The Temple of Elemental Evil Wrap Report

Development Timeline The contract was signed and work officially began on February 1, 2002, and the game was originally scheduled to be completed on June 1, 2003. About one full year into the schedule, we amended the deal to add two additional months to update the rules to 3.5 Edition D&D and add additional content, which put us at an August 1, 2003 completion. We delivered the final product on August 30, so we came in ultimately at almost one full month late. I think with the bugs we have discovered since then, we should have delayed this for at least one additional month beyond that. Our game would have been solid with an October 15 completion date.

The "why" part of this could probably fill several pages worth of notes and such, but the main thing to throw in here is that there were only two real months in the schedule for full-team testing. And with a game as complex as ours, we obviously needed at least two more months in our schedule just for testing. We delivered our first full-content product about August 8, so in one sense we were only a week late, and in another, we only gave ourselves 22 days to test the final completed product.

Changes and Enhancements I would guess the biggest thing that we changed from the original design was the change to the 3.5 Edition D&D rule set. When we discovered that Wizards of the Coast was coming out with the new book very close to the same time as our game was to be released, it just made sense that we upgrade our game accordingly. I think all parties involved agreed that this just made sense.

Another thing that was not in the original design or milestone list was time in the schedule for the tutorial to get done, even though a game with as complex rules as ours really needed one. A few very independent team members here just created this on their own and in the very wee hours of the night / morning. It came together rather quickly ultimately, and even though it would have been nice to add a few more lessons, it is incredibly polished and useful. Another thing added was the help text and roll history available at any time in the game. With the hypertext links, in some ways, the in-game help text is easier to use than reading the manual or the core books. This, of course, meant a lot of typing in text and summarizing important chapters, which we ultimately hired someone to handle specifically because the sheer volume of text put in was so huge. Another thing that comes to mind that we added were the opening vignettes. These came together very early in the development process, but required wholly new maps to be made for all nine alignments, plus dialogues and such to be written for them, none of which were in the original schedule. The vignettes basically give your party, based upon the alignment you have chosen, your reason for coming to Hommlet and an opening quest to get you started in the game.

There are a lot of other things added. For example our original design required us to implement only 82 spells. We ultimately delivered 223.

Major Challenges I'll mention here a few of the bigger challenges overcome from each area of development and then tell a story from one of my personal biggest challenges.

For programming, I think the biggest challenge was getting the party pathing in and working. We had a milestone for pathing very early on, but we had sort of fudged that one together using existing code from Arcanum and making it look pretty good... but the intention was always to rip it out entirely and come up with something new. So, a new milestone was added called "complex party pathing" for several milestones down. Well, when that one hit, for some reason we got that past Atari too, even though it really wasn't as complete as it should have been. Ultimately, pathing was being worked on and improved up until the very end... and there are even additional enhancements in there now from the patch.

For art, the map of Hommlet turned out to be a much bigger project than originally planned. I shared an office with two of the artists, Peter Delgado and Lucas Feld. Peter, who was given the Hommlet map, was scheduled for only one month on it, but ultimately needed three months to put the huge, beautiful map together. In hindsight, we may not have needed such a large map, or perhaps needed better guidance around the map's locations, but you can't argue that he did an excellent job at bringing that map to life. Because of the scheduling mix-up though, it was known as the "Hommlet Debacle" and the schedule was called the "Marcia" schedule, in reference to Lucas often being commended for finishing his maps early, to which Peter finally said, "Lucas, Lucas, Lucas..." as in the Jan Brady soliloquy from the Brady Bunch. The good news here is that it is all transparent in the final project, because everything on the art side was completed and on time.