Backroom Boys

by Francis Spufford

  And it was those competences that would be called on to manage the task of making something permanent out of the early 1980s moment of opportunity in computer games. For the ability to start small didn’t abolish the perennial British problem of finding a corporate form to do justice to technological creativity. The era of the home-programmed game was just a phase from which an ability to thrive on a larger scale had to grow. Elite cost Acornsoft very little to develop: some Acorn hardware for the boys, fees to Rob Holdstock and Philip Castle, and the printing and promotion bill. But games companies would soon no longer be able to count on the fact that (as David Braben put it) ‘the burn rate of a student is zero’. Some innovations have eased the process of writing games. The existence of compilers means assembly code doesn’t have to be laboriously composed by hand any more, for instance. ‘Middleware’ outfits will rent you their rendering engine, or their artificial intelligence system, so you don’t have to write your own for your game. But the cheapening innovations are enormously outnumbered by the expensive ones, which are all necessary if a company is to keep up with the relentless improvements in the state of the art. Games are now written by teams of coders, animators, artists and level designers, all on salary, all using incredibly specialised professional software packages it may cost hundreds of thousands of pounds to licence. To bring a game of professional quality to market now costs in the millions; and therefore the market has to be global if you’re to cover your costs. Then, there is no guarantee at all of the income a game will generate. Producing video games is a hits business, which means it abides by William Goldman’s rule for Hollywood movies: ‘No one knows anything.’ No one can say which will succeed, which won’t. When your products are this unpredictable, you do anything to reduce risk, which is why games companies, like film companies, produce so many sequels: they desperately milk any formula that seems to offer above-random returns. But ultimately the only way to prosper is to have enough bets on the table at any one time so that the winnings on the few hits cancel out (or, preferably, more than cancel out) the losses on all the rest. To keep your head above water year after year in the video-game business, as in the film business, you need something like a studio system.

  Which is why the outlook for the British games industry is uncertain. It employs about 6,000 people in 270 different studios – or many more, if you count all the shop staff who sell computer games, all the magazine journalists who review them, and all the PR types who arrange the vodka glasses at console launches. It turns out a steady stream of games as remarkable as Elite, as jaw-dropping, as genre-defying: Populous, Lemmings, Creatures, Tomb Raider, Theme Park, Black & White. It has produced a self-sustaining cluster of expertise down in Guildford that looks like a textbook illustration of how a relative advantage should be maintained. It continues to take an astonishing, disproportionate, unadvertised share of video-game markets around the world. (For example, in 2000, British studios authored 35 per cent of the Playstation games sold in the UK, 33 per cent in continental Europe and 11 per cent in the United States.) But it is still an open question whether the British industry is going to find a corporate expression for all this creative strength, or whether we’re just going to supply idiosyncratic artisans to the world. In the late 1990s, it looked for a while as if Eidos might attain the critical mass to operate as a kind of British MGM for games. But it turned out to be overdependent on its central asset, Tomb Raider, while its network of developers failed to generate other hits of the necessary size. It tried to merge with one of its international rivals, was spurned, and is now a faded star. Other powerhouses of the mid-1990s – Bullfrog, Psygnosis – were snapped up by French and American studios. It remains to be seen whether Rage or Warthog or Mucky Foot or another of the next generation will try for the scale required to compete with the global giants.

  *

  Meanwhile, in a Britain more violent, more ecstatic and more brand-named than it was when the rhythms of industrial manufacturing governed it, the makers of Elite have scattered. Acornsoft’s boom year in 1984 coincided with a terrible miscalculation by Acorn about the size of the Christmas market for the BBC Micro. By the spring of 1985, Acorn was owned by Olivetti, and the software division was out on a limb. ‘They didn’t seem to know what to do with us,’ said David Johnson-Davies. The BBC Micro itself succumbed when the IBM PC with its Windows operating system became the global standard, a development Hermann Hauser and the other Acorn chiefs hadn’t foreseen, in common with almost everybody except Bill Gates. (‘If we had, it might have been a BBC Micro world by now,’ he said ruefully when I interviewed him.) The lesson was learned in ARM Holdings, the offshoot of Acorn established in the late 1980s to licence Acorn’s chip technology. ARM’s chips for mobile phones are the global standard. ARM operates from the building near Cambridge airport that Acorn moved to when it outgrew the labyrinth at 4a Market Hill. David Johnson-Davies and Chris Jordan both run technology companies in Cambridge that have nothing to do with games.

  David Braben and Ian Bell grew apart. ‘It was a positive feedback loop,’ Chris Jordan told me. ‘They diverged a little bit, and then that made them diverge further.’ They spent much of the 1990s in legal dispute with each other. David Braben used the rewards of Elite to build himself a career in the games industry on his own terms. He is a businessman now, with a development company of his own just outside Cambridge. He worries about the Euro, and hopes to create games still bigger than Elite. At the moment, he’s working on a Wallace and Gromit game. Ian Bell lives quietly in the countryside with his girlfriend, a vet. He used the rewards of Elite to study Aikido and to get into the rave scene. He breeds pedigree Burmese cats and worries about American imperialism and Third World debt. He does a little exploratory coding now and again, but he doesn’t play modern computer games: too obvious, too violent. He tries to keep out of the stream of greed-inducing images. He doesn’t read fiction much either. Like the intelligent horses at the end of Gulliver’s Travels, he thinks it only says ‘the thing which is not’. He doesn’t much like the world he helped to create.

  * He had had them for about a year by then, but a year is a long time in an emergent technology, and a long time when you’re twenty.

  Four

  The Isle Is Full of Noises

  Eeeeeeeeeeeeeeee …

  Three men in a van, stuck in the dense lunch-hour traffic at Hyde Park Corner, late in 1984. Then a gap opened in the lane to their left, and the one of them driving slipped the van deftly through it and pulled away up Park Lane. ‘17 mins 30 secs,’ wrote the one in the back: ‘Point B. Turned north.’ He was calculating their location by dead-reckoning, as if the van were a ship and he were an old-fashioned navigator. The dinky little modules of the Global Positioning System weren’t available yet, so if you wanted to create a timed log of the route you took, this was how you had to do it. A counter on the back wheel of the van recorded the distance they’d gone. Later, he’d be able to put that information together with his notes of direction changes, and then they’d know pretty much where they’d been at each minute of the journey. The driver accelerated a bit, but not a lot: just to a steady trundle which infuriated the vehicles behind. Every few seconds someone overtook, delivering silent-movie insults and gestures as they passed. On the left the bare trees of Hyde Park slid by; on the right, the lobbies of the grand hotels spilled metallic gleams of luxury into the wintry air, gold and silver and bronze.

  The third man in the van was watching the output of the instrument connected to the twelve-inch antenna on the bonnet. Across the park about a mile away to the north-west there was a transmitter up on the roof of a tower block, the solid but less grand Royal Lancaster Hotel. In the frequency band shortly to be put in use for Britain’s first mobile-phone calls, 900 megahertz, it was sending out a continuous test signal. The antenna on the van was shortly to be the standard model fitted to cars; it was connected to what would shortly be the standard in-car transceiver, only this transceiver was rigged to make a continuous r
ecord of the test signal’s strength. Twenty-eight decibels, 30, 31, 27, 25: the signal strength never settled. It was always hopping about, always on the move, affected by the environment in ways they couldn’t always fathom yet. The third man’s job was to annotate the recording with observations of anything in the environment that might be a factor. ‘Tall buildings to right,’ he put down, being careful but not expecting the wall of luxury over there to have much impact. It was on the other side of the van from the transmitter. Hang on, though: the signal strength was going up. Thirty-two decibels, 35, 37. What the hell was happening now?

  If you send a signal down a wire, what happens inside the wire is so predictable that you can effectively ignore it. Wireless is different. The ‘air interface’, as radio engineers call the stretch of open space between a transmitter and a receiver, is not a neutral medium. It has qualities. It is an invisible ocean, with the ground as its ocean floor. Slow, coiling currents flow in it when the wind blows and alters the electrical resistance of the air. Waves of every length in the electromagnetic spectrum roll across it all at once. At the tiniest scale, it is pierced through with little voids where the radio waves cancel each other out, so that a truthful picture of it would have to look like the Sea of Holes in Yellow Submarine. The ground is its ocean floor; and as the ground rises and falls, the ocean has reefs and shallows and breakwaters where waves of different lengths are impeded to different degrees by different obstacles. The wave of a 900 MHz mobile-phone signal is about 30 cm long. Therefore, a 900-MHz signal that runs into the side of a building hits a surface which is much larger than it is and comes bouncing straight back off it, like a wave of water coming straight back off a flat sea wall, a little less energetic but still the same wave with its motion reversed. On the other hand, if it hits something about the same size as itself, like a tree branch or a pillar box, it bursts apart into electromagnetic foam. It was a wave reflecting back off the marble wall of Mayfair that made the signal strength rise unexpectedly on Park Lane. That was the kind of thing the men in the van were driving around to find out. They were learning the ways of this new sea.

  The basic theory of mobile radio had been invented in America nearly forty years earlier. In 1947, D. H. Ring and W. R. Young of Bell Labs published a memorandum describing a neat way to get around an apparently impossible problem. Anyone could see how handy it would be if radio could become a mass tool, like the telephone, used by millions of people to send and receive messages; but there just weren’t that many separate frequencies available in the radio spectrum, no matter how finely you sliced it, no matter how much of it governments released from military use now the war was over. The spectrum was finite. Ring and Young showed, nevertheless, that you could persuade a small number of frequencies to support a much larger number of users. The trick was to forget everything that had just been learned during the war about very long-distance radio communications, and to think small. You divided up the land area you wanted to cover into a multitude of little ‘cells’, each of which contained a short-range low-powered radio mast – a ‘base station’ – that could handle the traffic to and from all the radios in that one cell. You could then reuse your limited set of frequencies over and over again in different cells, because no one ever communicated directly from one cell into another one. The base stations were connected up by land lines. The only condition was that no two adjoining cells should use exactly the same frequencies, or there might be interference across the cell boundaries.

  But that was easy to ensure. You just split up your total number of available frequencies into, say, seven sub-sets, and arranged the cells on the ground in groups of seven so that no two cells with the same sub-set ever touched. It would be like tiling a kitchen floor in seven colours, or making a patchwork quilt with a repeat pattern in it. Ring and Young imagined the cells as hexagons, because the honeycomb regularity of hexagons made it easier to work out the repeats. In any one cell, then, there wouldn’t be very many frequencies on offer, but there’d be more than enough to cover the number of people who felt like making a call just then in that one small area. And there was no limit on the number of times you could repeat the pattern, so there was no limit on the number of cells you could have. Multiply the cells sufficiently, and this was indeed a system that could support tens of thousands, hundreds of thousands, millions of users at once.

  It was a lovely piece of lateral thinking. Unfortunately, building the system in practice depended on two crucial bits of automation. Every time someone initiated a call, the nearest base station had to be able to assign them the next free frequency it had available. Every time someone happened to cross a cell boundary in the middle of a call, the call had to be ‘handed off’ to the next base station and switched over to a new frequency without the caller even noticing. These both required computing power of a kind that wouldn’t exist for decades. The Bell company experimented with cellular in various small ways. Passengers on the Metroliner express train between New York and Washington enjoyed making novelty calls at 100 mph from a special Bell payphone. But serious development had to wait until the end of the 1970s. Till then, the commonest form of mobile in the world was taxi radio, a rough ’n’ ready system in which ten cars all on the same frequency made crackly progress round a city under the reach of one simple mast, with a human operator switching the calls on and off.

  Kkkhhrrr … Two kilos to pick up, John, at North kkhhrrr. Do you copy, John? … Say again, please, control. Say again, over? … Kkhhrrr …

  The men in the van worked for Racal, pronounced ‘Raycle’, one of the military electronics companies that had been left standing when the first Thatcher recession shredded all the consumer electronics manufacturers in Britain. But Racal was a defence contractor with a difference. It had begun as a breakaway by a bunch of Plessey engineers, and it deliberately went against the traditional cost-plus way of doing military business that held at Plessey – and at Ferranti, GEC and British Aerospace too. Instead of winning a government contract, building a product without worrying much about the cost and then charging the government the cost plus a profit margin on top, Racal designed products to a price and then sold them aggressively all over the world. Racal didn’t get contracts to build radios for the British Army, but it designed a backpack field radio that Arab sheikhdoms and African dictatorships could afford. Where other field radios used expensive metal casings, Racal’s radio used plastic, sprayed with a thin metallic film. The competitors would have liked to believe that Racal’s stuff was cheap and nasty. Plessey got hold of a Racal field set, a former Plessey engineer told me. They fiddled with it: it still worked. They cooked it, they froze it: it still worked. They took it up to the top of one of the two towers at the Plessey HQ in Ilford and threw it off. Now its corner was crushed in. They turned it on. It still worked.

  Racal lived and died by its sales, not by having an intricate webwork of connections to the procuring departments of Her Majesty’s Government. Paradoxically, this meant that Racal had exactly the right credentials in 1982 when the Thatcher government invited applications for two licences to run Britain’s first ever public mobile telephone services. Two licences, please note. In every other European country, the incumbent state phone company got a head start at mobile. There were going to be two networks from the outset in Britain. The Thatcherites didn’t believe in industrial policy but they swore by competition policy. They wouldn’t intervene to make things happen, or to offer goals to industry, but they’d step in purposefully to disrupt any set-up where they felt the magic of the market wasn’t getting its full chance. One of the licences was automatically going to go to British Telecom, that was understood. But the other was up for grabs, and the criteria in the application document were gloriously vague, this first time round, what with the whole idea being so new and untried, leaving the government free to pick a sympathetic candidate, not just an adequate one. Racal was not an obvious choice. It had no experience of customer service (unless the customers were Third World armies).
Whoever won would have to invest tens of millions of pounds in their network before they saw a penny of revenue. Rumour has it that Gerald Whent, the managing director of Racal, might not have bid at all if he hadn’t heard a whisper that the government would look favourably on him. Ferranti, Plessey and GEC were all in the running, each with the commercial partner the bid required; but when the winners were announced on 16 December 1982, the British Telecom-Securicor bid had won one licence – and Racal and its US partner Millicom had the other. Thatcher’s government preferred nimble and piratical. BT’s service would operate as ‘Cellnet’, Racal’s system for voice and data would be called ‘Vodafone’.