2. Evelyn's Idea
After our beta presentations, we had exactly six days to prepare for our final showing of the game. There was still a lot of work that needed to be done, and we were beginning to wonder if we were going to even make a significant dent in it in time. It was Evelyn, my fiancee and 'quality assurance lead', that saved the project that night.

At her suggestion, Rolf and Chris came over to my place the next morning. For the most part, with the exception of class time, during which we were together anyway, they never left until final presentation. Evy kept us fed and caffeinated, tested new builds as fast as we could copy them onto her PC over the network, and played the all-important role of team cheerleader. More work got done on the project in those days than in the two weeks between alpha and beta.

3. The Amazing Never-Ending Development Cycle
Guardians of Neverwood was playable at our final presentation in June, but it wasn't the showpiece we had wanted to produce. Much work needed to be done. In July, Rolf continued the newly-established tradition of practically living at my dining room table, and we cranked out a significant amount of polish, and a whole heap of new features, in a very short time. Without the added pressures of class, the mountain of paperwork we were expected to produce daily, and the looming deadline, we were able to re-approach the project with a relaxed, thorough candor that was simply impossible to achieve while at Full Sail. With only two months to go from an empty project to a finished game, we had been in crunch time from day one.

4. Microsoft Saves the Day
Due to our trials and tribulations with animation (see below), we were forced to make some hard decisions. We had heard through the vast number of people that we spoke to about the problem that Direct3D and its D3DX libraries did a lot of the work associated with animation for us. The fact that the Direct3D .X model format is one of the very few we could find that even supported the type of animation we wanted to do sealed the deal. The problem, however, is that we were less than a week from alpha, and were faced with the spectre of completely re-writing the entire graphics pipeline of the game, using an API we really weren't familiar with, and one for which little support is available at Full Sail.

The three of us had a meeting, during which we pulled in several instructors, lab assistants, and former graduates, and after about an hour of discussion made the decision to pull the trigger. For four days, I didn't sleep. Then, Rolf and Chris worked some debugging magic to help me out, and we actually managed to restore the game's graphics capabilities to match the pre-conversion status in time for our alpha presentation.

Direct3D allowed us to do some nifty things we might not have otherwise pulled off. The wrapper class we developed for it was intuitive enough that even the team members who hadn't spent the last four days immersed in the DirectX SDK documentation could use it without difficulty. The .X model format proved simple for our artists to work with, and while we didn't have animation completed at the final presentation, we were a lot closer to it than we would have been with OpenGL. We're very glad to have been exposed to the Direct3D API, as we feel this knowledge will be invaluable to us in future projects.

5. Play it Again, Sam!
After presentation, one of the first things we did in 'developer overtime' was add a series of over thirty unlockable features that can be earned through gameplay. Instantly, a game that takes five minutes to beat became worthy of several hours of enjoyment. Unlockable extras are fast becoming a staple of a value-driven game consumer's expectations, and we felt that it was important both from a polish standpoint and from a demonstration point of view to get at least a few things in. We were shooting for eight to ten, and the number just kept skyrocketing as we had more, and better, ideas. Some of the features we had listed in the design document were rendered impossible by unavailable art resources, but we balanced them out with a host of new ideas.

The real part that went right with the extras was the fact that we didn't kill ourselves to get them in. Most of the things we added were simple, under 50 lines of code to implement. Some, like cardboard mode, were achieved simply by tweaking a value or turning a boolean on or off. Our game was architectured from the beginning to make it easy to do those kinds of things, and it really paid off at the end.