The Unexpected Benefits of Making Your Own Tools

I Invented Not-Invented-Here Syndrome

In 1996, I was working at a little games company that shall remain nameless (but later became famous for creating an amply-endowed female protagonist fond of shooting small animals while plundering historical sites).

We were working on a rather less successful game that nevertheless had a full 3D engine (one of the first — we wrote our own perspective-textured, Gouraud-shaded polygon renderer that ran in 11 cycles per pixel on an 80286, for those impressed by such things) with a fully articulated inverse-kinematics-based character system. There weren’t really any tools out there to design characters or to animate them, so we had to make our own. And that constituted probably 95% of the development effort, and in fact its core delight; for me at least, the game itself was just the pretext.

One lunchtime I was messing around with the walk-cycle editor (all true innovation happens around lunchtime or weekends) just as one of our artists dropped in with a couple of friends. We lit up a joint, as only artists could with impunity in our realm, and started to mess around.

To demonstrate our new animation system’s flexibility, I rotated the walking test robot’s pelvis forward 90 degrees, and suddenly he was swimming; then, on a whim, I freed up his pitch control and connected it to the keyboard, and suddenly he was flying. We dropped in a trippy clouds-and-skies cubic environment map, fired up some ambient music, and five minutes later we had a little chill-out flying game.

“That would work for a chill-out room,” said one of my artist friend’s friends, who just happened to have a day-job projecting abstract visuals onto giant screens at r̶a̶v̶e̶s̶ live music events, “but I don’t suppose you could make him dance ?”

It just so happened that, with the help of a couple of sine-wave generators and a little string and glue, I could; and so, the following weekend, I found myself at a New Year’s Eve event at Alexandra Palace, making a virtual robot dance on giant screens for a crowd of 50,000 people.

“This is more fun than games,” decided 24-year-old me, and not very much later, when I inevitably got the sack from said games company (it was a blessing for all concerned), I took a long hard look at the pile of job offers from the games recruiter, tossed them in the bin, and started my own company. Two weeks later, I’d won my own first gig — making a virtual robot dance on a large television screen for a crowd of 20 people at a trance rave in Tunbridge Wells, for a fee of just £20. And the rest, as they say, is history.

The point : having complete control over my own tools led, completely by random accident, to a complete change of career. I moved from an industry that bored and frustrated me (games), to one that enthralled and delighted me (entertaining crowds of chemically altered idiots), while continuing to do the thing I loved most (build real-time 3D engines). And all because of that five-minute lunchtime messing-around session.


One thing led to another, I prospered, and I eventually found myself in 2002 learning how to DJ (and writing software to let me DJ without the faff of lugging records around, beat-matching, or looking cool), then joining a company started by a pair of friends, building a visual show for a then little-known band by the name of Massive Attack.

And not just any show — this was a new kind of show, one that would morph every day to reflect the real-time nature of the world of information (this was cutting-edge thinking in 2002) — today’s news headlines, stock prices, weather, air traffic, messages from the audience, show location, even a clock counting down to showtime (extremely effective, if never quite accurate). All in pinpoint sync and pixel-perfect digital clarity.

How did you build a show like that, in 2002 ? The answer was, you didn’t. You could try conventional tools, like AfterEffects, but you’d fail (you couldn’t possibly re-render the entire show every day, not on the computers of the time). You could try edgy real-time tools like Macromedia Director, but you’d fail (too complex, too slow, not malleable enough, not controllable enough, not rhythmic enough).

So we built our own. It was called Mosquito; it was stupid simple, Frankensteined together out of bits of the engine I’d built for dancing / DJing robots, three months from first line of code to first show, without the aid of version control, bug databases, or even post-it notes; and it worked. The show was a success, and it put us on the map.

The point : having complete control over our own tools meant we could produce a ground-breaking show with only three people, in a ridiculously short time. Whenever we found ourselves repeating something, we’d automate it, then turn the dial to 11 and fall about with amazement. We built only what we needed to get the job done. Totally useless for anyone else; perfect for us.


One thing led to another, we prospered, and eventually (in 2005) who should cosy up to the bar than a slightly better known band by the name of U2. This time, the challenge we had was to show the client how the content we’d made looked in three dimensions, for which, it became clear, no effective tool existed.

So we built our own (I said there’d be a theme). I sat down over the weekend and knocked together a simple 3D visualiser, and called it Dragonfly 3 (Dragonflies eat Mosquitos, you see). It was stupid simple, three months from first line of code to first show, and it saved our ass. U2 ended up using the new software on tour to drive their visuals, and it put us on the map.

Perfect for us; totally useless for everyone else. Or so we thought.

Turns out having U2 on your CV attracts attention. One thing led to another, and eventually dragonfly3 turned into d3 which turned into a product which turned into a company called disguise, and here we are today. If you’d told 24-year-old me that 47-year-old me would be running a goddamn media server company, I’d have looked at you like you had two heads.


Here’s the thing about tools made by other people for other people : they’re not made for you. They’ve got a ton of stuff you don’t need, which clutters up the interface and slows everything down; but the stuff you really need, they don’t have (or it doesn’t work quite right), which slows everything down even more. Make your own tools, and you get to make just exactly the thing you need, and nothing more.

Here’s the other thing about tools made by other people : they limit your distinctiveness. You’ve got the same moves and affordances and constraints that everyone else using the tool has. Make your own tools, and you can do things nobody else can do. And that makes you unique.

And that, in turn, pushes you out into the white water. Which is the right place to be, if you want to go places.

I once made the mistake of letting other people use my software; the result was Now I’m trying to figure out how to fix what’s really broken.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store