After a break, I have gotten back into my project.
I had written a new voxel engine in the summer that allows for streaming in of chunks and pretty much unlimited travel in any direction (I will post something about that later), but I choose to work on the old one that is a fixed area as there were so many other features already baked into it.
Over the last three weeks during the holidays, I’ve worked the following:
- Agent pathfinding that can now build paths to unreachable areas
- Agents gathering resources
- Agents building things based on player designs
- Commanding the agents in 3rd person and through a user interface
I sat back and marveled at my progress I have made as a solo developer. Then my kids played the game and asked, “What’s new?” I told them about the updates, and they said, “You had all that a long time ago.” I tried to explain these were new features built on reusable and extendable core systems, but they responded, “Looks the same to me.” Indeed, when looking back at old videos, I realized I’ve barely added anything new from the player’s standpoint.
All my struggles are stemming from trying to make the perfect underlying system. I have no idea what the core gameplay is going to be, beyond commanding agents to collect, build, defend, and attack—while the player focuses on strategy and third-person combat.
Yet this experience drove home a crucial point: wearing all the hats—code, art, and sound—I can’t afford to over-engineer every system. It’s tempting to plan for every obscure edge case, but I could easily spend years building a perfectly robust framework without ever delivering a single entertaining moment.
So, I’m flipping my priorities. From now on, “Less Tech, More Fun” means creating a minimal version of a feature, testing whether it’s enjoyable, and only then fleshing it out. Does the player need intricate pathfinding for a grand siege scenario if the basic combat loop isn’t even satisfying yet? Not really. Until the “fun” is validated, all those bells and whistles can wait.
This isn’t to say complex systems like agent AI, pathfinding, and resource management aren’t important—they’re vital for an RTS-style game. But they need to serve the experience, not overshadow it. By prototyping quickly and moving on when something’s “good enough for now,” I’m hoping to build a foundation that captures the essence of what makes the game fun—then bolster it with deeper mechanics once the core is solid.
Ultimately, I want to enjoy making this game as much as people enjoy playing it. Over-engineering is a luxury for studios with specialized teams. As a solo developer, if I don’t keep my eye on the fun, I might never finish. So here’s to the next phase: find the fun, deliver the fun, and let the tech support the fun—rather than drive it. Stay tuned for more updates, and hopefully, more fun.