Grand Thieves & Tomb Raiders

by Rebecca Levene

  Hauser represented the Cambridge University view – that computers were exciting and inevitable, and that there was already a nearby talent pool determined to get involved – and he encouraged Curry to make his own start. When Furber was building the first MK14 he had invited his friend and fellow University Processor Group member Sophie (then Roger) Wilson over to see it. Wilson was unimpressed. ‘Sophie looked at the MK14 and gave a characteristic “I could do better than that” reaction,’ Furber says. ‘And she went home over the Easter Holidays and came back with a design for the Hawk.’

  The Hawk was inspired – a real improvement on the MK14. When Curry and Hauser saw it, it was obvious that this was the future, regardless of Sinclair’s plans. At the end of 1978, they set up a new company to sell it, named Cambridge Processor Unit, a jokey riff on the central processor unit at the heart of their computers. It is doubtful Sinclair knew much of their activities, but they were certainly happening on his premises, as Curry admits: ‘Hermann came to spend a lot of time in the office in King’s Parade. It started before we had a chance to move into CPU’s own offices.’

  Sinclair finally shook himself free of the NEB by quitting Sinclair Radionics, with a golden handshake of £10,000. It wasn’t enough to prevent him from having to sell his house and his Rolls Royce, but it did leave him able to take full control of Science of Cambridge, where Curry’s efforts gave him a running start: there was product line to extend, and a consumer base to leverage. Curry’s absence also meant there was no objection to renaming the company once again, and it soon became Sinclair Research Ltd.

  The MK14 had been a great success for a small outfit, but with its open circuitry and Heath Robinson component list, it didn’t look like a consumer product. Sinclair’s previous business had been built on compactness and smart design, and he carried that philosophy over to Sinclair Research, along with a crucial third pillar: the products had to be cheap.

  Sinclair brought Jim Westwood over from Radionics, to weave his magic and minimise both the size and cost of a home computer. The Financial Times had run an article in 1979 that daringly predicted the appearance of a proper consumer computer for under £100 – meaning a machine with a QWERTY keyboard and full screen display. The newspaper set the timeframe for this at five years. Sinclair told his team that they had to make it happen within six months.

  Meanwhile, the Cambridge Processor Unit needed premises, staff and money. The core members were recruited gradually. Chris Turner, a well-regarded design engineer from Philips, was employee number one, and Hauser’s persuasive character and university connections helped CPU to assemble an excellent team. Furber started spending more time with the company, as did Sophie Wilson, who now had a completed prototype of the Hawk, her rival to the MK14. For premises, they found a small office squeezed down a rather forbidding alley. Money was harder to come by. The new company had yet to generate any income from products, so the team set about selling their sole asset: their expertise.

  Sophie Wilson’s first electronics job had been to make an automated cattle feeder, and in a sense CPU’s first consultancy job covered similar ground. A Welsh one-armed bandit manufacturer called Ace Coin wanted to make its machines more attractive to punters and statistically more precise in their payouts. Furber set about building a sound-and-light show, and then with a ‘possibly over-complicated arrangement’, used another processor to control the reels. There was a problem with this approach, though: the electronics of the era were susceptible to being shorted by nearby electrical pulses. In practice this meant that the machines could be made to pay out a flurry of coins simply by flicking on a cigarette lighter, and unfortunately this trick was becoming rather well known. Wilson devised a typically ingenious solution. An FM receiver was built into the casing that would detect these attacks and cut out the mechanism. Ace Coin was satisfied that it was now only the customers who would play against the odds.

  As a consultancy, CPU surfed a wave of quite random contracts – one they pitched for was to provide the graphics for the computer screens in the film Alien – but its core product was a set of modular circuit boards developed from Wilson’s Hawk design. It made sense to Curry to sell these to the same hobbyist market that had bought the MK14, and an expandable kit was put together.

  The £20,000 earned from Ace Coin was enough to put Sophie Wilson’s Hawk design into production as the ‘System 1’. It was sold in the form of modular cards that could be mounted on a rack. This approach allowed its users, mainly university labs but also some home hobbyists, to add floppy disc drives and more memory to the machine as well as – for the real computer experience – a monitor and keyboard. And it was all assembled by the user; in this market, knowledge of soldering and debugging was a given.

  But CPU needed a brand name to market its products, one that would reflect the company’s potential for growth and also lend itself to a logo that would look good as a silhouette in the smudgy, black-and-white adverts of electronics magazines. Although it had few competitors in Europe, CPU wanted to stand out in a global market rapidly filling with manufacturers, of which the most prominent was Apple – not least because it was always at the top of alphabetical lists. Curry and Hauser chose the name Acorn.

  CPU and Science of Cambridge were selling similar products into similar markets, and like Sinclair, Curry realised that although there was clearly a huge interest in computing, the ‘boffin’ user base was reaching its limits. There was a charm and a teasingly futuristic feel to their output – the System 1 featured as a prop in the Blake’s 7 BBC TV series – but US products, such as the Apple II and the Commodore Pet, were professionally packaged with built-in keyboards and branded monitors. Beside them, the British kits would be forever trapped in the world of Practical Wireless, and Curry knew it: ‘We realised after a year or so that selling modules that were really intended for the industrial market was not going to get far in the consumer market.’ With their exposed circuitry and the technical know-how required for assembly, kit computers were also forbidding and impractical for the markets with most profit potential: homes and schools.

  After his departure from Radionics, Sinclair had lost control of the Newbrain, but the same compact, cost-saving mantra that had driven its development was now echoing around the offices he had taken over in King’s Parade. The magic price point target was £100, psychologically important to consumers, but also fixed in Sinclair’s mind by that prophetic Financial Times article. His vision was of a single, smart box that would overcome the hurdle of high hardware costs by using a cassette recorder for storing programs, a clever trick for a built-in keyboard and, most challengingly, a home television for the display.

  Jim Westwood was given the job of making the computer send its pictures to the television through the analogue aerial input. The technology was known, but not at this price point – a modulator that could send a digital image to the analogue scan lines of a home television had never needed a mass market before. But eventually Westwood summoned Sinclair to witness a stable screen that read ‘Jim has done it’, a nod to the BBC’s Jim’ll Fix It programme. It was the first practical demonstration of the way that British players would see their computer games for the next decade.

  The rest of the machine was a collection of compromises. It had one kilobyte of memory – four times that of the MK14, but tiny compared to its American rivals. The processor was so slow that the screen went black when it was asked to do anything at all. And the miraculous money-saving keyboard consisted of a pressure pad broken up into 96 sections, each representing a key, which had to be pressed with real force to provoke a response. It worked, and had three parts instead of the hundreds of a fully moving keyboard, but was only barely usable.

  The computer was finished within Sinclair’s bold six-month timeframe, though, and it would sell for the uniquely desirable price of £99.95. A young industrial designer called Rick Dickinson had devised a sleek white plastic case, about the size of a hardback book, which made it look like a sli
ce of the future. Sinclair called it the ZX80.

  For as long as computer games have existed, the ‘platforms’ – the computers or consoles that play them – have determined the boundaries of the medium. Most obviously this flows from the technology, as graphical and computational power lifts games or holds them back. But the platforms also segment the games’ players – once they own a particular computer, they are locked into buying only the games that computer can play. This affiliation with a particular platform often feels deeply personal, like supporting a football team, and their chosen machine can subtly influence a gamer’s habits and tastes.

  Another, hidden distinction between platforms happens under the hood, one that was especially true of the computers used in 1980s Britain. Only a handful of microprocessors were widely available to computer manufacturers at the time, and even fewer that would make sense used in a home computer. Machines that looked entirely different on the outside could have identical technology at their heart and, wrinkles aside, converting a game from one of these to another would be quite straightforward. But where the processors were different, the game’s programmer may as well have been rewriting from scratch – the step change in effort could be the difference between a couple of days of tweaking and months of hair-tearing frustration. The boundary lines that gamers saw between machines could look very different from a developer’s perspective.

  The early eighties computer games industry – the ‘8-bit’ era – is a story of two chips: MOS Technology’s powerful but pricey 6502 chip, and Zilog’s cheaper workhorse, the Z80. The choice between them was often made at the whim of a developer, or due to the hard realities of cost. But each individual decision would help set the landscape of British computer games for a decade.

  Sinclair’s ZX80 was not named by combining the year of its launch with letters that sounded futuristic, although that impression was a happy one. The name was actually chosen by Rick Dickinson and his team to convey that it was powered by a Z80 chip – with an extra, unknown ingredient.

  For Clive Sinclair, that ingredient might as well have been success. He was back in the game, and his marketing was shameless. Adverts boasted that the ZX80 was ‘powerful enough to run a nuclear power station’, which even then required a very indulgent analysis to accept. He also made sure that the public knew that his children had helped him test it. This was a consumer product that looked smart in the home – it sat neatly under the family television as a tool, a conversation piece, or a mark of aspiration. For all that it invited hours of lying on your stomach jabbing fiercely at keys, it also opened a portal to a new world.

  Sinclair’s company had a secretive culture, so the team at Acorn found out about the unit at the same time as the rest of the country. By then, Curry and designer Nick Toop had been developing Acorn’s own machine, called the Atom, which was much closer in appearance to the American ‘beige boxes’. This was Curry’s intention: ‘The Atom was in many ways like the Apple II. It was smaller and cheaper, and had more bits to go with it as well.’

  The Atom was a reduced version of the 6502 based System range of computers that Acorn had been selling to the hobbyist market, packaged into a consumer-friendly machine complete with high-resolution graphics, colour and sound. It appealed particularly to schools: it was a tough unit with a full-sized keyboard, and, for the first time on a home computer, it could be networked using technology that Acorn had devised for sending data around its own offices. The Atom cost more than twice as much as the ZX80, and the education market was still tiny, but some schools – mainly fee-paying – could justify the extra expense of a computer that looked as if it would withstand an onslaught of children, especially if the teacher could manage all of the computers in the classroom. And for all Sinclair’s hype, even a cursory glance revealed that the Atom could do much more than its flickering, blocky rival.

  Acorn arranged to launch its new computer at an electronics show in 1980. Two months earlier, Curry had secured a commitment from a supplier in Hong Kong to have the first cases ready within six weeks – much faster than the twenty it would have taken in the UK. By the deadline, having heard nothing, he flew over only to discover that they hadn’t even started. Curry stayed in Hong Kong and refused to leave until the cases were finished. They were ready just in time – he took the prototypes straight from the airport to the exhibition.

  Sinclair Research and Acorn were now clear rivals. Both sold their products by mail order, as cost-saving soldering kits for enthusiasts, or ready-made for consumers. They looked to the same home and schools markets, segmented by price but not much else. And vitally, each machine had a small version of the programming language BASIC built into its hardware. Sophie Wilson had written an elegant, compact form for Acorn. Sinclair had outsourced the job to a company called Nine Tiles – which probably did the best job possible given the fierce deadline, although the end result was still visibly compromised. BASIC was famously easy to program, but notoriously slow at running the code once written. On the Atom, this meant a more sluggish appearance. On the ZX80, you might hear a faint buzz, the machine would be as hot as ever, and then the screen would switch off and ignore you altogether.

  Sinclair often seemed to target price before quality, and the ZX80 was certainly prone to overheating – some apparent ventilation vents were in fact painted on. But the Atom also had problems: a key component was designed to hang upside down in the case, and as it heated through normal use, it would slide gently out. And both companies were discovering that allowing inexperienced consumers to build their machines from kits threw up huge numbers of support issues. Furber recalls that one despairing customer wrote to Acorn to say that they knew that chips were heat sensitive so they had glued them in instead, ‘and it still didn’t work!’

  Acorn had also adopted Sinclair’s trick of using a cassette interface. Tapes were cheap and common, but incredibly frustrating: saving and loading took minutes and could still fail, while finding the file on the tape meant listening for gaps in the computer’s recordings, which to the human ear sounded like ungodly screeches. But tapes could also be made at home and swapped and sold, and for the first time computer manufacturers started to see a retail software industry grow around their products.

  Steve Furber’s only published game was for the Atom. It was a clone of the arcade game Asteroids that he had written at home on a 6502 machine he had built himself. It was quickly seized upon for the first ‘Games Pack’ tape released by a company called Acornsoft, which was run by the author of Moon Lander for the MK14, David Johnson-Davis. Founded with Acorn’s blessing in 1979 to maintain a supply of software, Acornsoft would claim many obscure achievements – the country’s first ‘zombie’ game, for instance, which appeared on its second Games Pack soon after the Atom’s launch.

  Meanwhile, ZX80 coders had found ways to coax real-time graphics out of a machine that was already famous for blanking its users. Amongst these pioneers was a company called Macronics based in a suburban Solihull house. By skipping BASIC and talking straight to the processor using machine code, Ken Macdonald and Ron Bissell developed intricate timing and hardware tricks that made gaming possible. And they were interestingly open about their technology. ZX80 owners wanting to play Macronics’ primitive version of Space Invaders had a choice: it was available as a pre-recorded cassette or, for a pound less, as a sheet of paper with a code listing for the user to type in.

  Sinclair Research and Acorn were first into the consumer market, but by the end of 1980 other manufacturers were actively contemplating similar moves. It was an exciting, ultramodern industry, which had only seen huge growth, and there was every expectation that a giant market remained untapped. Home computing hadn’t yet ‘broken out’ – it was still a niche, mail-order speciality bubbling away in a corner of the public’s consciousness. So when the BBC announced that it would be choosing a single machine to use in a prime-time television series to teach computing, both Chris Curry and Clive Sinclair leapt on the news as the biggest
marketing opportunity their young industry had ever seen. And so did everybody else.

  If the decade of strife fell upon Britain’s industries unevenly, it’s fair to say that the BBC was amongst the more insulated. Predominantly based in London, and midway through the 10-year cycle for which its funding was set, it remained functionally independent of the wider, increasingly gloomy, economic climate. It was not untainted, though; for all its belief in its own lack of bias, on some issues it felt a self-conscious need to lead the nation’s agenda. Having raised the question of the changes heralded by the microchip, the BBC felt compelled to provide an answer.

  So did the government. Soon after the Horizon programme, ITV had run its own series, called The Mighty Micro, which had delivered more optimistic predictions, and between them the two broadcasters had spurred the Manpower Services Commission, an autonomous government-funded body, to fill the terrifying silence that Edward Goldwyn’s programme had pointed to.

  Apparently, Goldwyn had been right about the government’s ignorance, because the MSC went straight back to the BBC to help it investigate. It gave the Corporation some money to help with the budget, which arrived at the Continuing Education Department and fell into the hands of producers David Allen and Robert Albury, who put it to good use on a worldwide fact-finding tour.

  If it was a junket, it was an effective one. Most developed countries turned out to have relationships with the microchip far in advance of the UK’s – not only in the United States and Japan, but also Germany and Sweden. The BBC’s line changed from panic to ambition. Allen’s conclusion was that the country needed a public awareness campaign, not only about computers, but also how to program them. ‘If we wanted to democratise the technology, rather than be dominated by it as some people seemed to think, we needed people to experience it and to control it,’ he says. ‘And in those days that meant programming. It was very much a hands-on philosophy.’