http://www.garrysmod.com/quickimg/bf...60981d1dc8.jpg
Gabe Newell is the cofounder of Valve Software, one of the most prestigious game development companies on the face of the planet, responsible of course for creating the Half-Life series, and without it, Garry's Mod would not exist.
Community member
mm3guy has interviewed him, check it out below!
mm3guy: Hey, thanks for being here today! So, so start this off, can you tell me a little about yourself?
Gabe: I grew up in the military. My first video game experience was playing Trek on a Burroughs mainframe using punch cards. I went to Harvard. I joined Microsoft in 1983, and left to start Valve in 1996 with Mike Harrington.
mm3guy: When did you first decide to get into the video games industry?
Gabe: When I was working on Windows, there was a point in time when the common wisdom was that while Windows might be fine to run spreadsheets and word processors, that it wasn’t possible to run games without writing to the bare metal. This annoyed me as it didn’t make any sense, and so I decided to find the coolest and most demanding game and work with that developer to get it running on Windows. At the time, Doom had just come out and lit everyone’s hair on fire, so I contacted the guys at id about the idea of some people working for me getting it ported to Windows and then giving them the code back. Bill McCormick was the engineer who started the port, and it was built on top of Chris Hecker’s new technology called WinG, a precursor to DirectX. Bill’s deceased now, and Chris most recently worked at EA on Spore.
Through that project I got to know several people at id. One of the smartest guys at Microsoft at the time was Michael Abrash. It turned out that Michael was one of John Carmack’s programming heroes, and John convinced Michael to leave Microsoft to work with him on the Quake engine. After Michael had been there a while, he started talking with Mike Harrington and myself about this new engine they were doing, and his feeling that it was going to create a lot of opportunities to do super interesting work in the game space (I think his exact words were “Quake will be the worst game done on the Quake engine” which wasn’t a slam at their own game so much as a statement of his confidence in what the tech enabled). Mike Harrington and I went down to Texas with Ken Birdwell and Kelly Bailey, and we left with the source code to Quake, and a mission to build what turned into Half-Life. Abrash is at Rad Game Tools, and his early involvement in Larrabee is one of the things that has us fired up about how significant that architecture will be as it proliferates.
mm3guy: So, what was it like for a brand new company like Valve making a game like Half-Life way back in the day compared to making a game now? What kind of challenges did you face?
Gabe: When we started working on Half-Life we didn’t have any idea what was important and what was irrelevant. Lots of people were confidently giving us what turned out to be, in retrospect, really terrible advice. Nowadays we have a lot more clarity about what we should do, but never seem to have enough bandwidth to get everything done that we would like to. I still want to go back and actually ship some of the ideas we put on the backburner when we pursued Half-Life. The toughest time for the company was the middle part of Half-Life 2 when the source code was stolen and released on the net and we were in a lawsuit with our publisher of the time.
mm3guy: When did you first realize how big of an impact having a large hand in your games' community has?
Gabe: We’ve always felt that we were part of a community. Movie directors of the auteur school have this fiction that of a personal creative vision that the viewer has to appreciate. The purest form of the experience is inside the director’s head, and the customer’s takeaway of the movie or even the version that is actually on the screen hampered by all those pesky production details are bastardized versions of the “real” vision of the director.
This is the antithesis of how a game company needs to think. You are always a partner with the player in the performance of the experience (admittedly the lead actor doesn’t have a copy of the script and has a tendency to improvise by shooting your NPCs in the face). If you aren’t collaborating with the player, your games are going to suck. This collaboration is a great model for everything we do – whether it’s provisioning game servers, or creating mods, or people just not being dicks when we play Counter-Strike – viewing ourselves as being part of a community that is creating all the fun rather than somehow separate and distinct.
mm3guy: What kind of planning did you guys do before deciding to purchase mods from the community such as Counter-Strike or Day of Defeat?
Gabe: Mainly we look at the people involved. The fact that a MOD is doing well is a pretty good sign that at least some of the people involved are the kind of people we all want to work with.
So we don't really "plan" things in terms of sales, market analysis, development funding, or an ROI analysis on a MOD. The critical issue is sitting down with the people and getting to know them and what they would be like if they were sharing our offices. And after we pick up a MOD, we don't analyze the success in terms of sales nearly as much as we analyze it in terms of "how are these guys working out and what have they been able to accomplish."
I don't think anyone in the industry could have predicted the relative volume in sales of Counter-Strike versus Day of Defeat, and then done so consistently based on a pre-purchase planning stage. On the other hand, we'd even with the sales variance, both purchases were very successful, and we're super happy that Jess Cliffe/CS and John Morello/DoD (for example) are still here.
mm3guy: When did Garry's Mod get your attention?
Gabe: That would be better answered by Erik. Erik?
Erik Johnson: We noticed Garry's MOD very because it was so easy to keep an eye on, I'm pretty sure he was releasing early versions before the SDK was even released. We try to get a full playtest in for most projects that are released, but Garry's MOD had the advantage of communicating most of what it was about through screenshots we would send around the company. Also, it seemed like everything produced with Garry's MOD was pretty funny.
What set Garry's MOD apart from a lot of other projects was the frequency that he was producing updates to his user base and listening to their feedback. We saw it grow over time as a result of his work, to the point where offering the game for sale on Steam made a lot of sense.
mm3guy: So what are the chances of you guys venturing out of the First Person genre and experimenting with other styles of games?
Gabe: Well, we’d certainly like to. When we were originally thinking about games we could build back in 1996, only one of the three, Half-Life, was first person. Prospero was third-person, and the unnamed underwater sub game was a … sub game. With each game we try to pick off a piece of technology and design work which then can be used by the rest of our games, like physics and the facial animation system from Half-Life 2, or the Director from Left 4 Dead. We just need to piecewise extend ourselves towards a broader set of games. The motivation and desire is there, we just need to move things forward in a way that gives our customers the most cool stuff while we get there.
Everyone once in a while we get all fired up to jump into the beyond and do an RTS-style MMO, we start making hand waving gestures about how utterly awesome it will be, and then some bastard points out that we still haven’t shipped the TF-2 updates on the 360 or Episode 3, and we come crashing down to earth.
mm3guy: Not too long ago you guys mentioned the idea of a full length (or at least, longer than a simple short) Team Fortress 2 movie was being tossed around, have there been any developments on what the plans are for the movie making side of Valve?
Gabe: Our approach to the movie making is incremental. If you look at the series of pieces we’ve made, starting with the DoD shorts through “Meet the Spy”, they each represent a step forward technologically and/or in some aspect of story telling. We’re also getting a better handle on how they relate to the games and what our customers like. The next three pieces won’t be significantly longer, as they are intended to let us work through different issues, but after those, at least with our current roadmap, we’ll start doing longer format.
mm3guy: So, do you have anything information about HL2 Episode 3 that you can feed to your horde of information-hungry fans?
Gabe: We don’t have anything to say about Episode 3.
mm3guy: Steam Community has been a huge hit so far, with hundreds of thousands of people utilizing it every day. Do you have any future plans for enhancements for it, like new features, or even an API so people can make their own stuff for SC?
Gabe: Yes, we are going to continue to add functionality but we don’t have anything to announce yet. The lack of an SDK is obviously a problem for us and for the community.
mm3guy: Again, thank you for being here today, just too bad you weren't able to say anything about Episode 3.
Here's some relevant links:
[...]
