tisdag 6 november 2007


First: Sorry for the slow updates in the Swedish blog, when things speed up in the creation process, so will the reporting.

Lately, I've been reading frequently about object management in different types of communities, books, articles etc, and I am thrilled to see that there is no real universal solution to the "problem". On one hand, you can do it quick, simple and well adjusted to your project. On the other you can spend a bit more time and have it more general. Really, what you choose depends on what you need, and this is where it becomes tricky. When you start a project you may not know all the details of the completed game and you will not know entirely what you will need your engine to be capable of. Some might say that this has to do with being able to plan ahead, but some questionmarks will remain. As for the object management, choosing a more general solution basically means it will be harder to create more complex objects, like tanks, cars etc, that need several types of physical components all tied together, the communication between them might get a bit mixed up. But if you choose to use an inheritance tree in a situation where you need many similar yet different types of objects, you will end up with far more classes than you need, and a lot of copied code.

What I'm trying to say is that at this point in the developing process, we really need to decide what the engine should be able to handle. And this is always hard. Everytime you start out, you want your compiled code to be totally independent of what the specific game will be, but in reality, you will probably have to sacrifice either speed or reusability or something else.

More questions - fewer answers this time.
Tomorrow: Sweet Nineteen, wish me a happy birthday and stay tuned.

Inga kommentarer: