Attempting physics

My cog logic was working, but I knew I would need to add more complexity to the system soon. I felt it was important to consider whether using entirely kinematic cogs was the right way to go, or whether making the cogs physics based would open up more possibilities.

I thought that physics might solve:

• More realistic cog movement (because the teeth would actually be pushing each other)
• Easy to add variation to cog shapes (e.g. mutilated cogs - technical term)
• The balls colliding with the teeth and falling down the same layer as the cogs themselves, rather than the convoluted foreground layer of “pie-wheels” (not a technical term)

I expected that a physics-based method would be too unreliable, but I wanted to try it and see for myself. 

I was right! No surprise, but cogs and gears require a lot of precision to move correctly. It’s possible this could be achieved with a strict grid system and careful design, but that didn’t seem like a time investment that would be worth the effort. Unfortunately I don't have a screenshot as I messed up my git branches in the process... so you'll have to take my word for it!