After I had finished the original Faery Tale Adventure, I went on to create a few other programs for MicroIllusions such as Music-X and Discovery 2 (the latter being a collaboration with Joe Pearce).
I had grown increasingly uncomfortable working with MicroIllusions and had started to distance myself from them, primarily by doing contract work for other companies. I teamed up with Joe, who was now also my roommate (as well as Greg, Bonnie and Allison), and together we worked on a bunch of projects for Commodore, including the Amiga Installer, the Amiga MIDI driver (CAMD) and CDTV Welcome disk. We used to joke that “I would start projects and Joe would finish them”.
MicroIllusions was floundering pretty badly by this point, Jim Steinert had turned over management of the company to his marketing director David Boyles. I won’t say much about this except that I never received another royalty check from MicroIllusions after that. The justification for this was that there were no sales of any of my products (although they kept asking for bug fixes and updates to Music-X, some of which we delivered). Boyles renamed the company to “AMP entertainment”.
The Dreamers Guild
It was now 1991 and I was starting to get a little tired of being an independent contractor, mainly because I had been badly served by the general incompetence and lack of imagination of the various game companies I had worked for. And I knew a lot of other independent developers who felt the same way.
It was Robert McNally who suggested the idea of forming our own company. We had long conversations (at the Good Earth restaurant in Northridge) about what kind of company we wanted to create and how we would get it started. Joe Pearce and Allison Hershey were brought into the discussions early on, as well as some friends of Robert’s like Dan Pinal, Charles Wilson, Walt Hochbrueckner, and Paul van Schlifgaarde. We also kidnapped Joe Burkes, a young and talented programmer who had impressed me with his sprite editor demo at a user’s group meeting.
It was Robert’s idea that we call the company The Dreamers Guild. The inspiration was taken from the trade guilds of old, where apprentices would learn their skills under the guidance of master craftsmen.
A complete history of the Dreamers Guild would take a book, and this is really about The Faery Tale Adventure, so I’m going to skip over a lot of bits (like the Northridge earthquake!). What’s important to understand, however, is that we were a bunch of starry-eyed idealists with no cash and no resources. We managed to bootstrap the company and get some initial contracts and money coming in on the strength of our reputations, but we were always cash-starved. Our “sweat equity” company was always more sweat than equity.
The biggest mistake we made was grabbing a hold of our own destiny, and then immediately handing control of that destiny to someone else.
In 1994, we had finished Inherit The Earth: Quest for the Orb for New World Computing and were looking for a new project. The company had grown to about 50 people at that point, but we were still losing money on every project, despite paying low wages. Our employees were extremely devoted to the welfare of the company (a result of our egalitarian policies — for example, everyone who worked for more than 2 years was a board member) and worked their hearts out, but it wasn’t enough.
(More pictures here. Thanks Joe!)
At this point, the day-to-day management of the company was in the hands of Bill Simpson and Richard Pferdner. Neither Robert nor I had much influence any longer. Richard and Bill strongly believed that if we could just get to a certain scale, we would be profitable. Robert and I disagreed — the reason we weren’t profitable is that we weren’t able to execute the contracts we took in a timely fashion, and that was because we grew the company so fast that we weren’t able to give each project the strong tech leadership needed. This disagreement degenerated into Machiavellian political maneuvering, and Robert would eventually leave the company as a consequence. In my case, I was too busy writing code and being a technical and creative lead to pay too much attention to what was going on in the front office.
Wait, someone wants a sequel to Faery Tale?
Somehow — I don’t remember the details — we got a contract from Encore Software, a relatively new publishing house, to do a sequel to The Faery Tale Adventure. I had toyed with the idea of doing a sequel in the intervening years, even had sketched out some code, but I was too busy with other projects to make a serious effort.
But now that we had a contact in hand, that was a different story. We immediately set to work on the new project. This would be the largest and most complex thing I had ever done.
One of the first things we had to do was clarify who actually owned the intellectual property for Faery Tale Adventure. Recall that I had done the original game as a contractor. As part of a complex deal involving some Music-X bug fixes and upgrades, I took back ownership of Faery Tale from AMP.
Defining the Look
One decision made early on was to use an isometric tile system — tiles on a diamond-shaped grid, first seen in the arcade game Zaxxon. We had used this successfully on Inherit the Earth, but I wanted to double the resolution and use tiles that were 64 pixels wide instead of 32. Since our target platform was MS-DOS with the VGA/VESA graphics standard, we knew that we could develop for a 640 x 480 screen instead of 320 x 240 as we had in previous games.
Another innovative idea came from our art director, Brad Schenk. Brad had been experimenting for several years with 3D rendering, but didn’t like the plastic look of Phong-shaded models (material shaders had not yet been invented). Instead, he created animated, 3D versions of all of the characters and monsters in the game and then rendered them in grayscale. We then had our art team hand-paint over the top of the 3D rendered images, giving them a more “painterly” look, while preserving the preciseness of the 3D animations.
Another artistic decision made early on is that all objects and characters would have black “toon” outlines. I felt that this made objects easier to see, and made it visually distinct from other, similar-looking games.
Because we were targeting a 256 color display, I developed a special palette for the character artists to use, which divided up the available colors into groups of 16. Colors 0–15 (dark red to light red) would be used for the shirt, colors 16–31 (dark blue to light blue) would be used for the pants, and so on for each part of the character’s body all the way up to color 255. The game engine could then remap these colors dynamically through a lookup table as the sprites were drawn. This allowed us to “recolor” characters — so for example, if you were attacked by a band of goblins, each goblin could have a slightly different skin tone. A sprite representing a farmer or a princess could be used dozens of places in the game world, and have a unique appearance in each instance.
Engine and Tool Development
The basic rendering code was developed fairly quickly, requiring just a few weeks of work by me. David Richter and Evan Olson were hired to assist me in developing the game engine and art tools.
As part of this work, I wrote a new pathfinder algorithm, which allowed characters to move from place to place in the world, avoiding obstacles in their path. I worked hard to make the character’s movement seem ‘natural’ — for example, if there was a trail or path leading to the desired destination, the character would prefer that route to cutting across the grass or flower beds — unless the advantage of taking the “shortcut” was too much. Similarly, I wanted the characters to avoid making sharp turns like a military drill team, and instead prefer gentle, curving turns.
An important part of the project was artists’ tools — the tile and sprite editors. One of the things that I had learned from Inherit The Earth is the idea of an asset pipeline — that getting a large team of artists and programmers to work together isn’t just a matter of tools, but training people to use a set of procedures in handling the files. So for example, when a sprite was finished, it would be processed using certain programs, and then copied to a particular shared folder on the network file system.
As a result of this, we were able to re-use all of the expertise we had build up during Inherit the Earth. We were able to start producing characters and terrain almost immediately.
SAGA turned out to be surprisingly powerful. One of our “scripters” (I guess you would call them level designers or content producers today) was Steven McNally (brother of Robert and Michael), who could make SAGA do tricks that even I, its creator, had not anticipated. For example, he created an interactive debugging utility that used SAGA dialog-trees to present an on-screen numeric keypad that would let you teleport to any location.
In fact, one of the things that I had to keep telling the scripters was to report bugs in the game engine instead of using SAGA to code around them!
SAGA, the game engine and the artist tools were the most successful parts of Faery Tale 2, in fact we re-used the exact same technologies on Dinotopia (which started later but finished first).
I need to say something about the cover art for the game. For years I had admired the work of Greg and Tim Hildebrandt, and I was lucky enough to attend a panel at Comic Con where they were speaking. After the panel, I walked up to Greg and said “So, how much would you charge to paint a cover for my computer game?”
The price turned out to be five grand, which in my opinion was a bargain. So that’s how I managed to get a Hildebrandt cover.
Content Creation Troubles
Like the original Faery Tale, I wanted the sequel to take place in a huge open world. However, the problem was filling that world with enough interesting content to keep the player entertained. At the time, I believed that we could solve that problem by throwing money at it — by hiring enough talented people to generate all of the content we needed.
We hired a designer and several writers to help flesh out the story line and the various plots the player would run into. Unfortunately, these were not experienced game designers, but rather game enthusiasts who liked writing. None of them really understood the technical limitations involved in creating a game world.
You see, all large world games — even modern games such as Skyrim or Dragon Age —rely on clever re-use of art resources. The landscape may be covered with thousands of trees, or bushes containing valuable herbs you can collect, but the artist doesn’t draw each individual plant — instead, they create a few variations and then replicate them across the world, using subtle tricks to try and disguise the fact that it’s the exact same artwork in each case.
Similarly, when you are in a dungeon, often the designer will create a wonderfully-detailed stone corridor with mossy stone walls, and then use that same corridor many times within the same dungeon.
Unfortunately, one of the designers that we had hired didn’t like this idea of re-use, he thought that any cutting & pasting of content made the game repetitious and stale. I spent about three months arguing with him — he would produce these giant design documents, every one of which would require all-new artwork, not only new terrain but new character animations as well. I simply could not get him to create designs that leveraged the work that had already been done.
Another one of the writers thought that it was funny to embed Monty Python references in all of the character dialog, and no matter how hard I tried I could not get him to stop doing it. I blame myself for this — I was too unsure of myself and lacking in people skills to really put my foot down.
You see, we knew that the project was behind schedule and over budget — we were going to run out of money before even a fraction of the game world was populated. And we were afraid to fire anyone because we felt that it would make the project even later. Everything would be OK if we could just get them to understand…!
I think that about thirty people worked on Faery Tale 2, and roughly half of those were artists (the rest were mostly engineers, scripters, designers, writers and musicians). I was running around frantically trying to coordinate everything — some people actually started slapping post-it notes on my sleeve to communicate with me as I whizzed past their desks.
We also had problems on the scripting side. The people who were in charge of hiring (Bill and Richard) had a tendency to treat coders as interchangeable units — if we didn’t have enough programmers on a project, well then, just hire some more! Unfortunately, the people we had to work with were very much a mixed bag, some were young enthusiasts with lots of talent but no experience, others were coders that had worked in other industries and were feeling burned out. Not everyone was up to the challenge.
Out of those thirty people, there were at least three (I won’t name them) who had an overall zero productivity — it was not that they didn’t work hard and put in a lot of hours, but every single thing that they produced had to be thrown away and redone by someone else. I learned an important lesson about apparent vs actual productivity.
Richard Pferdner was lobbying very hard to reduce the scope of the project. But how? We had already created the game world, re-drawing the map to make it smaller would cost additional time and effort.
The Bitter End
By the end of the project, I was working 12 and 14 hour days, 7 days a week. We knew that The Dreamers Guild was facing bankruptcy, and I was trying to get the game done before that happened. Even after we decided to shut down the company, Evan and I worked unpaid for three more months to try and deliver something.
Even worse was the fact that our sponsor, Encore, was itself facing bankruptcy. The game was eventually pushed out the door by Ignite and K.O. Interactive, but it was a pale shadow of the game I had dreamed of creating. It did not do well.
For me, I lost my project, my company, and (around that same time) broke up with my girlfriend. I fell into a deep depression that lasted about fifteen years. I remember one episode in 1997, standing in a parking lot late at night, where I was feeling so bad that the very air surrounding me felt malevolent.
I could still work and be productive — for me, work is an anesthetic — but I would never again be able to emotionally commit myself to a project so strongly. I was burned out.
I went on to do a few more game projects for other companies, but in each case I was just a software engineer with a minor role, and none of the games I worked on were successful until I joined Maxis in 2002 and worked on Sim City 4 and then later The Sims 2.
In 2007, I decided that I’d had enough of the games industry (I was working for Electronic Arts and I was not happy with them). Michael McNally was working at Google at the time and he, along with Guido van Rossum (creator of the Python programming language), invited me to come work at Google, which I did.
My depression has been under control for the last 5 years, although I still suffer occasional bouts of melancholy and loneliness. Fortunately, I have a lot of creative projects that keep me busy and my mind occupied. My health is good overall, and I am financially secure. I have a lot to be thankful for.
You can see some of my more recent musical creations on my SoundCloud page.
At the moment I am working for an incredible startup called Nimble Collective, which makes web-based tools for 3D animators. I figured that since I had always enjoyed writing tools for artists, that I would enjoy doing this, and so far that has turned out to be true.
Will there ever be another Faery Tale? I attempt to address that question in part three.