The vision
by cfoust (04/02/2023)
Cube 2: Sauerbraten is probably one of the most popular open-source FPS games of all time. For a while at the end of the 00's it had a thriving community with a clans, competitions, and players who churned out thousands of maps in the game's intuitive map editor. While the player base has waxed and waned over the years, the game has never seen popularity like it once had. Consumer tastes have probably moved on from arena shooters.
But that does not mean the game's time is up. The Cube 2 engine is simple, powerful, and fast and its content creation tools still hold up 15 years later. The popularity of games like Roblox, where players create experiences for other players with a set of simple tools, demonstrates that there is demand for the sort of thing that Sauerbraten once popularized. It might be true that the core gameplay of Sauerbraten is no longer as compelling as it once was, but it could be the foundation for something even better.
I started Sour with that in mind, but it's time to start thinking about what's next. The original goal of the project was to make Sauerbraten available to more people while providing advanced functionality to the existing community, and I have achieved that: as of writing, the main Sour instance gives players access to the complete game experience, including all of its maps (1,800 of them), assets, and game modes. Players can create temporary private game servers to play on with their friends and connect to any of the servers operated by the community.
Yet it is clear that this is not enough on its own to grow the player base. The game needs more, and I think that we are on the cusp of being able to do it. From my perspective, there are two main areas in which the Sour project could make serious gains: building a user-created world and doubling down on the competitive scene.
The goal is for players to be able to create assets, maps, and game modes that other players can access immediately on any device. The joy of games like Minecraft comes from being able to build things together and for them to be there when you return. The game world is the manifestation of your collective fantasy, whatever that may be, which has always been a compelling experience. My hope with Sour is to build something like that in Cube 2: users should be able to craft worlds that have whatever geometry and rules that they want.
That sounds far-fetched, but it's already partially implemented. Every Sour user is assigned a space, which is a game server with an associated map and rules that they determine. When the player edits the map, the Sour instance replays those on a copy of the map and saves the result. Put differently, when the user disconnects and returns, any edits they made are still there. If other players visit their space, they will see it exactly as the owner left it.
Sour therefore has the foundation for persistent game worlds, but it needs expansion. Players should be able to upload arbitrary assets such as sounds, models, and textures. They should be able to define custom game rules both on the server and client side. There should be a way for other players to discover the worlds users have created and even "fork" them. Players could even generate game assets using generative ML models such as Stable Diffusion. If it sounds crazy, it's not: this is all within reach.
It's clear that many players still derive a lot of joy from Sauerbraten's competitive scene, and there are some simple things that the Sour project could do to support them better. For example, a consistent request is to be able to view demos in the web, something that Sour actually has support for, but still has some bugs that need to be worked out before being made available to users. Sour actually already supports competitive matchmaking with ELO rankings (akin to any modern esport) but that functionality is not widely used, probably because of awareness. That's something we could improve.