This has been an interesting project for me so far. In all of the years I've been programming, I've never experienced anything quite like this. On the surface, it's such a simple thing. To date, it's just a handful of D modules with a very limited interface. It's that latter bit which has been the challenge. I never realized how tough it is to design a minimal interface.
The goal with this project from the beginning has been to enable getting a game off the ground with Allegro and D quickly and painlessly. No wrapping of the Allegro libraries, no complex abstractions. Just a collection of routines to package up a bunch of boilerplate so that you can get to the game code almost as soon as you pull up your IDE for the first time. I thought it would take a few days of dinking around. After all, I've got years of experience under my belt. I could do this in my sleep, right?
From the very beginning, I stumbled over which approach to take with the interface: free functions or static class methods. Ultimately, I settled on the latter approach. Then I started implementing the outline I had sketched out. As I progressed, I realized it was a bit overly complex. Did I really need a templated resource manager for a finite number of known Allegro resource types? So I scrapped the first pass and went back to the drawing board. More than once. It's been a while since my last major rewrite, but I think there were a total of four. Some of that process was documented here on this blog.
Over the past couple of months, I've not done much more than a bit of tinkering here and there due to my busy schedule. But now that I've come back to the project in earnest, I can say it's looking good. I've got it to the point where it's ready to use for a game. And that is, in fact, my next step. I'm going to port over the A5teroids demo that ships with Allegro. And probably implement a few other old skool games just to get an idea of what else I need to add to the core.
The weird part for me is how much refactoring and thinking went into getting such a pitifully small amount of code put together. Of course, it would have been much easier had I been more familiar with Allegro in the first place. I used it a lot back in the late 90s, but that doesn't count given how long ago it was and how much Allegro has changed. Still, I will never again approach a "simple" interface with the idea that it will be simple to implement.
As the project moves forward, I'll be working on some utilities that could be useful for any game project, Allegro-based or otherwise. I'm looking forward to getting into some of the features of D I've not yet had much experience with, like ranges. Before all of that though, once I get a couple of small games knocked out and know for sure that the current interface achieves my goal, I'll do some proper DDoc comments and put all of the source up on Github.
