PC Zone 55 (October 1997)(uk)


Taken from 75eb8c9a274e6cfd696c0a1ad606ff82 PC Zone 55 (October 1997).pdf (thanks to Pix2 on UG)

Cover Disc

There are AirQuake, QRally and Quess, repackaged.

Oi! ... Carmack What's Your Game?

He programmed Wolfenstein. He programmed Doom. He programmed Quake. And now he's really, really rich. But it hasn't stopped him from caring about games. David McCandless corners John Carmack of iD Software.

Pretty spartan, is the official line on John Carmack's office. A mere six workstations sit humming on a row of desks. A couple of biros and a bit of paper can be seen. Oh, and a bunch of real life, real sharp medieval swords are piled in the corner. Carmack is articulate and boying looking. He could easily be 14, but he drives a rather expensive Ferrari and is regarded as one of the greatest programmers in the world. He's not exactly a party person, though he's rather 'good' at writing real-time, perspective-correcting, dynamically-lit 3D game engines. But you knew that.

PC Zone: Did you always anticipate Doom and Quake to take off in quite the way they did?

John Carmack: It's worked out how we expected it to, really. There have been conscious decisions in Doom to make it more amenable to the modifications. Then in Quake, we went a further step and created a really excellent extension mechanism. So yeah, I would have to say we made all the right decisions, but we're still surprised at how extensive it has been.

Does it please you?

Yeah, it's gratifying to see.

And does it please you that you're employing people now from the community who you've indirectly trained?

We've known for a long time that people from the community would get jobs because of what they did. Now you can do such cool stuff with Quake, and show that you can work with an established code base, do something interesting, innocative and worthwhile. Any kind of product is worth so much more than where you went to university or any past job performances. It's a good, solid level playing field for people who want to get into the industry.

So what is iD's next step?

We're going to continue to push the technologies. The thing which makes the industry so interesting is that we ride the technology currents. We're getting twice as much processing power every 18 months. I have vague ideas of what the next two geenrations are going to look like.

So what are they going to look like?

Okay, we're now moving into curves and fractals and atmospheric effects that we couldn't attempt before tbecause the geometrical calculations were too high. We're also going to be riding the 3D acceleration wave, which is great for us. Wolfenstein to Doom to Quake have all been riding the CPU wave, so on this next we also have these graphics accelerators. We'll be able to get twice the polygon rate, twice the geometrical complexity of the worlds we create. That's going to be nice.

Is this going to mean more detailed, better graphics for games?

A fairly bad Doom scene or a moderate Quake 2 scene can have 5,000 triangles visible. I expect the next game generation aimed at hardware to be averaging 10,000 and peaking around 20,000 triangles visible in a scene. So when you're looking at a guy, instead of seeing planes on the sides of his head, it'll be smoothly contoured using bump-mapping. And when you get right up to walls with pictures of pipes, the pipes will actually bend out of the wall. They'll have shape to them rather than just being a veneer over a flat geometry.

What about lighting?

I've got three or four major directions I'm still investigating. It's probably going to be like Quake where we'll try everything and figure out which one works best. Dynamic shadoes from static lights will be cool. For instance, you could have a monsters's shadow creep up against a wall, projected from a low-lying lantern. I've got that stuff pretty much figured out. The big problem is dynamic shadows from dynamic lights, and you can just go ahead and do that - intersect every light with every polygon, project onto every polygon behind it - but that's not going to be feasible to maintain an interactive frame rate.

Anything else?

Atmospheric effects are going to be a big thing. When you think 'fog' in computer graphics right now, you think of the haze where everything fades into white. People use it as a crutch instead of proper scene management, so they render to a certain point and then stop you seeing large areas. The future will yield volumetric effects. Instead of a consistent atmospheric absorption, you'll be able to have low-lying mist. You'll toss a grenade down onto the floor and you'll see the clouds rush away from it. All effects like this are possible.

Everyone in the universe seems to be doing a Quake-style game or a Quake-style engine. Do you not feel threatened by Prey, or Jedi Knight, of even Microsoft's new engine?

If you step back in time to a year ago, Quake had been out for around a month and every one was talking about how these Quake killers were right around the corner. A year ago. And still none of them - not a single one - has shipped.

But Jedi Knight's coming out very soon, isn't it? Surely that comes close to Quake?

Yes, with Jedi Knight the content's good, but it's not a technologically superior engine and they know it. They're doing the same thing that I am - but they're still half a step behind.

But surely there are other 'Carmacks' out there somewhere?

There are very few people who have the right ressources and focus to do this. It's getting tougher all the time. People look at Quake and say "Wow 3D graphics", but not even a third of the code is in there, let alone a third of the effort. There's a gulf between the demo you show to people and an architecture you use for production work that's going to ship a commercial game. Integrating the sound, doing the network communications, handling the control and the game extension language - this all has to be done. You can't just say: "I've got a little demo and I can fly around a 3D world and it looks like Quake, and can I be a millionaire now?"

So you're not bothered about your rivals?

I'm not really worrying about being scooped to the next generation, no.

Just what was going on during Quake's much rumoured problems?

We were actually getting concerned about development cycles. Wolfenstein took six months. Doom took a year. Quake too 18 months to develop. That was really too long and a lot of people were fraying really badly. Quake was messed up due to internal issues rather than conscious decisions.

So what does this mean for iD Software games? Are we going to see any different game styles from you guys?

The idea of making a simple game versus a complex game is definitely going to stay.

Oh, okey, so I've got this great idea for Porn Quake. You know, 'spooge cannons' and loads of tunnels. I want to license the Quake engine - how do I do it?

Well, we've almost got a clean, simple sign-it-here license. You can negotiate for more upfront and less royalties. The ranges have been from several hundred thousand to a million dollars, balanced against the royalties.

Eek. Might have to get a loan then.

It works out well for us because we're still getting a substantial amount of return on it.

I bet. But doesn't it get kinda… competitive?

There's some sniping about who's doing what and all that and when people start treading on each other's release dates, then it gets a bit ugly. But you can look at it and remind yourself that if Ion Storm have a spectacular sucsess we'll probably make more money than anyone there makes out of it - because we've got a big chunk of the royalty.

Are you going to bund the Quake engine out into the community one day like you did with Wolfenstein?

Yeah, definitely. There's a book coming out this year that's basically 'Inside Doom', with all the source code. I look forward to the time when we can do that with Quake. For one thing, Quake has the cleanest, most elegant internal structure of any of the projects - that's easier for people to read and digest and learn things from.

Will Quake 2 be Quakeworld compatible?

It's not compatible. QW was my research project that lots of people got to play with. A lot of that technology will be rolled into Quake 2. We're going to make it as good an on-line game as possible. So yes, the network game will have some significant enhancements.

Do you think there'll be a stage where a future engine will become the standard, an operating system if you like, for games?

We're a lot closer to that right now. After Doom was released, there were 50 teams doing Doom-style engines. Most of them failed miserably and I think many companies took this as a learning experieice. And in the end there were only a couple worth a damn. There was the BUILD engine which 3D Realms used and there was Dark Forces. Quake is a hell lot more complicated to duplicate than Doom was. But when you balance it against say, half a million dollars, for most companies it makes good business sense to license the engine with proven technology instead of taking a blind shot. We stopped after about six licenses but we're going to open it up after Quake 2 ships. We're going to wash our hands of it. We're on to other things. Let everyone else fight it out over content.

Is this going to improve the quality of games?

If you have eight companies doing roughly similar games with the same technology, then they'll have to compete on content. It's almost like a console for PC games. The console has the same hardware capabilities and everyone has to differentiate without radical technical innovation. It's like there's now a gaming platform with the Quake engine and people differentiate on content.

Which can only be good for the gamesplayers, right?

Absolutely. We're in a good position to capitalise on it because we're taking a sizeable chunk out of everyone's pie. For the gamer, it's clearly going to be a good thing. There will be more games to choose from, but the entire breed should be improved by the competition.

What are you going to do next?

We might do a Quake 2 mission pack. It depends on how much time I spend on research stuff. I've actually spent more time improving Quake 2 than I intended to. I'm really looking forward to taking my clean sheet of paper and starting over again.

Are you going to go on holiday?

I don't believe in vacations. Last year I took five days and went to England. My God, I was nuts by the time I got back. Actually, I had a laptop with me but it wasn't the same.