AEI 1010 computers (Resurrection 43) reminded me of the time I was working in adjacent laboratories, during the early 1960s, leading a team responsible for the development and design of the AEI Type 1200 telecontrol and computer display system.

The development of the type 1200 was a commercial attempt by AEI in 1961 to break into the emerging remote monitoring, control and display markets of the public utilities, offering digital systems with greater flexibility than existing analogue ones. The type 1200 used digital circuits from the AEI 1010 computer to realise all the logic functions for the outstations and a control centre. They were of hard wired fixed logic design, to generate the necessary outstation control algorithms and error correcting messages. AEI’s experience in developing radar displays led to a decision to use a 21”CRT on which to display outstation information using electronically generated English characters, numerals and special symbols on a 4 x 4 matrix, together with basic line diagrams. In 1961, no such systems were commercially available, although today they are commonplace.

The CRT display system was of the random scan type, allowing the CRT beam to be positioned at any point on the screen, rather like a pen being positioned on a piece of paper. Once positioned, X & Y deflection signals derived from the selected 4 x 4 diode matrix drew the character on the CRT screen. The CRT display unit required sophisticated analogue circuitry, using ultra linear deflection transistor amplifiers, magnetic focussing and deflection of the beam, pin cushion magnets to correct for distortion, dynamic focussing of the CRT beam to all points of the display area and rapid defocusing of the beam, should the scanning system fail, to avoid burning a hole in the phosphor. The L4 fluoride phosphor, developed by Ferranti, was flicker free when refreshed 12 times per second, giving a pleasant orange on black display when viewed through a neutral density filter fixed to the face of the CRT.

The AEI 1200 Display

The display file holding the required alpha numeric character codes was based on the AEI 1010 magnetic core store, together with the same read-write circuitry, for the required 4096 words of six bit length, instead of the computer’s 44 bit length. The character 4 x 4 matrix was a committed diode array mounted on the same 10 ″x 8″ PCBs as used in the 1010 computer. The magnetic core store was read continuously 12 times a second and a one MHz clock drove the character generator at 50,000 characters per second for a flicker free display.

The computer display prototype was demonstrated in 1962 to Dick Grimsdale, who had left Manchester University to join AEI (Automation), and to Max Jervis (CEGB). The CEGB had contracted AEI (Automation) to produce an Alarm Analysis System using the AEI 1010 Computer at the Oldbury Nuclear Power Station then being constructed by AEI. A type 1200 computer display system, interfaced to the 1010, using the same logic PCBs where necessary and housed in type 1010 cabinets, was then ordered by the CEGB, together with a number of 21″ CRT displays, to help reduce the size and number of control and indicator panels in the Control Room. The overall system became operational in 1964, being one of the first, if not the first, commercial application of a computer display system. The prototype won “Best Industrial Exhibit” at the IEA Exhibition at Earls Court in the same year.

A second system was ordered for use in another Nuclear Power station, interfaced to a GEPAC 4000 computer which had been acquired by AEI (Automation) under licence from GE of America, to replace the AEI 1010. If I recall correctly, another was used in a steel works application using a GEPAC 4000.

Another computer display system known as the type 1400, interfaced to Ferranti type 1600 computers, was ordered for use in an Admiralty simulator project. The type 1400 was a modified type 1200 computer display, with a character generator capable of producing 250,000 characters per second for a multiplicity of 24″ circular, horizontally-mounted CRTs.

The AEI type 1200 telecontrol system was never put into production. The Midlands Electricity Board acquired the prototype and installed it at the Hanley Control Centre where the late Horace Heath (MEB) and Mike Jennions (AEI) carried out some interesting pioneering experiments on digital telemetry and control of electrical distribution networks.

The AEI 1010 - an Additional Note

Due to an editing error, a paragraph was inadvertently omitted from Ron Foulkes’ article on the AEI 1010 in Resurrection 43. Readers will recall that there is a discussion of the 1010’s ability to timeshare. Although Resurrection has considered the introduction of timesharing in some detail, the means by which it was achieved have received less attention. Ron went on to describe how it was done on the 1010 as follows -

The process was controlled by a continuously running housekeeping routine. It could instruct the central processor to stop working on one application and start or restart working on another. It was a piece of software that had sole use of 99 blocks on the magnetic drum and 11 blocks in the working store. The housekeeping routine allowed for up to four parallel programs which, between them, could have up to nine parallel branches. This allowed the programmer to use parallel working within one application program.

Forthcoming Events

Every Tuesday at 12:00 and 14:00 Demonstrations of the replica Small-Scale Experimental Machine at Manchester Museum of Science and Industry.

Every day Guided tours and exhibitions at Bletchley Park, price £10.00, or £8.00 for concessions (children under 12, free). Exhibition of wartime code-breaking equipment and procedures, including the replica Bombe and replica Colossus, plus tours of the wartime buildings. Go to www.bletchleypark.org.uk to check details of times and special events.

Lecture Programme

13 November 2008 London seminar “Iconic machines: Exhibiting a History of Computing”. Speaker Doron Swade.

18 November 2008 NWG seminar “Elliott’s Contribution to Computers”. Speaker Laurence Clarke.

15 January 2009 London seminar “ICL-Fujitsu Technology Collaboration”. Speakers Tom Hinchcliffe & John Vernon.

20 January 2009 NWG seminar “Some CAFS Applications”. Speakers Hamish Carmichael and Martin Wright.

17 February 2009 NWG seminar “Computers in Telephony”. Speakers David Parsons and Nigel Linge.

19 February 2009 London seminar on “EMI Computers - A Reminder”. Speakers Michael Knight & John Bradbury.

12 March 2009 London seminar “The Archaeology of Very Early Algorithms 1948-58 : Strachey’s Love Letter Generator”. Speaker David Link.

23 April 2009 London seminar “JANET - the First 25 Years”.

14 May 2009 London seminar “BBC Domesday Book Project”.

Details are subject to change. Members wishing to attend any meeting are advised to check the events page on the Society Web site at www.computerconservationsociety.org for final details which will be published in advance of each event. Details will also be published on the BCS Web site (in the BCS events calendar) and in the Events Diary columns of Computing and Computer Weekly. London meetings take place in the Director’s Suite of the Science Museum, starting at 14:30. North West Group meetings take place in the Conference Room at the Manchester Museum of Science and Industry, usually starting at 17:30; tea is served from 17:00.

Queries about London meetings should be addressed to Roger Johnson at , or by post to Roger at Birkbeck College, Malet Street, London WC1E 7HX. Queries about Manchester meetings should go to William Gunn at .

Committee of the Society

[The printed version carries contact details of committee members]

Chairman  Dr David Hartley FBCS CEng
Vice-Chairman  Tony Sale Hon FBCS
Secretary, Chairman DEC Working Party  Kevin Murrell
Treasurer  Dan Hayton
Science Museum representative Dr Tilly Blyth
TNA representative  David Glover
Bletchley Park volunteers representative  Pete Chilvers
Chairman, Elliott 803 Working Party  John Sinclair
Chairman, Elliott 401 Working Party  Arthur Rowles
Chairman, Pegasus Working Party  Len Hewitt MBCS
Chairman, Bombe Rebuild Project  John Harper Hon FBCS CEng MIEE
Chairman, Software Conservation Working Party  Dr Dave Holdsworth CEng Hon FBCS
Digital Archivist & Chairman, Our Computer Heritage Working Party   Professor Simon Lavington FBCS FIEE CEng
Editor, Resurrection  Dik Leatherdale MBCS
Web Site Editor Alan Thomson
Archivist  Hamish Carmichael FBCS
Meetings Secretary  Dr Roger Johnson FBCS
Chairman, North West Group  Tom Hinchliffe
Dr David Anderson
Peter Barnes FBCS
Chris Burton CEng FIEE FBCS
Professor Martin Campbell-Kelly
George Davis CEng Hon FBCS
Peter Holland
Dr Doron Swade CEng FBCS

Point of Contact

Readers who have general queries to put to the Society should address them to the Secretary: contact details are given elsewhere. Members who move house should notify Kevin Murrell of their new address to ensure that they continue to receive copies of Resurrection. Those who are also members of the BCS should note that the CCS membership is different from the BCS list and is therefore maintained separately.

Aims and objectives

The Computer Conservation Society (CCS) is a co-operative venture between the British Computer Society, the Science Museum of London and the Museum of Science and Industry in Manchester.

The CCS was constituted in September 1989 as a Specialist Group of the British Computer Society (BCS). It thus is covered by the Royal Charter and charitable status of the BCS.

The aims of the CCS are to

• Promote the conservation of historic computers and to identify existing computers which may need to be archived in the future
• Develop awareness of the importance of historic computers
• Encourage research on historic computers and their impact on society

Membership is open to anyone interested in computer conservation and the history of computing.

The CCS is funded and supported by voluntary subscriptions from members, a grant from the BCS, fees from corporate membership, donations, and by the free use of Science Museum facilities. Some charges may be made for publications and attendance at seminars and conferences.

There are a number of active Working Parties on specific computer restorations and early computer technologies and software. Younger people are especially encouraged to take part in order to achieve skills transfer.