[Prior to LD31]
I payed attention to the voting results from rounds 1-4 and tried to predict the theme. My prediction was unicode snowman, and if not that, artificial life. Before snowman results, it was looking like artificial life for sure, so I started brainstorming on that. The idea I came up with was a robot that escapes from a scientists lab, the robot runs out into the wintery world and jumps into a snowman to hide. The game then jumps ahead in time to a boss battle between you (robot inside snowman) vs the scientist. After I saw how well snowman scored, I started thinking about taking that idea and making snowman the primary theme. From that I settled on the idea of a fighting game sort of like SSB. Instead of damage making hits push you farther, I decided to make damage reduce friction, and have the battlefield be ice. I wanted to make controls reasonable to be playable on a single keyboard because I don’t have a joypad/controller to test with, so I ended up with the following control scheme:
up: jump
left/right: run forward/back, double tap to dash
down: guard on ground or smash in air
Damage is done via collision rather than having buttons to attack with.
The only character would be snowman, and when he dies little kids would come with balls of snow and rebuild him quickly for the respawn animation.
Keep in mind, this is just my brainstorming before the theme was announced.
In the final voting round, I voted for artificial life, unicode snowman, and everything on one screen, because my idea would work for each of those themes.
[LD31 begins]
To my surprise, everything on one screen won out, but my idea still fit, so I got right to work.
I started out trying pomodoro technique for time management. I began by creating a list of tasks that would need to be done for the game, keeping in mind how many sets of pomodori I had for a day. Next I ranked each item on the list. From my previous attempts at ludum dare, I have learned to go into it expecting to cut features, so prioritizing what to implement when is crucial.
[Physics hell]
Late at night my code base began to grow complex enough to have physics bugs crop up, and I realized how hard writing the physics code for this actually is. I’m not using a physics engine like Box2D or Chipmunk, and I have never made a game like this before.
During this time I realized that hard debugging is best done without the pomodoro technique.
I think it is better to slow it down a gear and avoid letting frequent breaks interrupt your flow, although the pomodoro technique was definitely a huge win in the beginning when it helped me organize & plan, keep track of my progress, and take care of my health with frequent breaks.
The vast majority of my development has been spent fighting with my physics code. In honour of this great battle, I have decided to name my game…
So guys, physics is KINDA HARD

So far I have managed to implement forward and back dashing, jumping, collision damage, deaths, respawning, players face each other, life display + damage display, a bit of quick artwork, and a couple sound effects from bfxr. I’m currently trying to get the bugs ironed out of forward running. I don’t expect to have time to polish much, but my goal for this ludum dare is to try get all the core gameplay features implemented before the deadline 