Valve's Gabe Newell on, well, everything

I think we were trying to be responsive to the feedback we were getting from customers, in the same way that Half-Life 2 tried to respond to what we were hearing from people after we released Half-Life 1, and Half-Life 1 was our own response to what we saw as being the positives and negatives in the genre at the time we did it. So, we wanted to move the story forward faster, we wanted to be able to adapt to hardware changes faster, we wanted to try to manage risk on the project rather than making them five years long which really limits you in your ability to try out various things. Those were some of the main reasons.

We're really just tinkering with the formula. I sort of wish in general that game developers would try more experiments, because as an industry we're all going to benefit if everybody's not trying to duplicate exactly what was successful by somebody last time, because we're really going to learn. So some people are, you know, the Sam & Max stuff is coming out on something like a monthly basis, so it's going to be interesting to see what kind of experience they have from doing that, and we're taking the amount of time we are and that will be instructive in terms of how much we can get done versus how long it takes us to do and that kind of thing, so we're all going to be learning a lot from each other over the next couple of years as to what works best for customers with which kind of properties. Some properties I suspect and some genres I suspect will benefit more from shorter release cycles and others will work better with longer ones. So we're all going to learn. We're going to do three episodes and sit down and talk to our customers and find out what worked and what didn't about that approach and adjust accordingly.

Well we're looking at a variety of ways of funding these development projects and it probably benefits us less than other kinds of developers. We're very successful with the retail model of large-scale projects; we could have just kept turning the crank on that. I think where advertising-supported projects will be interesting is the degree to which it continues, you know the same way Steam hopefully broadens... one of the issues that it solves is broadening the distribution of games like Red Orchestra, which deserve to reach an audience. I think that we're also going to start to see games that would struggle to get traditional publisher funding find that advertising is a great way of finding and developing an audience, and that's why we're putting the effort to make it possible for people to use Steam to do advertising-supported games. So we really see this as another option that we and other developers can use to figure out how to fund projects going forward.

What I would hope to see is that small developers can give away their titles for free and garner ongoing development support by generating advertising revenue, and we've done all the work to make that possible through the work that we're doing in Counter-Strike. That's certainly the hope. Another aspect in addition to broadening monetisation options and funding options, especially for new developers, is the possibility of segmenting your audience. So any time you can give people more pricing options, that's always been a good idea. Some people will prefer it one way and some people will prefer it the other way, but it really requires us - especially as the technology provider that a lot of developers are starting to depend on - to do it first and get it out there, work out the kinks in the system so that they can then use it to trial things themselves.

We haven't really thought about that. If they want to talk to us about it then they can. I think in general... there are between 150,000 and 180,000 servers around the world, so I think we've traditionally done a good job of supporting those people and giving them what they need to be excited about running and hosting servers. That's why it's an order of magnitude larger than any other server community for any other group. We're certainly always interested in engaging with any server operator to find out how we can do a better job.

The Steam business? It's going great. I think that we're continuing to get a clear idea of what we need to do next for consumers, how to make Steam more useful to them, how to solve problems. I'm pretty excited about finally getting some display drivers - the ATI announcement included that - where rather than having the situation they have right now where there are literally tens, close to hundreds of display drivers out there on people's machines, that everybody will have the most current, the most up-to-date driver, automatic bug-reporting and things like that. That's a nice step forward.

That'll also I think provide a lot of value to customers as well. We also need in addition to make the customers happier. We need to make sure we're listening to and developing the features and functionality that other software developers need to make it more useful. There are little things, like being able to have better control over international pricing, and have day and date releases in Russia and things like that that as we talk to people and find out what's important to them - we have lots of those kind of check-boxes to check, and as long as we keep doing that I think we'll continue to see excitement about Steam building.

Well, you know, as long as you've got Microsoft willing to lose six or seven billion dollars to push a hardware standard, it's hard to see, you know... as long as people are willing to absorb losses on hardware. Sony's doing the same thing. I think they just announced three quarters of a billion dollars in losses. As long as you've got people willing to do that, it's hard to see why gamers aren't going to want to have subsidised hardware from those people, so I'm not sure how soon we'll get to unification.

I think the thing that - from a technology perspective rather than the hardware issues, is really exciting - is this collision that's coming up between the GPUs, which have had these really light execution capabilities and the CPUs, which have had great compatibility stories and great single-thread execution capabilities. Those two design or architectural characteristics are going to converge and we're going to end up being in a world with all these huge numbers of homogenous cores and it's going to require us to rethink our approach to how we design game engines, but I think it's going to vastly improve the kind of game experiences we can give people; the kind of huge acceleration of just graphics quality that we've seen over the last couple of years is going to apply to all of the other aspects. AI, physics, game behaviour - all of that's going to suddenly be able to benefit from the same kind of scalable growth that we've been seeing on the graphics side, so that'll be I think one of the major characteristics of the next five years is watching that sort itself out.