I wanted to make a cross-platform game, and it became quickly obvious that using SDL/OpenGL was the only sensible choice for a very 3D-intensive game. So I went in search of books on SDL… There really aren’t any that were of interest. Quite a few on OpenGL, 2 of the better ones are now in my library. So I read and followed some tutorials online, and did some cute little technology demos. I’m not one to not do something because it’s “hard”, but it became quickly obvious that a lot of it was at a lower-level than I had hoped it would need to be. Some very simple (in my head) things were exceptionally complex. I found a series of great articles on SDL on IBM DeveloperWorks … But the game discussed is vapor still, even though the articles are old (c1999), which isn’t so encouraging.

I then started looking at *gasp* DirectX. [TANGENT] Despite what some otherwise cool people say, Microsoft IS evil. There are people (some that I know/have met, even) that are not, but any company that tries to upset the homeostatic balance of nature and churns out crap is evil. This does include companies that claim they aren’t evil, by the way. [/TANGENT] Tangent aside, DirectX is the defacto standard in game development, so it was worth a look. At first glance, DirectX wasn’t much/any better than SDL/OpenGL. In face, some things were even MORE low-level… Then I discovered I wasn’t reading about the higher-level APIs that take care of a lot of minutia for the developer… Wow. Power, feature, flexibility and… ease. And tons of documentation.

So I spent some time (3 weeks) doing parallel development. The engine modules I was working on would duplicate everything. For every class (where applicable), there would be two: one with _dx9, and one with _sdl. For every function (where applicable), there would be two: one with _dx9, and one with _sdl. This was exceptionally useful for evaluating the ease of coding (SDL/OpenGL requires more code to do the same thing, usually) as well as the performance (on my development system, DirectX renderings were faster, and took less of a footprint, in general than SDL/OpenGL).

I am still writing all of my engine code very agnostic, so theoretically, I can swap out libraries, rewrite some engine code, but the engine API will be stable and the games built on top of them can run unchanged. But for now, DirectX is the only real way I can feasibly write a game of the size I’m working on, without it dragging on even longer than it already will….
Sometime soonish, I swear, will be the big reveal on what the hell I’m writing.

So, in all my whacked-out wisdom, I decided to write a game- a big game- in Windows… Using C++ and the DirectX system. I decided a few weeks ago to blog about the progress … and, well… I am.

So, to catch ya’ll up, I’ve written several thousand lines of C++ thus far (6412). Writing a number of parallel engines: 3D, 2D, Input, Networking, Inventory Control, and… my favorite… wait for it… wait for it… Scripting. I, uh, embedded the Perl interpreter into my scripting engine… So I can write Perl scripts to implement game-related shit. My A#1 goal is to expose absolutely as much as possible to the scripting engine, thus minimizing the amount of C++ I have to write, and maximize the amount of Perl. This has the added benefit of making the game infinitely configurable/scalable/adaptable by myself.

NPC AI is done through scripting. Assembling the world is done through scripting. Rewards and punishments… Scripting. Random encounters….. Yeah, scripting. Mission database? 100% scripting.  I’m doing a series of little test gamelets that show off various pieces in order to make sure my engines- and scripts- actually work. I may post some here for the fun of it… soonish…