Create-ivity

AI Wars is done

It’s finished. The Flash frontend and Neko backend are done, and my classmates have produced a nice AI client for demonstration purposes. The whole package is roughly 0.5 MB.

To make things a little easier, I’ve written a batch file that starts both the backend and the frontend. All that’s left then is connecting the AI clients and starting the game.

The gameplay

Continue reading AI Wars is looking for AI agents!…

Lacking inspiration?

Recently there have been some threads about ‘how to get inspiration’ on the Gamedev.net forums. Sometimes it feels as if every good idea is already taken, and that every addition or twist you come up with turns out wrong.

That was written in a game design board, but I’ve also seen it come up many times on level-design forums. Take Counter-Strike for example: so many custom maps have been created that there’s virtually no original setting left. Or Half-Life 2: try to come up with an original physics-based puzzle for once. That’s harder than it looks.

Here’s how I handle it

Continue reading How to get inspiration…

Here’s a small image hinting at the surprize I wrote about in my last post. It’s a puzzle game, it’s almost completely finished and it includes an editor. I’ll post more info about it later, once it’s fully playable and polished.

Meanwhile, I’m finishing up AI Wars - it’s almost done now, except for a few communication issues. I’ve also written a few additional Python tools. Since I’m painting quite a few tilemaps for Aural Fighter, I figured a few tools wouldn’t hurt. Today I added one that takes an image, asks for the desired tilesize, and outputs a serie of unique tiles as separate images. It’s using Python Imaging Library (PIL) to load, compare and save the images. It then optimizes them, using PngOptimizerCL, the command line version of PngOptimizer. The tool took me about an hour or two to create, and it’s less than 100 lines of code. And, except for the optimization phase, it’s quite fast, too.

Something has bugged me for quite a while: Aural Fighter reacted slower and slower the longer I played it. I was absolutely sure it had to do with the rendering code, but even a bare-bone test game slowed down after a while. Confused, I ignored the issue for a while, but I realized I couldn’t release Aural Fighter with such an issue unresolved, so I just had to fix it.

But whatever I tried, nothing seemed to cause the problem. Finally, I decided to create a minimum piece of code that still reproduced the problem, to post on the HaXe mailing list. While doing so, I found out that the game only stopped respondeding after a lot of keypress events had happened, not just after some time. I was confused: was I handling input in a wrong way? It should be a perfectly normal method… but while stripping away unrelevant code, I finally figured it out: I had added another keyListener in a pre-loader frame, one that was never removed afterwards. This one started to intervene with the other keyListener. Removing it finally solved the problem.

On a side-note, I’ve added shadows to all enemy fighters and I’ve drawn a few new enemies and tiles. I’m also working on another game - as a school project - so Aural Fighter is proceeding slower as desired, but the experience I gain from that project should be pretty usefull for later games. I’ll write some more about that project later.

Level 1, but now with updated art. It’s still not really final - the large trees could use some work, the jungle borders are still somewhat flat and the water is too obviously tiling - but it’s definitely moving in the right direction. It took me quite some work to make sure the enemies were sticking out enough from the background. I’m using more saturated and bright colors for them, while I try to keep the background darker and less saturated. It surely takes some testing to get right.

So, that’s what I’m working on now, and the more I do it, the more I get the hang of it. Pixeling is fun and with my in-game editor in a usable state, creating these levels is actually quite enjoyable. Let’s hope it shows in the final product!