Oberon Games Studio
First posted on 23 August 2007. Last updated on 21 October 2008.
To adventure game fans, Scott Bilas is best known for his work on Gabriel Knight 3: Blood of the Sacred, Blood of the Damned. A graduate from Iowa State University, he has worked as both an engineer and a system architect before joining the ranks at Sierra On-Line. Bilas has been credited as being responsible in leading the technical redesign of Gabriel Knight 3: Blood of the Sacred, Blood of the Damned during its long development. Since leaving Sierra On-Line, he has worked at Gas Powered Games and is currently the Development Director at Oberon Games Studio in Seattle, US. In addition to Gabriel Knight 3: Blood of the Sacred, Blood of the Damned, his other game credits include A Series of Unfortunate Events, BeTrapped!, Inspector Parker, Candy Crisis, Dream Day Wedding, Dungeon Siege, Galapago, Scrubbles, and even the inbox games in Windows Vista. Bilas is a frequent speaker at the Game Developers Conference. He has also published in Game Developer magazine and Game Programming Gems, and has been a section editor for Game Programming Gems II.
We are privileged to have this interview with the multitalented game engineer, designer, and developer. In the interview, Bilas speaks about his past at Sierra On-Line, the technical challenges in the redesign of Gabriel Knight 3: Blood of the Sacred, Blood of the Damned, his current works at Oberon Games Studio, and what holds for him in the near future.
- What was the first adventure game you played? What was the first game you programmed as a hobby?
- Would "Adventure" on the Atari 2600 count? It's not an adventure game in the sense people typically think of them, but, well, it was called Adventure! If not that, then one of the Infocom games, or perhaps a pirated King's Quest. So long ago. My favorite adventure game is easy though: Tim Schaefer's Grim Fandango.
The first game I programmed as a hobby was awesome. I was probably 8 years old. It was called Phoenix Duel (because 'Phoenix' has an 'x' in it, which makes it cool), and was all in C-64 BASIC. I didn't know how to do interrupts, non-procedural programming (frame, input, logic, render, loop), or even do a basic simulation, so if one player shot at the other, both players would freeze while the shot moved across the screen. It was really more of a weak game of chicken, but that's a feature not a bug - and let's not forget it was awesome.
- What game development or programming experiences did you have before joining Sierra On-Line? Who were your role models as a game developer?
- My mom originally got me started, teaching me some BASIC on an old North Star (CP/M). My dad gave it a shot too with some HP-RPN and then C, but I didn't really "get" programming until much later. I did the usual teach-yourself-to-code thing, and like everyone I struggled with pointers in C for years before I saw the light with Turbo C++ 1.0. Later, the classes I took in college were pretty worthless - learning Fortran and reel-to-reel tape sorting (I'm not kidding) was kind of fun but not really useful. In my last couple years in college I worked for the Center for Nondestructive Evaluation and got to make robots whirr around, scan eddy currents, and draw pretty graphics of the results. This is probably where I finally figured out how to write decent code and not just hack functions together. I studied the OWL framework from Borland (their equivalent to MFC) and learned a lot about how big systems are structured. That's where I originally got interested in systems and architecture.
I've never really had "role models" that I remember, although I've always liked working with or reading books by smart people I could learn from. Going way back, some early authors I recall reading cover-to-cover are Lippman, Sedgwick, Stroustrop, Meyers, and Maguire.
- When and how did you get recruited to Sierra On-Line? What was your first project at Sierra On-Line? How much contact did you have with Ken and Roberta Williams?
- A friend of mine I worked with at Edmark, Jim Napier, recruited me to Sierra to work on the next S.W.A.T. game, but I ended up on Gabriel Knight 3 instead on the day I started. I never met Ken or Roberta - I think by the time I joined they were already gone. I did see Ken speak at a local game developer gathering one time, sounds like he's had a pretty interesting life. He had said that he was getting tired of golf and thinking about getting back into programming again. It's been a while, I wonder what he's up to...
- You were credited as the System Architect and Technical Lead for Gabriel Knight 3: Blood of the Sacred, Blood of the Damned? Titles aside, how much did your role change over time over the course of the project?
- I originally came on as a grunt to help finish a couple puzzle levels in GK3 so it could ship a few months later. It was immediately obvious that the project was in deep trouble and that I'd be doing more than just some puzzles. The team building the game was experienced, but not in 3D. It was very buggy, few scenes were implemented, it looked bad, etc. Artists would have to wait two to four weeks before seeing their work appear in the game, which made iterating nearly impossible. I met with the producer and the general manager a few times to figure out how to get the game out of the hole it was in. We decided that the game needed to be rebuilt nearly from scratch, and that I should take over the technical lead role to make it happen. This was a big gamble for me - it's not like I knew what I was doing either, but I thought I could figure it out, and talked them into it.
I kicked off the rewrite by getting the foundations of the new game engine done in a sleepless weekend. After that, we rapidly added new systems like font rendering, a console, user interface, the "Sheep" scripting engine, etc. Then we poured content into the game as fast as we could to get it built, as well as bringing on a few new people to build Sidney and write scripts. Throughout the project we were able to rely on Jim's solid "GEngine" for the graphics and core engine services.
- You joined Sierra On-Line at a time when the company was in turmoil from management restructuring and from the fallout of the infamous mass layoff on "Chainsaw Monday" in 1999. At the time, how had this affected you and others on the project team? How much of an effect did these events had on the team's morale and productivity?
- When I joined Sierra they were at a tipping point of projects that were over budget and late, some famously so like Mask of Eternity. At least four new 3D engines were in simultaneous development in the Bellevue office alone, which is a huge amount of risk and duplicated effort. Every company has to make tough decisions to remain profitable. I have seen many independent developers rise and fall with regularity since I started in this business, and it's not easy to keep things going as long as Sierra has. Nobody likes having their project canceled, and nobody likes layoffs, and sometimes there's nothing you can say to make things any better.
That all said, Sierra handled things super badly. For example, I remember that they moved the Babylon 5 team all the way up from California to Bellevue only to lay them off a short time later. To say it hurt morale would be a pretty big understatement. On our team, we lost some people, but in our specific case I thought it was handled fairly well - lots of advance warning, some severance pay, etc. I've certainly seen a lot worse.
- How often did you consult with creator Jane Jensen on the technical aspects of Gabriel Knight 3: Blood of the Sacred, Blood of the Damned? What elements in Jane Jensen's original vision of the game had to be dropped or curtailed because of technical limitations that subsequently surfaced during the game's development?
- I didn't work with Jane all that much, as I was spending most of my time in low level systems supporting the engineers and artists. Jessica Tams, who wrote nearly all the scripts that shipped in the game, was Jane's main interface to the engineering team. We had to drop a number of Jane's designs due to lack of time or because of technical limitations. For example I remember an action sequence in the game where you had to stomp on rats and whack bats, or something like that. We just didn't have time to make the game engine support action, so it had to go. There were many other things like this, though it's been so long I can't remember most of them.
There is one that sticks in my mind, though. Jane had a puzzle that we had to kill which was unfortunately replaced with the famous "cat hair mustache" puzzle that the game's producer designed. The gaming site Old Man Murray gave us an award for killing adventure games because of the cat hair puzzle, as I remember. The team hated that puzzle, but we were trying to ship a game, and so we just let it go. Funny to think about it now.
- What was the day when the development of Gabriel Knight 3: Blood of the Sacred, Blood of the Damned was officially declared complete? For the (historical) record, what was the last feature or bug fix implemented during the final moments of the game's development in "crunch mode"?
- Wow, I wish I could remember. I quit the company a couple days before shipping, so I should know this. Early November 1999 I think. I definitely can't remember any of the final bugs or anything. I'm sure they were compatibility related fixes, though. 3D hardware games back then were nightmares to support. It was probably a bug with some 3dfx driver that made us hack in a workaround. 3dfx wrote the worst drivers (unless you used Glide, which was already in decline at the time).
In the final moments of the game's development we weren't in crunch mode any more. Perhaps the test team was, but development stopped completely. The way it works is you go into lockdown, waiting for 48 straight hours of testing without finding any critical bugs. On finding and fixing a critical bug, you reset the 48 hour timer, wash, rinse, repeat. So the final week or two of a project is very stressful, but also can be very boring, while most team members wait for approval to go gold and finally relax.
- What were the greatest technical achievements (such as graphical engine and scripting system) made in Gabriel Knight 3: Blood of the Sacred, Blood of the Damned? In hindsight, what parts of the game itself were you most disappointed? Why?
- Well, most of what I'm happy with, at least on the technical side, nobody outside of the development team would notice. We built a lot of great tools for speeding up game development, such as the scene definition language, scripting system, console, "construction mode", dynamic reloading, and advanced developer feedback tools. I was so happy with how these turned out that I've taken the best ideas and moved them forward into most of the games I've worked on since then. And while the noun-verb-case system is a very old Sierra concept and definitely nothing that the GK3 team invented, it's something I continue to find useful today as a smart way to organize large amounts of storytelling logic in games.
Our graphical engine was very impressive at the time, particularly in our support of both hardware and software rendering modes. Credit for this mainly goes to Jim Napier who originally wrote the GEngine, as well as Peter Freese who came on later to enhance and optimize it in many ways.
The most disappointing part of the project was probably our animation and "fidget" systems. While the final results were good, these were constant sources of pain during development. They were unstable and prone to breaking frequently. I think the problems we ran into were totally understandable, though, because skinned bone deformation systems in games were brand new. Nobody really knew what they were doing at that time, industry-wide. Today every second year kid at Digipen has probably written an animation system.
- After leaving Sierra On-Line, you worked at Gas Powered Games to develop Dungeon Siege for Microsoft Game Studios. How different was the working experience at Gas Powered Games compared to Sierra On-Line?
- After Sierra I worked at Gas Powered Games (GPG), an independent developer published 3rd party by Microsoft Game Studios. The working experience at GPG was extremely different from Sierra in just about every way - people, culture, technology, game design, company size, location, publisher relationship... It would be impossible to describe what is different between the two working experiences without a whole lot of beer.
There is one important thing in common, though, which is my desire to never work in the hard core game industry again. I'm too old to want to put up with the overtime and stress.
- It was perhaps with a lot of irony that, after joining Oberon Games Studio, you worked on Inspector Parker and BeTrapped!, both of which were games created by Jane Jensen with whom you also worked while at Sierra On-Line. How different was your working relationship with Jane Jensen between these different projects?
- Jessica from GK3 kept in contact with Jane over the years, and hooked up with her a little after Oberon (originally called Odyssey) got started. Jessica got her brother Eric involved and then me as well once I decided to quit GPG. (I should also mention that Jess, Eric, and I all worked at GPG previously). My relationship with Jane at Oberon was pretty similar to at Sierra. I just worked to support Eric as he built Parker. When we got working on Betrapped!, which was much later, I had a closer role in the project but not much. I was distracted building a new multiplayer system for our Game Center.
- You had been a contributor to Game Developer magazine and an invited speaker at the Game Developers Conference. What motivated you to take part in these endeavors? What was the most rewarding experience you gained from these commitments?
- Well, for the record, I've only been invited to GDC one time, for GDC Europe to do my Continuous World talk again. For all the others I sent in a proposal and got approved the old fashioned way. It's been so long since I gave my first lecture that I can't remember why I started speaking at GDC, but I imagine that I originally did it to get a free pass to the conference (GDC is very expensive).
Now, I've continued to do lectures at GDC, and more recently at Casuality, and to contribute to books and magazines mainly because I like meeting new people who are working on interesting things. After a lecture, random people will come up and chat, or send me email, often years later. I'll learn a lot talking (and drinking) with them about different things they're working on, different ideas they have. Some become good friends, some I end up hiring. I love talking about tech stuff with anybody who will listen, even if they're just being polite.
The other big thing I like about doing lectures and writing is that it forces me to organize my thoughts on a subject that I care deeply about, which in turn improves future projects I work on. I recently did a talk in Amsterdam on the development pipeline that made me think about all of the different things that go into the daily work of game developers, and what we can do to make things faster or more efficient. I happen to be doing tech planning for 2008 right now for my team, and trying to figure out how to explain an optimized development pipeline to an audience was immensely helpful in this planning.
- What are your current projects? What is a typical workday for you these days?
- I can't talk about current projects, but two weeks ago we shipped Dream Day Wedding which looks to be a big hit in the casual games space, go check it out! We're very excited.
A typical workday for me is rare. Most of my days are very different from each other. Earlier this week I went snowboarding for a few days. Today I'm sick with the flu, doing an interview, fixing some bugs in our engine, and doing brainstorming on one of our projects. Last week I worked on version control, wrote some code, did some tech planning, researched some new tools and tech, figured out Visual Studio's new service pack, and got hammered at a launch party. The week before I was in Amsterdam at the Casuality conference giving a lecture and enjoying the local coffee shops. Next week I'll do some research, more tech planning, play with the 360 devkit, do an employee review or two, train some team members on tools, and catch up on some casual games I've been wanting to play (like Virtual Villagers 2).
I guess a typical workday is:
1. Work on the top 5 things on my list of 1000 items
2. Add 50 more things to list
3. Prioritize list
Sometimes I think I'll never get my head above water.
- How do you describe the current state of the adventure genre? How has the adventure genre evolved over the last 5 years? How similar or different do you believe between the profile of an adventure gamer and the profile of a causal gamer?
- I think adventure games were getting far too hard core for people. Every year a shrinking group of people were willing to buy them. A very small but very vocal group of players kept saying "it's too easy, it's too easy!" and pushing the games further away from ordinary people. One of my favorite adventure games is Syberia, which I thought was way too hard! I had to refer to gamefaqs.com at least 20 times to win the thing. I thought it was insane. I was talking with one of the developers of this game, and they said that they had gotten nothing but complaints from players that the game was too easy! This is crazy.
I think this is what ruined adventure games: the hard core adventure gamers. Casual gamers just won't put up with that stuff. Sierra was particularly awful about this with their famous "learn by dying" mechanic. Lucas did it plenty of times with their pixel hunts. I used to just put up with it, but lately I've been valuing my spare time a lot more. Every couple years I'll check in and see what's new in traditional adventure games, and I've always been disappointed that they're rehashing the same old tired gameplay.
Today, the best elements of adventure games are everywhere. Just about every platform type game (Ratchet & Clank, Zelda, Spyro) or RPG (Diablo, Anachronox, Oblivion, World of Warcraft) has large amounts of traditional adventure elements in them - dialog, puzzles, inventory-unlocks-content, etc. They just aren't the primary gameplay, and have minimal content attached to them, so traditional adventure gamers probably won't get into these as much. This disappoints me as well.
In the future, I think it will be in the casual space that adventure games make a comeback. It may take a while, though. Most attempts I've seen so far have been decidedly hard core, covered with thin casual dressing. I give it three to five years before I see a good approachable game that I'd call primarily "adventure".
- What advice can you give to someone who is interested in pursuing a career in game development? As a game developer, what advice can give you about team and project management?
- I can only offer advice to engineers. No clue what it would take for an artist or designer to enter the field. Engineers are easy, too. Just make a cool demo, or be part of a team that makes a cool demo or mod. Be smart and flexible, don't marry your tools. Learn current technology but don't obsess about it, focus on making fun and not fancy algorithms. Don't be either overconfident or sycophantic. And get your name out there. The industry is always hungry for smart people that have shown they can deliver. But you must have demonstrated experience in some form - a simple resume with no shipped titles or finished demos or mods that someone can play will not get very far. You must have an online, playable demonstration of what you can do. Oh and use a spell checker, at least until you get the job.
Regarding team and project management, that's a giant subject that totally depends on team culture and makeup, the demands of the market you're working in (making games for cell phones requires radically different management structure than Xbox 360 titles), and many other things. All I can generally recommend is to treat people with respect and honesty. Love the people who are solid and do good work, and fire the rest.
- What can we look forward in you over the next 5 years?
- I'm getting married this summer, so in the next few years I'll probably get distracted by a baby or three. Other than that, I'll be doing more of the same. :) Do some more lectures, ship some more games. I'm expecting a couple huge revolutions in game design and technology to happen in the casual space within this time, and I intend to be a part of it!
- Thank you for this interview. We look forward to hear more about you and your endeavor in the future!
- Thank you!