Grand Thieves & Tomb Raiders (2 page)

He was also a clever boy, albeit by his own admission one who got by on ‘flair rather than hard work’. In the way that very capable children sometimes do, he hunted out pursuits that
suited a lively mind. And where these fell short, he invented new ones for himself.

Bartle was one of the first people to try the new hobby of role-playing games. These were very different from board games: they used conversation, adaptable combat rules, and verbal
improvisation to generate free-form adventure stories between groups of people;
Dungeons & Dragons
remains the most famous example. Bartle usually took the part of ‘games
master’ – a kind of writer-producer role, who established adventures for the other players to immerse themselves in. He spent hours creating stories and games, including many of his own
design. In one, the players took on the roles of
Victorian explorers traversing vast maps that Bartle had drawn on paper stitched together with Sellotape.

He played obsessively, inventing new games with his brother and playing them with school friends most evenings, but soon his appetite stretched further. ‘Play-by-mail’ games, where
gamers sent moves through the post to strangers across the country, were gaining in popularity amongst enthusiasts. While traditional role-playing games could run like improvised radio dramas,
postal games were more like playing chess by exchanging post-cards with your opponent, one move at a time. They were achingly slow, but the sense that there was an unseen opponent was utterly
compelling.

Bartle’s interest went deeper than simply playing: he created and published his own fan magazine about the games he loved. This wasn’t that uncommon in the emerging gaming scene,
where hand-made, photocopied ’zines proliferated. Even Bartle’s choice of title,
Sauce of the Nile
, followed the convention of using puns around historical or fantasy themes.
It had the colloquial feel of an amateur publication, but where other ’zines reviewed, commented and joked, Bartle was more ambitious: he used his as a platform for developing new games, and
linking readers across the country to play them.

The most successful of Bartle’s home-grown postal games was called
Spellbinder,
which he themed to his tastes: players took on the role of wizards, attempting to defeat each other
with spells. By now a veteran of playing board and strategy games, Bartle’s design allowed for emergent, complex gameplay, and players lapped it up. It was popular enough to be repeatedly
revised during the following decade, usually to make it more complicated.

And Bartle’s isolation in Hornsea had another, quite unexpected, benefit. The cabinet-sized computers of the seventies were usually the preserve of universities and large corporations and
it was common for schools to arrange some access for promising students on a visiting basis. But Bartle’s school was given its own subsidised phone line to connect to a computer located at a
nearby BP plant. Rather than having direct access to the computer itself, students could dial
up from a terminal in their school on a 110 baud modem – about a
millionth of the speed of modern broadband connections. Though sometimes a laborious process, for a curious pupil like Bartle it opened the door to programming far wider than would have been
possible in most of the country: ‘It was BP’s way of saying thanks to the community for letting them pump fumes into the air from their chemical works fourteen miles away,’ Bartle
says.

Inevitably, having learned to program, he applied those skills to writing games. There would not be a retail games software market for years, so his was published in a completely different form
– an ingenious role-playing adventure that ran as a single-player game.
The Solo Dungeon
foreshadowed the adventure gamebooks that would become popular in the eighties, but was more
complex – closer to the logic of computer instructions – and yet also quite charming. Its typewritten introduction opens with the unusually honest remark that ‘we believe that
there are no further errors to be found’.

It seems obvious now that the hobbies that were invading some British teenage bedrooms in the late seventies were uncannily suited to teaching computer game design, before any such discipline
existed. By the time Bartle sat his A-levels, he had absorbed the core principles, managed a community of players, cut his teeth on writing for public consumption, learned the foundations of
programming, published his first ground-breaking game and designed plenty more.

Given these demands on his time, perhaps it’s no surprise that the flair which so often substituted for graft wasn’t quite enough in his final mathematics exams – he had hoped
to go to the University of Exeter, but when his results came back in the summer of 1978, he found himself headed to Essex instead. In retrospect, this change of destination seems fated, because,
for what Bartle found there, he could not have been better qualified.

In 1978, Rob Trubshaw, a student at the University of Essex, made two fantastic discoveries. One was a lucky find. The other was an inspired insight into the mechanics of a
state-of-the-art computer
system, which worked around its careful security. It was a small, clever hack, but it created a completely new way of using the university’s
machines.

Trubshaw could most often be found in the computer science department, making use of the university’s giant DECSYSTEM 10 computer. This resembled a blue filing cabinet, with a panel of
switches which offered the uninitiated little clue as to their function. Even users normally couldn’t touch the DECSYSTEM 10 – it was kept in a controlled, air-conditioned room,
connected to terminals outside. Those terminals enabled multiple users to operate it simultaneously, sharing its computing power, although access was strictly rationed. This was the kind of
equipment that established computing as the province of boffins – a discipline where byzantine operations served an impenetrable purpose. Computer use was no longer restricted to scientists
in lab coats – anyone in the university could join the computer society – but to the layman’s eyes it was still dominated by a hallowed circle of experts.

Trubshaw was one such expert. He had started his degree in computer science a year earlier, and even within his department he stood out as exceptional. He would program the machine not only in
the common languages of the day – MARCO 10, BCPL – but also in its more obscure but powerful assembler language. Although as an undergraduate he was a junior user, with all the access
restrictions that such lowly status entailed, he was also one of the few people on the campus to own copies of the manuals for the system. He came to know the workings of the machine in a way that
most people never needed to, and found ways to use it that even the computer’s makers hadn’t imagined.

It was during a conversation with his fellow student Barry Scott that he devised a programming technique that achieved something previously considered impossible. He and Scott were pondering
ways to allow two users on the same machine to access each other’s work, but from different terminals. Normally this would be done by the first user saving work to one file, and the other
opening it afterwards.
The method was workable and safe, but slow and limited. And this was deliberate; affecting something that was happening on another terminal could lead
to damaging accidents, or even vandalism, and was intentionally locked down. From day one, the computer’s design had always been intended to keep users such as Trubshaw and Scott in their
boxes. But what these first year undergraduates discovered was a highly technical, but ingenious, way to completely bypass this limitation. There was a tiny piece of memory that was already shared
– if they could alter it, they would be able to communicate without special privileges. The two students raced through Trubshaw’s manuals, eager to be the first to find a loophole in
the system.

Scott won. He found an instruction that allowed two terminals to look at, and write to, this shared memory without the usual permission. The pair wrote a piece of code to prove their concept
– and when text was typed on one terminal, it appeared on another. They looked at their work in awe – it was a small step, but in the locked down, centrally controlled,
permission-oriented world of mainframe computers, this was a significant breach. For the first time, users could interact with each other.

Trubshaw’s other discovery occurred nearly a year later, and in a sense was far more mundane. As a curious programmer at a time when programming resources were scarce, he devoured examples
of code wherever he found them. To help, he wrote a clever program that kept tabs on every file that was saved on the university’s drives. Usually this turned up nothing, but one morning in
late 1978 it produced a list of files that Trubshaw immediately recognised.

He was an avid player of an early computer game called
ADVENT
, short for adventure. It had originated in the US, and was played almost entirely within universities. The game worked like
an interactive book: the text was written in the second person, advising the player on what was happening, and on what they could see. The player then entered simple commands – no more than a
verb and a noun – and the computer would write another paragraph describing the impact of their actions. Although a simple concept, it was applied with brilliant
diversity, producing a virtual world of compelling puzzles.

What Trubshaw’s virtual watchdog found were the source files for
ADVENT
. These differed from the program that most people used to play the game, which had been compiled into a
form the computer could read, but which scrambled the original programmers’ code. The source files gave the complete code listing as it had been written, with comments from the programmers
who had put it together. Trubshaw immediately realised that he had stumbled upon a coder’s goldmine, and within minutes the files had been discreetly copied to his personal tapes.

Although it was a joy to look at this famous code, it was also disappointing. As the game had become more popular, it had been expanded and developed by different people, and the code that
Trubshaw saw was, as programmers call it, a ‘kludge’: a functional but messy compound of different authors’ work. Routines intended to achieve one task were stretched beyond
recognition, and elegant ideas had ugly extensions grafted on to fulfil the ambitions of later contributors. Yet Trubshaw was encouraged by this: he knew he could do so much better. In that year he
had learned how to let two computer users talk to each other, and decided that he was easily qualified to write an adventure game. In between these two events, he met Richard Bartle.

Bartle arrived at Essex keen to write more programs, and needed access to the mainframe computer. At the time, the only way to get this was to become a member of the Computer
Society, and at first he didn’t even manage that, having missed its stand at the Societies Bazaar for new undergraduates. When he eventually tracked the society down, he found a ‘hacker
culture’ of kindred spirits, and amongst them Roy Trubshaw. ‘We delighted in programming for its own sake, and intuitively saw the power and potential that computers had to
offer,’ Bartle recalls. ‘It was inevitable that we would meet, just as it was inevitable that we would meet every other half-decent programmer at the university.’

Unusually, Essex University gave its computer society the
resources to use its precious computers for non-academic purposes, and hardcore users, Bartle and Trubshaw
prominent among them, could gain regular access. Over time they would have to fight off various attempts to restrict their access, but their right to use the computers for non-academic work
prevailed – ‘having fun’ was part of the university’s computing culture. The pair didn’t have the free rein they might have done at a cutting-edge US university, but
Essex was certainly as supportive.

Bartle’s arrival exposed Trubshaw to a gaming culture he had barely encountered before. Encouraged by Bartle, he had already started creating a text adventure game that used some of the
characteristics of
Dungeons & Dragons
by the time he discovered the
ADVENT
source code, and he already knew how to connect users in real time. It may have been serendipitous
that the pieces for multiplayer adventure gaming had fallen into place, but they had landed in the lap of an excellent programmer.

The pair set about building a multiplayer adventure with the energy of undergraduate students. Trubshaw’s time-limited access to the computers meant that he developed the code in his head,
and then wrote it down longhand – a discipline that made it concise and very efficient. Keying in code involved using punchcards – a system of recording programs through holes in index
cards – and correcting errors meant sending these to the computer in giant batches. It was immensely time-consuming, and throughout all of this Trubshaw had to complete enough of his genuine
studies to keep him from being thrown off the course.

But it worked. Trubshaw based his first tiny virtual world on his own house, and the
Multi User Dungeon
was born. Although Trubshaw and Bartle’s peers were used to text games like
ADVENT
,
MUD
had something entirely new. If two players stood in the same room at the same time, the description would tell each of them that they could see the other and, vitally,
their instructions would allow them to interact with one another. What had once been a game where a single player faced the challenge of a pre-programmed environment,
much
like Bartle’s
Solo Dungeon
, was now something far more powerful.
MUD
could facilitate competition or teamwork, enable mutually supportive or disruptive strategies, or simply
allow communication between players. It could be a social network.

At that time, though, it was yet to be any of these things. Bartle had been looking over Trubshaw’s shoulder while
MUD
was being written, but now he came on board officially, and
set about both building a full game, and harassing Trubshaw to expand the feature set. He had an instinct about how role-playing adventures should work, and the tools needed to make them happen on
a computer.