Interview: EPO talks to Darren Stone about Faceball 2000

faceball 2000It was the Early 90’s and developers were looking for a few way to put players inside the action of games. Giving birth to the First Person Shooter genre of games, at least on the handhelds, was Faceball 2000. No clue what Faceball 2000 is? Go here.


Don Komarechka:
What was your title at Xanth Software F/X?
Darren Stone: Programmer (they call this a “Software Engineer” nowadays!)

Don Komarechka: At what point in the company’s history were you employed?
Darren Stone: I was employed by Xanth from late 1991 until mid 1993. It was my first job out of university.

Don Komarechka: What were some of the projects that you had helped develop, and what areas of these projects saw the most input from you?
Darren Stone: The only released project, and thus the only one I can speak about in detail, was Faceball 2000 for the Sega Game Gear. And even that game wasn’t released during the time I was there. I learned about its Japanese release after I left the company. I was the programmer to start the project and carry it through almost, but not quite, to completion.

There was no artist required for the Game Gear project — wherever possible, we reused artwork from the Game Boy version. You can blame me for the nasty colorization of the greyscale Game Boy art!

Some really good music and sound effects were composed and I integrated them into the code. I recently found the Faceball 2000 main theme song extracted from a Game Gear ROM image floating around the Internet. I probably hated listening to that song over and over and over at the time, but it’s my cell cell phone ring now, 13 years later. That’s full circle! I love chiptunes.

I also did PC prototypes of Faceball for both Windows (more of a limited P.O.C.) and DOS (almost playable, including a trippy, moving landscape “maze” and networked multiplayer arena). Those were shown around a bit at the time and at a CES, but we never received publisher funding to go ahead with them.

Don Komarechka: On a technical level, what was the most challenging part of the design process? You?d be dealing with a fraction of the memory sizes available to the SNES development team, what had to be scaled back in order to accommodate the hardware?
Darren Stone: Two main challenges:

1. The nature of Game Gear display RAM. Just like the Game Boy, there’s no bitmapped drawing surface on the Game Gear. This would be a strange platform for programmers used to today’s PCs and consoles and even handhelds. These old handhelds were designed with fixed or scrolling (tile-based) background, sprite-based games in mind. You’re always in text mode, essentially, so if you want a big bitmap to draw into, you generally write a block of characters onto the screen (”ABCDEFG…”) to fill up whatever area you want covered. Then you redefine the 8×8 character definitions for each of those characters in another part of RAM so it all pieces together and gives the effect of a continuous bitplane. And you do this over and over and over again, every frame, whenever you change anything on the screen. And then you realize it’s all too slow, so you optimize it. And optimize it. And you can only do this during a vertical (or horizontal) blank interval, otherwise you’ll see flickering on the screen. Rob Champagne showed me how he did all this on the Game Boy, and it was quite genius — so I don’t want it to sound like I was the first person to solve this or anything.

2. CPU. Think of all the perspective transformations, clipping, collision detection, and other math-heavy operations required for rendering first-person perspective. Now think of doing that with -spare- cycles on a low-end Z80 in assembly code. It seems pretty crazy that we attempted to do this at all. Faceball was never as limited by ROM as it was by RAM, so lots of ROM lookups were used to quickly approximate the heavy math on the fly and we took advantage of the symmetries and constraints of the Faceball maze environment. Again, I mostly took what Rob did on the Game Boy and replicated it on the Game Gear.

As far as I know, Faceball was the first 3D multiplayer first-person shooter for a handheld platform so it was all worth it.

Don Komarechka: Did any of the efforts you put forth at Xanth ever see the light of day in the North American market?
Darren Stone: No, not that I’m aware of. The publishers I dealt with were based in Japan and their responsibility was their own domestic market.

Don Komarechka: Other versions of the game, the Gameboy version to be precise, had a few revolutionary multiplayer ideas that were slapped down by Nintendo. In the development of the Game Gear version, what was the multiplayer scope intended to be, and how far did it progress?
Darren Stone: We kept the Game Gear multiplayer mode simple. Two devices, head-to-head. It would have been nice to cook up a crazy 16-player cable for the Game Gear like Xanth did for the the Game Boy, but budget and time didn’t permit. Thus, the Game Gear version was designed for head-to-head play and arena play with two humans and optional AI drones. I should note that I started, but did not finish, debugging and tuning the final multiplayer code — I’m not sure who finished it!

Don Komarechka: Were any easter eggs stuck into the code for fun?
Darren Stone: Yes. All the level warps, sound tests, secret credits, and other tricks you know from the Game Boy version are in there. And a few more for reasons of color and sound testing, since the Game Boy was more advanced that way. In most cases, the same codes and cheats you can use on the Game Boy can be used on the Game Gear version. For example, the old 180-degree turn-and-shoot trick at the end of Level 10 and the “config_sys” name trick are in the Game Gear version also. I don’t remember all the codes… that was 14 years ago!

Don Komarechka: From beginning to end, how long did the development of the project take?
Darren Stone: Definitely at least 6 months in my hands, probably more. We reused the same level data from the Game Boy, so we had the benefit of not needing to play-balance test every level or design new levels again.

Don Komarechka: Anything you would have like to add to the game, but due to time constraints had to leave out?
Darren Stone: I wouldn’t have “added” anything. The game is pure and abstract and well-balanced (it has its Midimaze and Game Boy heritage to thank for that), so it had all the features it needed.

I can think of two aspects I wish I had more time to improve though.

1. Color. The Game Gear, unlike the Game Boy, is a color platform, but it imposes very heavy constraints on what you can do with color. You can’t, for example, just arbitrarily plot pixels of any RGB color all over the screen. Again, it comes back to the nature of the device being character-mapped, which is very suitable for games that use sprites and have a tiled or scrolling background, but very unsuitable for treating the screen like a big, randomly accessible color bitplane. Since the hardware was designed for tiles and sprites, you’ll sometimes see more colors on the screen of simple Game Gear side-scrollers than I could pull off in Faceball. Sad, but true.

2. Frame-rate. See the RAM and CPU issues, above. Really though, I’m not sure much more could have been done to improve this. I was moving so much memory around, so often, and on such a slow machine that the frame rate suffered. We were just doing things those little handhelds were not designed for, plain and simple. Once you get immersed into the game though, you see past framerate issues.

Don Komarechka: What were the best and worst moments during your time at Xanth?
Darren Stone: The best moments were writing and sharing code with the other programmers there. It was just a very talented group, and that’s always nice to be a part of. It was a big, open environment with lots of show and tell, group coding and debugging, and I think we all learned and grew a lot. (Well, I did, anyways.)

Meals and parties together were always a lot of fun, too. We had a generous boss, James Yee, and we were fortunate to live and work near Granville Island in Vancouver, a very hip and healthy part of the city. So good food and drink were always part of the company culture. Everyone worked long hours but we goofed off and partied just as hard.

No “worst moments” stand out.

Don Komarechka: Were any of the other unreleased Xanth projects as innovative as Faceball 2000 was?
Darren Stone: Yes, I think so. Several epic projects were undertaken, on several platforms. That’s probably as much as I should say, since I’m not sure where Xanth’s intellectual property ownership ended up.

Don Komarechka: After your time at Xanth Software F/X, where has your career lead you?
Darren Stone: It has been almost 13 years since I left Xanth. I’ve been in and out of the games industry a few times, but always in software development. EA Sports for a couple of years (AI, stats, and audio for Triple Play ‘98, ‘99, and 2000 on Sony Playstation and Windows). Relic Entertainment for a few years (Homeworld, Homeworld: Cataclysm, and Impossible Creatures).

Outside of games, I spent a few years programming Fortune 500 cashflow and investment management systems. Sounds boring, but writing code that suggests where to transfer a few hundred million $ every day has its own rush! Bugs are very costly — good experience!

I did some independent consulting during the tech boom of the late 90’s. And for the last few years, I’ve been cofounder and CTO of Verrus (http://verrus.com), developing cell phone software, SMS entertainment and marketing apps, and mobile commerce.

When nobody will pay me to write games, I write them anyways. I’m always tinkering with AI and math for stuff like Go.

Don Komarechka: Any final comments?
Darren Stone: Thank you, Don, for keeping the spirit of Faceball 2000 alive. It’s easy to forget the heritage of today’s multiplayer first-person shooters, so thank you for documenting this part of it.

Comments

  1. October 27th, 2008 | 9:52 am

    MIDI Maze - External links…

    MIDI Maze External links MIDI Maze Utilities…

Leave a reply