Stephenson's Racket: Why the Snow Crash author wants to crowdfund CLANG

It wasn't one big light bulb going on - I've been interested in this idea of sword fighting in a game environment at least since I wrote Snow Crash, which is over 20 years ago. I sort of thought that it might have been done by now.

There are a lot of perfectly good reasons it hasn't been, having to do with the state of hardware and the way that technology works. In the last couple of years it feels like things have been changing. Hardware has been changing in an interesting way; there's been a whole movement around the revival of historical Western swordsmanship in the last decade or two that has given us much more information about the details of how people used to use these weapons.

Putting that together with the potential of crowd funding, we thought we might have a chance of actually making this work.

I think it's a matter of getting the right people to do the right job at the right time. Starting something from scratch, thinking about it and working on it in the background for a few years, coming up with ideas and experimenting with them, those are all things that I feel I can do in my spare hours when I'm not actually sitting and writing.

I can't be the head engineer on a serious videogame development project. So there has to be a point where a line is drawn and we say, okay, we've learned a hell of a lot from experimenting and thinking and talking to people. We're not going to learn a hell of a lot more from this approach so it's time to turn this into a real engineering project. In order to do that we need to get real engineers on it and I need to move to the role of a consultant, an advisor, and let them get on with what they do.

Zooming way out to the big picture view: some writers spend their whole careers producing only prose, which is a perfectly fine thing to do. But many writers, at one point or another, engage with other media. Over the last few years this has typically meant film. This has become a well-trodden path, and a reasonably respectable thing to do - I always think of William Faulkner in Barton Fink.

The question posed by the rise of the game industry is whether, and how, writers can engage with that medium as well. In some respects it's a better fit, for a novelist, than film. Novelists - especially fantasy and SF novelists - tend to be world-builders, and modern gamers have come to expect not just entertaining gameplay but a fully realized world that makes the gameplay relevant to some larger story or theme.

My own instincts run closer to the game world than to the film industry and so I've been thinking about how to engage with games for a while. 2012 seems like a good time to have a go at it.

As to how this fits with my overall career as a writer, I would just caution people to avoid swinging to either extreme in how they interpret this. It doesn't mean that I've stopped writing novels. In fact, one of my chief goals over the next couple of months is to transfer the responsibility for this project into the hands of competent engineers so that I can get back into a productive novel-writing mode. But neither does it mean that I'm somehow not taking the game project seriously.

Well, when we got guns, we forgot how to sword fight. A really odd thing happened in Japan: they got guns and decided not to use them for several hundred years, so they never broke the lineage. So, in the East they just did a better job of curating their old martial arts than we, maybe pragmatic, Westerners. Then they just really nailed it on the image making front. All of these pop culture icons turned up in the '60s and '70s, like Kato in the Green Hornet, Bruce Lee, the Kung Fu television series - innumerable martial arts films from Hong Kong.

"one of my chief goals over the next couple of months is to transfer the responsibility for this project into the hands of competent engineers so that I can get back into a productive novel-writing mode."

The term martial arts became synonymous with the martial arts of Japan and China, which is fine - they deserve all of the attention and glory they can get for themselves. A different point of view seemed to take hold around the way in which people fought in mediaeval Europe, which was predicated on the idea that a person in a suit of armour is so weighed down that he can't stand up, can't mount his horse, can't move freely, that the weapons are heavy and slow and ponderous, that everyone is nasty and brutish and stupid.

So that just became the standard kind of shorthand used by film makers whenever they were depicting people in the West. It's a remarkably difficult habit to break.

Even when people did wear full armour, it was much more lightweight and flexible than people realise. They could do shoulder rolls, run. There's plenty of YouTubery around this at the moment. Just within the last few weeks there have been a couple of mediaeval fighting tournaments where people have been going at each other in full armour, one's called lists on the lake, in Texas. The swords are far lighter and more mobile than is generally appreciated.

But as you say, it wasn't all people in full armour - a lot of these arts were intended to be used by people dressed in ordinary street clothes. They presume a much higher degree of lightness and mobility that we see in movies.

It has to. It has to have levels and modes that are just about the straightforward smacking around. That's key to making it work at all. We're sort of following the template that's been laid down already by first person shooters where you always have that. You can go into, say Halo, and set it to easy mode, and it's easy.

Many people don't want to go beyond that. Many people beyond a certain point say, 'I know, I'd like to make this a little more interesting, a little more challenging.' By changing difficulty settings or advancing to a more challenging level it becomes possible to increase the level if challenge and thereby get drawn deeper into that world. So, that's a tried and true approach to game design that ought to work perfectly well with what we're doing.

The technology isn't really there to do a full simulation. Even if it were I don't know if that would be terribly satisfying because a lot of what people do if they're just playing intuitively is going to fail.

The way to keep things practical from a tech standpoint and interesting from a player standpoint is to embody the fighting styles and martial arts which the historical fighting community has been reviving the last couple of decades because that narrows down the tree of what you can do to a far smaller set of variations. They are variations that make sense martially. By selecting those moves you're offering a selection of things that actually have a chance of actually succeeding in a fight.

The technological challenge is to bring together some elements of recorded movements and kind of free IK based character movement and blend that together in a manner that will be satisfactory.

The demo, which you can see snatches of in the video, is actually running in Source, but we're leaning towards Unity for the next iteration.

I'm not in love with middleware as a term for what we're doing. I'll say that people love to mod games and to contribute their own content. We want to make those people happy. We also want to open the door to people in the historical swordsmanship community and give them a chance to take their ideas and research into historical swordsmanship and place those into a functioning game and universe.

It's a hugely diverse and argumentative community of people. We're never going to make all of them happy by making one sealed and shrink wrapped game about one particular fighting style, saying 'here it is, it's finished.' In order to do even one version of one game, we have to sort of build the underlying framework anyway. As long as we're building that framework, we may as well document it and allow people to add their own content or improve what we did if they think we've got it wrong.

It always ends up happening, so the right stance to take is encourage and celebrate it.

I think that with any engineering project like this, an engineer can add a layer of hardware abstraction which says, beyond this fence there's some hardware that's giving us some information and it can always be swapped out. So that's almost an automatic feature of any well engineered code, I think.

One of the few pieces of information I've managed to pick up from hanging around with tech people is that any plans you make for more than three months in advance are completely worthless. [laughs] Three months later everything is different and your plan is not only wrong but completely besides the point.

The closest we've got to having a plan is just a general idea that if you can get a sustainable kernel of activity: something that works, people using it, some cash flow coming in, then as long as you pay attention to what the users are saying, and are responsive to what's going on, it's possible to iterate around that and get rid of what's not working. The later on, ten years later, you can look back and claim it was your plan all along.

I think a lot of people coming to this from a kind of physics/engineering mindset have got hung up on an implicit assumption that it has to be a fully realised 3D physics simulation that can handle all possible eventualities. It turns out that, if you start to drill down into the actual martial arts systems, almost all of the possible outcomes are sort of pruned away because they don't make sense. You're left with a smaller and more manageable set of outcomes.

One example is that in a lot of existing swordsmanship simulations, you see these gigantic roundhouse swings that look great in a movie or a game, but they're incredibly stupid form a sword fighting standpoint - if you miss you've followed through too far and you've taken your sword out of the action and you can't defend yourself. Any real sword fighter would just dodge back half an inch, let it go by and then counter-attack.

If you're trying to write a game in which that gigantic roundhouse swing can be violently blocked, then you've set yourself a very difficult problem in terms of haptic feedback. But if you look at how these sword arts really work, no-one ever does that. Instead, since all you need to do is barely touch the person in order to win the fight, you're trained to make controlled attacks and pull them so that even if they miss, they're going to stop very close to where they would have been blocked anyway.

So the difference between the blocked and unblocked version of that move is a matter of a few inches. Therefore the divergence of the position of your hands and how they're displayed on-screen is much smaller than I think people tend to imagine. Once it's that small, it's much easier to deal with by haptic feedback and user cues that are well within the realm of technological feasibility right now.

I doubt it. His approach is the very occasional and very useful nudge. Getting heavily involved with a company, or investing a lot of money in it, is a little bit like taking a puppy home. It can seem appealing at first, but it can also bring a lot of headaches and responsibilities further down the line. I think he's wisely steering clear of that level of involvement.

He's a person who can show up from time to time, when least expected, make a three word suggestion or hit a crowbar in a video and then he's gone. I'm sure that's not merely a personality quirk but a well considered strategy on his part. [laughs]

There's been quite a lot of that already. [laughs] I think a lot of it is a somewhat knowing interaction between Gabe and his fans, a multilayered 'I know that you know that I know' sort of thing.

I think this is one of the areas where there's been some confusion between our immediate goals with the Kickstarter campaign and our eventual, longer term goals. As I mentioned, with our Kickstarter goals we have to pick something we can do and raise enough money to do it, that happens to be a PC game.

It'd be foolish not to get it up on as many different platforms as we possibly could once we reach our initial goal. If we went out right now claiming to do that I think it would be sort of proving to the world that we don't know what we're talking about.

We're keeping it real for now by focusing on this immediate thing that we think we can do. If you are a mac gamer, or a console gamer and you want to see something like this on that platform then the best we can do is to ask you to take the long view, contribute to the Kickstarter so that we can have something to build on and expand to other platforms once we're up and running.