Grand Thieves & Tomb Raiders

by Rebecca Levene

  Although at first counterintuitive, in the right hands vast worlds clicked with the 8-bit generation of computers. The fixed hardware meant that coders could ferret out every last byte of power, and although the graphical tricks could be jaw-dropping, a palette of lo-fi images meant that some of the repetition needed for scale was forgiven. But they also worked because home coders enjoyed technical challenges, and for logical, creative minds, stretching the limits of home computers was a compelling pastime. For them, each new breakthrough simply goaded them on.

  Peter Cooke, the teacher who had written Invincible Island, had been as impressed by Elite as everyone else, but he was really interested in playing with the technology. He found himself wondering at another game, Gyron, published by Firebird for the ZX Spectrum a year later, and written by the same team that was converting Elite for the ZX Spectrum. It was a first-person maze game – hypnotically pretty though quickly dull – but it did feature an extraordinary graphical trick: giant, solid spheres roamed the maze. Elite had included solid suns, but they had visibly slowed the frame rate. The speed at which these spheres were drawn would normally need a series of pictures stored in memory, and there was nothing like the space for that in a 48 kilobyte computer.

  Cooke was keen to make ‘solid’ 3D graphics work, but it seemed that the hardware wasn’t up to the job. ‘With a 4Mhz Z80 it was impossible to do full 3D using points and surfaces with a decent frame rate,’ he says. Eventually, he hit on the idea that shapes didn’t need genuine 3D calculations, but instead used a table of pre-calculated data which was scaled up or down with the distance of the object.

  Whether or not this was the trick Gyron had used, it worked, and Cooke kept extending it. ‘I tried adding a light-dark shading,’ he says. For any shape with a vertical line of symmetry, a shading effect could be dynamic. If the drawing routine changed colour a consistent fraction of the way from left to right, say at twenty per cent or fifty per cent, it would look like it was properly shaded. For a sphere, or a tower, or a giant robot shaped like a chess piece, the gradual shift in shading would look like a light source drifting around it. ‘When I first had the shading code working I took it in to show the lads in the computer club, and they were very keen,’ Cooke recalls. ‘So I could see it had potential.’

  He started thinking of game scenarios that could use his new trick. Elite included three axes of rotation, but Cooke’s technique could only accommodate one. Happily, this fitted neatly with a scenario that gamers had been hoping for since Bell and Braben’s game had been published: ‘The game had to be set on a planetary surface,’ says Cooke.

  It would use the lighting routine to portray a day and night cycle, so he needed a planet with a sun. He chose to invent a satellite of Tau Ceti, a real star and – in astronomical terms – a close neighbour of Earth. It gave his creation its name and as Cooke saw it, Tau Ceti was going to be another big game like Elite. The player would fly a laser-armed craft at ground level across a vast planet, studded with cities and sprinkled with enemies. There would be a mission, with puzzles, clues and an unfolding story. And it would be huge.

  Bell and Braben had procedurally generated their universe, but Cooke wanted his cities to be designed by a human hand to create a compelling adventure. This was far beyond the capacity of a 48K ZX Spectrum, and Singleton’s trick of using a small amount of data for each co-ordinate simply wouldn’t be detailed enough. So Cooke used a different technique, in which the original data is mathematically squashed into a smaller space, and then unpacked when it’s needed. City by city, the planet would unfold before the player. He had literally given them a world to explore.

  Games like Tau Ceti and The Lords of Midnight marked a shift in the focus of the medium. From its launch, the BBC Micro had been a better computer that cost more, the natural home of landmark achievements like Elite and Revs. Now the real developments were happening where the gamers were, on the ZX Spectrum, with the Commodore 64 and the Amstrad CPC not far behind.

  And Tau Ceti was one of the first of many games with solid, 3D graphics for the ZX Spectrum. Eventually a system called Freescape would allow complex 3D shapes and scenes – it became available for the Commodore and Amstrad machines too. But the BBC Micro was left out; its gaming market fell away as its rivals built theirs. With games like Elite, it had moments of glory. But they were only moments.

  Revs, The Lords of Midnight and Tau Ceti were deep, expansive games that stretched the ambition and the state of the art of the industry. But they weren’t Elite – the overwhelming behemoth that had shaken the medium, and ballooned its scope from petty entertainment to a social-life-devouring universe. How could they be? Elite had claimed so much ground, fused so many technological and gameplay innovations, that the next step, however large, could surely never match its shockwaves.

  But there was a game that might have done. Elite had captured the thrill of open world gameplay and autonomous exploration, but the core activity of besting other spacecraft in combat remained essentially the same throughout the game, even if it was blissfully rewarding.

  Elite didn’t have one important feature that would become a hallmark of the open world genre: a universe filled with autonomous beings – who have rules governing their behaviour, but can also act quite independently, and most importantly, can interact with each other. As these elements are brought together they can create new situations that even the game-makers might not have considered. Later versions of Elite on more powerful platforms did feature something like this: police ships flying in formation, breaking off to attack pirates who had set upon a passing trader, all without the player’s input.

  But the game that really introduced these ideas arrived in the dying years of the 8-bit era. It is often overlooked now, but revered by the gaming cognoscenti for so comprehensively realising one of the medium’s finest innovations: emergent gameplay. It’s called Exile, and it’s what the other two boys from St Albans School made, after Elite had made Ian Bell rich.

  In 1985, Peter Irvin and Jeremy Smith each returned from university with a game published and no interest in pursuing their studies. Smith’s expertise was in modelling physics, for which his game Thrust had gained widespread respect. Irvin, meanwhile, had been working on a 2D, side-on and rather linear portrayal of a wizard heading down a randomly generated passageway. Irvin had no idea where either the passageway or the demonstration was going to go – like many home-developed projects it was an experiment to entertain the programmer.

  The two of them decided to combine Smith’s physics with Irvin’s landscape engine. Both would be rewritten endlessly over the two and half years the project would consume, but they provided a canvas, and the pair started sketching. There was no story at first, just ideas piled onto the player’s avatar and his world. He was issued with a jet pack, countered by the planet’s realistic gravity and, in the game’s first achievement of many, objects were made to collide with a momentum exchange, bouncing off in a way that felt satisfyingly genuine. ‘Exile might have been the first game with a complete physics environment engine,’ says Irvin, cautiously, but not having learnt anything to the contrary after twenty-five years.

  The story came together gradually along with everything else. The player controlled Mike Finn, a jet-pack-wearing space commander on a rescue mission to an isolated planet, Phoebus, where exiled geneticist Triax is holding hostages. It sounds hammy when summarised, but it all fitted their technology rather well: a side-on adventure, a physics engine, particle weapons and an interactive environment. They also hit on an effective conceit to keep it fun for the player, and allow them to explore – Finn’s suit would detect when he had taken too much damage, and teleport him back to the last place saved by the player. Suddenly the game opened up. Any experiment in this playground was worth a punt – the worst that could happen to the player was to be yanked away from the danger.

  Their first job was to create a landscape, a giant space vast enough for their plans. ‘It’s a massive map – there�
�s no way you could fit that into memory in the way it is normally,’ says Irvin. So, as with Elite’s galaxies, the pair let a routine feature the map. This was a more complicated proposition, though, as their map needed to feature a coherent, usable set of connected caves and tunnels. Irvin and Smith generated hundreds before settling on one that would become the obsession of the developers and players for years. It won by being coherent, and having a very large cave near the start.

  But the star innovation of the project wasn’t the map, or even the physics, although both were vital. Their world was populated. It was filled with semi-intelligent creatures with their own plans and territories, ready to interact with the player, or to defend themselves. The game world was alive.

  Irvin and Smith developed an emergent ecosystem, with different species that related to each other in complex ways. Birds followed the player and ate bees, bees swarmed and stung passing animals, monkeys stole objects and knocked things over, and robots pursued or protected the player according to their programming. In a sense, these things had all been programmed, but in subtle ways that, to the player’s eyes, made them seem autonomous. The animals appeared to have ongoing lives that didn’t require so much as a keystroke from the gamer.

  ‘It wasn’t an accident,’ says Irvin. ‘It wasn’t artificial that the animals had certain behaviours towards each other.’ He and Smith assign attributes to each object or creature: a type one baddy, a robotic baddy and so on. Certain groups were hardwired to hate or love other types. But separate to this, each had strategies or tactics. The program included very sophisticated – for its time – line-of-sight vision. Animals couldn’t see round corners, which was handy for the player, but they could spot one another. And from this their behaviour would emerge.

  Take the bees. They were programmed to like other bees, and so would normally circle around one another. Every few seconds, though, each of them would have a rethink, and if one saw something else that it liked or wanted to attack, it might pursue it. The other bees would then follow, perhaps themselves locking onto this new object. From simple rules a complex swarm intelligence emerged, and it was uncannily realistic. ‘You start to see how the natural world might work with these very basic programming ideas,’ says Irvin.

  The two creators had been rigorously rewriting and testing each other’s game code and were confident that towards the end there were no errors in it. But the adventure world they created was a different matter. It was perfectly plausible for the creatures to all obey their instructions and yet create impossible situations for the player, or for one species to become dominant over the others. The pain Smith and Irvin had saved from debugging the code would be spent on debugging the world.

  Balancing the game was vital if they were to give their audience the sense of total freedom they had planned, while making sure that players couldn’t find some shortcut or way of exploiting the world. ‘You weren’t meant to feel railroaded through a route. It was: there you are, there’s your planet to explore,’ Irvin says.

  The narrative worked with the game world too. Rather than filling the story with memory-hungry ‘scripts’ for the animals and robots to follow, they were instead put in situations where the game’s rules would guide them, and challenge the player. For instance, the final challenge for defeating Triax didn’t involve killing him, but trapping him between two teleports that kept the villain in an eternal loop. Even that wasn’t a watertight solution: he was intelligent enough, or at least randomly curious enough, to occasionally push his way out again.

  The absence of artificial direction helped give the player a sense of freedom. ‘You weren’t quite sure whether you were doing the right thing,’ Irvin says. ‘But eventually you might open a particular door to access another area of the cavern, so you kind of knew you were succeeding.’

  In 1987, after two long years of development, they were ready to approach publishers. Jacqui Lyons brokered their deal with Superior Software, which had bought out Acornsoft’s games catalogue. Its managing director Richard Hanson knew that the pair had something special. The publisher positioned Exile as a premium product: it had a pack-in novel commissioned from one of Irvin and Smith’s school friends, and a teaser campaign promoting it as a landmark title.

  As if having a name with only one letter different from Elite wasn’t a strong enough signal, Superior Software pulled in David Braben for an endorsement on the box. ‘He offered it,’ says Irvin, ‘and then regretted it later, tried to pull out of it.’ Braben did keep to his word, though, and his quote appeared in a yellow splash on the advertising. It was another sign that the circle of Cambridge programmers was still very small. ‘If it hadn’t been him, it would have been Ian.’

  Exile had a big launch in the autumn of 1988, and the reviews were universally positive, but to the writers it was a letdown: developers could expect little feedback from gamers at the time. The truth is that the reception was probably mixed. It was a complicated game with a steep learning curve that some players would find impossible to like. The first weapon was only found after a couple of hours gameplay, and the puzzles didn’t have smooth, clearly signed solutions – anybody looking for a quick fix would loathe it.

  But to its fans, it really was a masterpiece. The combination of the physics and the wildlife was wonderfully compelling – hours could be lost fighting and strategising around the other inhabitants. Some of Exile’s greatest admirers never progressed particularly far with the game – it was enough to dive in and watch the world unfurl, occasionally tossing in a grenade to see what happened. A review for a conversion in the first issue of Amiga Power highlighted the different reactions. The main reviewer had immersed himself for days, and gave Exile 89% – as high as a title brought over from an ancient machine was ever likely to achieve. But there was also a sub-review, a small boxout from another playtester that acted as a brief sanity check. He was far more circumspect: nice physics, but can’t see what the fuss is about.

  Elite is famous, but the odds were always against Exile achieving that kind of recognition. It came out at the tail end of the 8-bit era, on a computer that had lost its momentum as a gaming machine. Peter Irvin has stayed in the industry – he was one of the programmers for the Elite sequels. Jeremy Smith sadly died not long after the conversions were finished. The impetus for a sequel stopped then, and has never picked up again.

  Irvin is still tremendously proud of Exile, though. He even, grudgingly, credits the punishing hardware restrictions for the success of the final game: ‘If the memory had been ten times as big, it would have been finished in a quarter of the time. And it would have been a tenth as good.’

  Exile is a gamers’ game. The industry favourite magazine Edge, known for its detached scepticism, ran a special review of it more than a decade after its release, in which it was awarded a rare top score. The magazine had given out only one other retrospective ‘ten’, for Elite.

  It’s almost lazy to say that brilliant programmers cajoled amazing things from tiny boxes in the 1980s. But they did – curious, restless inventiveness pulled astonishing feats from the machines and squeezed worlds of implausible size inside them. It’s partly a reflection of the way bedroom coders worked. They didn’t have the limits and pressures of corporate targets or expectations, and if they wanted to try something directionless and experimental, or monstrously time consuming, or apparently impossible, they did. It set a pattern for the British games market, where a landscape of conventional games was studded with some truly incredible ones. And those were almost always home-made titles, because games needing years of work with untested technology were not projects that a commercial developer would be inclined to commission.

  But this hobbyism had a more far-reaching legacy. The scope of games changed in the hands of bedroom coders – they were in a different league, and had different ambitions to arcade titles. The pace slowed and the scale increased, and this all happened using the technology that they had, rather than waiting for a generation that would come
. Rival markets using consoles, or more powerful computers with disc drives, could attract gamers with more content in conventional settings. For Brits, it was taken as a challenge to fashion giant, novel experiences using primitive tools.

  A few of these games were directly inspired by Elite, but all were under its shadow. In the decades that followed, Elite would be a yardstick for technical advances, a byword for an industry changer, and a fallback answer for the most important British game. Its innovations, accompanied by some of the ideas from other games of its time, can be traced through the generations that succeeded it to some of the bestselling franchises in the world. But even in the eighties, Ian Bell and David Braben were held in a special reverence. They showed that while a legion of home coders had created an industry, individuals could revolutionise it.

  6

  Technical Failures

  At the height of Acorn’s success, Chris Curry went missing, feared dead, and the company’s share price tumbled.

  It was October 1984, and Curry had returned to the Conservative Party fold. He was an admirer of Margaret Thatcher and the drive for entrepreneurship that she championed. And the Tories were rather brazen about asking for his money: ‘It was quite blatant in those days,’ he remembers. ‘They said the Tory party is the party of sponsorship – you tell us what you want, you give us money, and we’ll make it happen.’ So he had been invited to the party conference in Brighton, and was staying in room 426 of the Grand Hotel, four rooms away from the Prime Minister, on the night that the IRA bomb exploded.