The Process

All games start out as an idea. Without an idea, what would you write? Just random code in the hope that sometime in the future it might turn into a masterpiece? Although this might work for abstract painters, software developers can't do anything without a plan. However, just because you have the idea doesn't mean it's time to start writing code.

One of the biggest mistakes I've seen and continue to see no matter how many times I tell them otherwise is developers who jump right into coding. Unless the code you're writing is so brain-dead simple that you can do it in minutes, it is never a good idea to begin writing code with just an idea. Jumping right into coding without a plan will cause you nothing but pain, rewrites, and a longer development time.

Before you write your first line of code, you should develop a design specification (spec for short). This document should contain specific information on how your code will be designed, how the objects will interact, and the various properties your objects may have. Starting coding directly without this step will most times lead you to solve the wrong problems in most cases. Although it may be possible to solve these problems quickly, it doesn't do you any good if you're not solving the correct problems. You'll end up spending more time making your solutions work for the real problems later.

You should also spend the time to commit the idea to paper (or computer file, if that suits you better). You don't want your idea for the next great computer game to be so vague that you can't figure out what problems need to be solved for your design specification. Try to make sure your idea is completely fleshed out. Show your friends, have them ask questions, and ensure they are answered by your document. If your idea is something like "Write an awesome first-person shooter game," chances are you haven't done enough thinking about your idea. What kinds of weapons will you have? What kinds of game modes will be available? Will it be multiplayer? You need to have your idea well thought-out before proceeding.

If you are attempting to develop a game for a commercial publisher, you will also need to develop a game proposal. Although it is important to have the documentation mentioned previously for this proposal, in the majority of cases you will also need a demo that is pretty far along. It is rare for a publisher to give you the green light on a project based only on an idea (unless you already have the reputation as a top-notch game developer, in which case, you probably don't need to teach yourself the beginning 3D programming this book covers).