Play TerraBellum here
As is traditional with my ludumdare entries, my most recent attempt was a bit crap, but was a good learning experience for working with Unity3D, the engine I implemented the game with.
I spent forever trying to come up with a decent idea for the theme ‘Tiny World’. Initially thought I might try and create a happy ecosystem game but decided I really wanted some explosions instead, so I went for something a little more violent.
What went right
The game looks reasonably nice. Very close to how I envisioned it. (Initial visual mockup)
Core gameplay mechanics were implemented in time. The player can build and edit their island and then send their island off to war. Serverside code for saving, loading, authenticating etc all turned out reasonably robust (though you could cheat with a little effort).
What went wrong
Everything else. Most of what I did with this game was new ground for me in terms of Unity implementation, so took longer than I would have liked to get working. That said, I learnt a whole lot of new things, and discovered that some of my ideas of how Unity deals with basic things like prefabs and object hierarchies were quite wrong.
I used tilt-shift blur to try and make the worlds look tiny, but I don’t think it worked very well. It looks more like a shonky depth of field, and it doesn’t play nice with the particles. I think the models were too low poly, and there wasn’t enough variation in distance to make the tilt-shift work. Perhaps if I’d gone with a more realistic look for the islands it might have turned out better.
The editor is incomplete and is buggy. Islands can be built in invalid configurations, sometimes money seems to mysteriously disappear or be created from thin air.
Audio didn’t get much attention, last minute job with (the ever useful) bfxr, and I didn’t have a chance to even attempt music (probably for the best).
The battles. Damn. What a joke. Needed at least three or four more hours to sort the fighting out. If I’d started immediately on this game idea I might have had time, but who knows. The turrets don’t aim and just shoot (almost) straight forward, with a little random deviation. I thought people would just modify their island designs to compensate, but instead they just quit when they realise their guns aren’t going to hit anything. The battles time out after approximately 30 seconds, so they shouldn’t have to wait long, but apparently the lack of feedback suggesting that a timeout would occur doesn’t help people wait patiently.
Mostly though the game lacked sufficient feedback. It’s not obvious what is happening within the system unless you wrote the system.
The Future
TerraBellum deserves to be cleaned up, debugged and made available for play. I’d like to play the game myself even, so it can’t be completely worthless. Perhaps after ludumdare voting ends the pain of coming last will motivate me to fix the game up (because LudumDare is all about winning and losing).
I’d quite like to expand the game a bit further too, add some additional complexity through additional building types, the ability to upgrade turrets ala tower defence.
). Blender importing doesn’t work well for me with new versions of Blender, but exporting to .fbx from Blender and dragging the file into Unity for an instant import is quick and painless, I’ve not run into any of the problems I’ve experienced exporting from Blender to Panda3D. Unity’s .net reflection magic near-instant super-compiling is amazing (or is so the 99.9% of the time that it doesn’t bug and require a unity restart) and makes for a tight code-test-code cycle, comparable to interpreted python with Panda3D.