Post Mortem of HanniBall
Nearly two days later, I can call HanniBall a success.
After two finished LD’s, I gave up during LD31 and missed LD32.
But this time, I managed to finish a game. Woohoo \o/
It’s a little 2-Player Battle Game to find out, who the real Hannibal is.
What went wrong
- I was not prepared enough, I struggled the first 2 hours with bringing a simple box with a light to the screen, just because I forgot, how to do so. And it took forever to record the few seconds of annoying music.
- I didn’t invole people for playtesting or give me some advises or ideas. Aftwards, friends came up with nice little ideas, which would have made the game real fun to play.
- I lost a couple of hours doing nothing. Or watching a movie. I could have use my time for creating properly 3d models – Now, I have only balls.
- The floating island is dynamically generated, because I thought it would be more fun/ challenging. But no one noticed it. I would have saved hours, if I simply would have modelled it with blender.
- Although I was streaming the whole progress on twitch, I had nearly no visitors. I should have advertise this more, and do more socializing on IRC, Twitter, etc.
What went right
- I chose the most simple idea. And it was one of the first idea – playing with the name “Hannibal” (my girlfriend is currently reading the novels) seeded into this game. No, it’s not shoehorned.
- To chose plain JavaScript and the incredible THREE.js framework was the right decision. I loved to come up with such a nice result despite not using a full featured game framework
- I finished the gameplay development first, and had afterwards enough time for building up the floating island, making music and to add the postprocessing.
- It’s a game I can play with my 7y old daughter – despite the theme
- My PC was capable enough to do a FullHD stream to twitch while developing with WebGL stuff. Nice baby
Tools
- JavaScript, THREE.js
- THREE.Terrain
- gimp
- bfxr, bosca ceoil, audacity (see the tools section)