Broke ground on the October Project a few days ago. So far, it is going well.

Unfortunately, it will be some time until this can do anything as sophisticated a full-blown Minecraft clone, so my first target "test game" will be something with a simpler visual representation and UI. Details still being determined, but likely some kind of top-down, one-layer-at-a-time sort of system.

The core parallel "tick" system is in place, as is the basic representation of the cuboid data structure (the 32x32x32 base units making up the world). The sketch of the design for how mutations are applied, how cuboids are asynchronously loaded/generated, and how data can be queried/observed are all in place. I need to add some more complex mutation types, some basic mechanisms for how a user would move around an interact with things, as well as the basics of the network layer next.

I am concerned that the actual UI and graphics will be a great deal of work, even for this simple design. It also kind of bothers me that this system isn't being prototyped in a fully top-down way, as a result of wanting to delay this heavy work. Still, it is my hope that I can build out a reasonable automated testing interface as a sufficiently representative "top" from which to continue prototyping "down". At the moment, however, much of what I am building is frustratingly "bottom-up", so I hope that I am not walking blindly in the wrong direction.

Work continues,
Jeff.