At GodotCon Amsterdam 2026, “Game Programming Patterns for Godot” started with a tiny Reddit find, a cached getter and grew into a tour of underexplored patterns for real Godot projects. Then the video went online, and the comment section did what comment sections do: it argued, it questioned, and it asked for more.
This sequel answers back. We’ll address the most common (and most spicy) objections from the community head-on, with code, and then go further: new patterns that didn’t fit in the first talk, each demonstrated in a real, shipping Godot project.
The talk has three acts:
Act I: A New Hope, Recapped. A fast, visual recap of the first talk’s core idea.
Act II: The Comments Strike Back. The top community objections and questions from the YouTube release, each answered with working code
Act III: Return of the Patterns. Three patterns that didn’t make the first talk, each shown in the context of my own released game, Tiny Drift:
- State machines beyond the basics: pushdown automata for input and UI flows that plain enum-based FSMs handle poorly.
- MVVM for game UI: separating game state from UI scenes using observable custom Resources; why this is underexplored in Godot and what it buys you in practice.
- The Repository pattern with custom Resources: abstracting save data and content databases behind a data-source interface, and where the classic IRepository approach breaks down in GDScript.
At GodotCon Amsterdam 2026, “Game Programming Patterns for Godot” started with a tiny Reddit find, a cached getter, and grew into a tour of underexplored patterns for real Godot projects. Then the video went online, and the comment section did what comment sections do: it argued, it questioned, and it asked for more.
This sequel answers back. We’ll address the most common (and most spicy) objections from the community head-on, with code, and then go further: new patterns that didn’t fit in the first talk, each demonstrated in a real, shipped Godot project.