Tuesday, August 18, 2015

games

I can't get anyone, period, to talk to me about business. They pick me for an idiot, and the fact is, they're right. But I think I'm right, too. Maybe if I could publish.

Computing looks so advanced, and so complete, but I think it's in a complete muddle. In the near future - this is what I predict - we'll think what we were doing with computers today was pretty silly. We'll also be glad we got past it, and can finally have some real fun.

In a word, CAD and virtual reality are the future of computing - they're really what computing is, but that's a long story. Today, at any rate, they are regarded as the periphery, outliers. Real computing is about e-mails, files, apps. CAD is for professionals only, and virtual reality is for gamers.

Let's call it games. The pros can keep their CAD. I don't want it. But the games we play, in the future, won't be today's frenetic fantasies, they'll be mellow deliciousness, and quite serious - seriously fun.

You will enter a world, even a universe, through your screen. It will be your world, your universe, and there you will play to your heart's content.

You will play, there, in blissful solitude. When you want to, you will invite friends. You will invite them to your party palace, or planet, or star, or galaxy, and when you want to be alone, you will go into the hills, to your hidden temple, and be ... that.

You will have all the tools you could want, saws and drills, and endless benches, in giant studios, supplied with stacks of fine timber, steel bar and plate, great heaps of clay, wax and coal for casting, fine paper and cloth, and on and on.

And now, to begin.

There is the problem of the interface. It has one solution: xml. The fact that this isn't universally recognized is evidence that computing, today, is in a muddle. It isn't that there's anything wrong with the computers. They are improbably perfect. The problem lies with the software. But I won't go further into that here.

Then there's the problem of rendering. You will experience your universe as images on screens. I, by the way, could care less about immersion. Computing is a medium. If it were necessary that a medium be indistinguishable from the rest of existence, then books and art would never have been invented. Well, I'm old fashioned. But you will experience your universe that way, as image on screens. These images will be constructed from a model. That's the xml.

In case you haven't noticed, I'm talking about very spacious models. If you want to explore galaxies, and not just a few of them - deepest space - in your model, you will be free to do so. If you want to explore stars, and the planets that orbit them, you will be free to do so. If you want whole continents, and the seas between them, to build out, all for yourself, they will be yours. If you want forests full of trees and wildflowers to wander in, wander. Very spacious models.

No limit on the spaciousness of your universe is imposed by the hardware, even today. There is no limit, even today, even on the amount of detail, or its precision. The problem resides with the software, and the unanswered difficulty is that of rendering ... infinity.

There is a mathematical formula that solves this problem, but first we must discuss more precisely what the problem is. The model contains information about countless objects that comprise our universes. We view these objects on our screens. They are rendered there. Extremely compact statements about the objects in our universes are exploded into arrays of pixels by the rendering algorithm. In the model, objects are so small they almost occupy no space, like magic, but when rendered, they become vastnesses of data. Therein lies the problem.

Viewed from a particular place, a very few objects are big enough to see, or, to put it another way, we are looking at them. The vast majority are out of sight. (But, not out of mind. That is the beauty of games, as contrasted with the inscrutability of lists, though, of course, lists are beautiful. But we can easily find our way to any location in a universe, and we can remember, if we've visited before, or even only if it has been described to us, what is there, what was there. This is the way the human mind is constructed, not so that we don't need to visit places again, but so that we can.) This is what the renderer must do, before it shows us its images: it must decide, or know, what is in the picture, out of all that is in the model, and what is not. It only renders what is in the picture, and it does not render what is not in the picture, and thus, it is not, any longer, limited, rather, it is infinite.

This is not, in any way, an insoluble problem. It is quite soluble. In fact, I feel certain it has already been done. But it is regarded as something highly esoteric, as I said, peripheral. Or perhaps it has not been done, but only because of that, and not, in any sense, because it is exceptionally difficult to do. (Admittedly, it is a little beyond me, myself, at the moment.)

Even in the vastnesses of SecondLife, for example, I always feel the pressure of amorphous limits, on me. This may not be entirely because of the rendering algorithm, but I think it is because of that, partly. I think it is because that algorithm imposes some kind of distant limit, and I can feel it. And I recently read that in MineCraft, if you wander far enough, the physics start to break down ...

Let us not be under any illusion: this game will be free. You will need your hardware - your terminal, your devices - and your Internet. That is all. There will be no up-front price of admission ... if you have the devices. But, have no fear, our company will introduce an abundance of delightful advertizing into your universe, and we will prosper greatly from it.