With my last post on a forum over a year ago (and last journal article exactly 2 long years ago today), the longing finally took over when I went on break from school. First came the boredom, then the realization that I was still (very) interested in making games. A couple mini-projects to jog my memory in DirectX, plus some adventures into XNA got me hooked once again.
Then the lurking began, and the main page RSS feed, and finally it has come to this.
Anyways, figure I'd get my GDNet+ worth, and try and start frequenting the forums once again. I return to game development with 3 years of Computer Science and Computer Engineering pounded into my head for better or worse. In that time I have finally worked on some large scale team projects, worked 2 summers writing embedded code with a team of 20+ other gents, and finally have a new perspective on coding and especially, documentation.
Documentation and Planning is actually where I feel I have always been lacking. I had always seemed to start projects with lofty goals, but no concrete plan, and my code suffered until it was virtually unusable/unreadable. Well, with a course in Software Design and Documentation fresh on my mind, I set out once again on the road to game development.
I decided that I wanted to make a basic rendering framework, nothing spectacular, just something to use, modify, and expand for my own personal projects. I looked at all of my old, dusty coding projects trying to decide where to start. Then, like a true CS nerd, I turned to UML. Hate me if you must, but I have always been the kind of person that needs structure and a nice neat chart to be productive. UML, it its more basic forms, was exactly what I needed. Just last Friday, I came up with this gem:
It is still a work in progress, but every class outlined already has at least minimal functionality and my test app works flawlessly. It may not seem like much, but I accomplished this in about 8 hours total of coding time. Most of my previous attempts at a simple, flexible renderer have not even touched this much functionality, and the ones that did took me weeks to implement. This framework allowed me to seamlessly hide DirectX from the client, while still making it very easy to use. I set up a spinning, Vertex/Pixel shaded mesh, with less than 60 lines of code. The engine itself is very neat and I feel like there are probably more comments than lines of code.
Hopefully I will be able to have enough implemented to start working on a simple game by the end of the week. And to all my buds on gamedev, you will definitely be the first to beta test... err... play it.
