Detailed Plan

Procedural generation is the art of making content without manually authoring it.

You define a process, a recipe, and it generates a mesh, a scene or any type of 3D content. It's non-destructive, and has the potential to create really complex results or infinite variations with little effort.

What's the use case?

Procedural generation can be used to:

• Automate asset creation
• Get different variations at little to no cost
• Provide a different experience to the player
• And more

Take a game like Minecraft for example, the world is not handcrafted, it's procedurally generated, so you get a new one when you start a new game.

Others, like Spider-Man (PS4), make heavy use of procedural generation too, even if the final product is the same for every one. The game takes place in New York, and the "small" team did not handcraft every single building, they procedurally generated them, so they could fill an entire city and iterate on the design faster.

So what's ProtonGraph?

ProtonGraph is:

• Free and Open-Source Software (FOSS)
• MIT licenced
• A graph editor that gives you full control over the procedural generation
• Can run in headless mode, so you can package it with your game (ongoing work)
• A software that already exists, but needs improvements

On the technical side, it's built with the Godot engine, and the code is structured in a way that makes it easy for new contributors to add new functionalities. The feature set is limited, though. Unlike some alternatives, ProtonGraph focuses on static geometry and composition. More will come in the future, but the priority is to get a solid foundation first.

What about the alternatives?

- Houdini established itself as the industry standard, but it's proprietary, and firmly forbids its use at runtime. This means, you can only generate your content in advance, not when the game is running.

- Blender recently introduced Geometry Nodes, but it's GPL licensed. If you package it with your game, you'll also have to make it GPL, which is not an option for most.

The roadmap and how we plan to use the funds

Port the software to Godot 4 (Ongoing work) and improve the existing set of nodes, which involves:

• Reviewing the existing features
• Study the alternatives solutions
• Provide enough building bricks to cover most common use cases
• If there's still time, improve performance

The requested funds cover 3 to 4 months of work for two people working part-time on this project.

I'm already working with another developer, Iulian, who specializes in UX and procedural geometry. He's very interested in the project, and his main goal is to iron out the final user experience and make it more accessible to game developers.

As for me, I've been working as a full stack developer for 7 years now, doing open source and game dev in my spare time. My goal is to make ProtonGraph production ready as soon as possible.

The requested funds should cover for around 600h of work, split between us.

Metric for success

At the end of this 3 months period, we plan to have:

• A working build of ProtonGraph
• Enough features to generate buildings.
• Landscape tools

Building types should range from simple wooden cabins to skyscrapers and anything in between. Landscaping tools would be limited to heightmaps type of terrain. Voxel terrain is not part of the scope but is a planned addition for later.

The idea is that if the software provides enough building blocks for these two use cases, it will probably cover most other kind of content as well.

Who is using it?

Educhainment is using ProtonGraph for parts of its level editor. The goal is to expose primitive shapes to the level designer, and ProtonGraph uses it as an input to generate the actual geometry.
More details can be found here:
<https://cardano.ideascale.com/a/dtd/Sandbox-MMO-for-learning-languages/381663-48088>