Yesterday we had a discussion on Discord about a possible mechanic to progress through the multicellular stage. It was initially suggested by @C009, so I’m opening this post to discuss the mechanics better and have some feedback.
The idea was initially suggested as (from @C009):
[…]
Basically, I’ve come up with giving multicellular editor a scale slider (or maybe better a precise number selector).
Placing cells on hex would still be the same on bigger scales, but what would change is actual amount of cells in that one hex. So, say, the scale = 1, which is the scale at the beggining of the stage, would make everything be the same as is, but scale = 2 would have each cell sort of multiplied by some amount, yet still clumped together in the same hex position.
[…]
I was thinking of increasing scale making game more difficult in some way, so there is somewhat constant difficulty if the player aims to grow more, but actually it could maybe do opposite: make growing bigger easier via slider rather than placing a ton of cells that would impact performane
And yeah, scaling would also come with LOD in my idea, so it doesn’t smelt any CPUs
Who has also provided some cool concept diagrams:
Personally, I really like the idea, and it gave me the inspiration to analyze the technical details to implement this.
How would it work?
My idea of this is quite consistent with the timeline shown in the images, which results in a smooth progression from simple colonies, to tissues, to multicellular. So, the question is: how would the LOD system work?
Colonies
Initially, the colony looks like the current multicellular phase; as soon as we proceed, we consider the cells as groups or clumps of cells. As the scale increments, we start treating the cells as discs or points, or as an Alpha-Shape connected together through a graph that can be computed with a Delaunay Triangulation. The resulting mass of cells can also be efficiently simulated as simple soft bodies (like suggested here) without necessarily implementing destructible mechanics that would cause potential bugs and programming headaches.
Other organisms
Other organisms, such as single-celled ones, would be scaled down and eventually reduced to simple primitives or even non-simulated points floating through the environment that may be used as food. These could be treated as point masses as we scale further.
Dirt and particles
Like the other organisms, these would be scaled down as well, and eventually become point-like structures.
The physics of point-like structures could also push the player to develop digestive structures to capture and digest this debris.
Editor
The editor starts with the hex editor we already have, and then turns more “continuous”, where the player can “draw” the tissues they have developed in the early phases. Eventually, the properties of these tissues may be modified with a system of sliders or pickers.
Transition to macroscopic
At this point, the transition to macroscopic would be much smoother, but it is still unclear how would it happen technically. I was thinking of calculating a 3D Alpha-Shape of the graph used in the multicellular simulation, and then transition to a mesh-based simulation and editor for the macroscopic.
Conclusion
I really like this concept, and I think it would be nice to discuss on this both over the technical and gameplay related aspects. Thank you.