Before diving into the meat of it, here is our process for kickstarting a game.

Spend as long as possible on the idea phase. Start by going as wide as possible, shotgun all your thoughts onto paper. Don’t zoom in. Let it ferment a bit. After a while some ideas will start to stick out and you can develop them further. Liberally use Wikipedia, image search, google to feed the process.

After that, sketch it out into features. Now remove as many features until your idea breaks, and add the last one back; it might not seem like it but you have nowhere enough time for all the fluff.

In our case, we settled on a terrarium / sandbox toy of sorts mixed with a card mechanic to drive events forward based on an old concept. Since we’re not too original we’ll name it Little Fortune Planet. A nice technical challenge mixed with the potential of some nice visuals. In this case I wanted to see how the java to js (web) transpiling would deal with a non trivial non optimized game design.

To keep things relatively stress free, start with the hardest tasks first. Typically for me this means graphics and music beforehand but in this case @Flaterectomy and @MeatMachine took care of that. Phew!

We targeted desktop and web with the java framework LibGDX and an Entity Component System framework called artemis-odb. I’m a big fan of artemis-odb as it allows quick prototyping and helps keep code clean and specialised, even when in a hurry. It has a time saving fluid-API that gets rid of some of the verbosity (disclaimer, I occasionally work on artemis-odb!)

So let’s jump right into our world simulation!

Read Part 2: Gravity of the simulation, man! or play it here!