A structured learning path for Game Development and Extended Reality (XR). Designed to take you from fundamentals to production-ready work.
Named in honor of Dr. Randy Pausch, professor of Computer Science and Human-Computer Interaction, creator of Building Virtual Worlds and author of The Last Lecture.
This roadmap is open to everyone. It is maintained by Next Tech Lab and serves as the primary curriculum for Pausch Lab associates.
Last verified: September 2026 | Engine versions: Unity 6.3 LTS, Unreal Engine 5.6, Godot 4.5, Horizon OS / Meta Quest 3/3S, visionOS 26
- Complete Prerequisites below — do not skip.
- Choose a track: Game Development or Extended Reality (XR). You can do both, but start with one.
- Build in public. Each track ends with milestones — ship them to GitHub.
This roadmap teaches skills, not shortcuts. AI can scaffold code, but it cannot fix motion sickness from a 35 ms pipeline, broken presence from bad interaction, or a 300 draw-call scene on a mobile SoC.
Fundamentals first → build once from scratch → then use production SDKs → profile on device → ship with a write-up. See the full rationale in XR.md — How to Use This Roadmap.
Complete these before specializing. They are non-negotiable for either track.
Vectors (dot/cross product, normalization, projection), coordinate spaces (local/world/view/clip/screen), matrices (translation/rotation/scale/projection), quaternions and Euler angles (gimbal lock, SLERP).
Reference: Freya Holmér — Math for Game Developers
Pick one from each row:
| Purpose | Options |
|---|---|
| Application scripting | C# (Unity, Godot) or TypeScript (Web, WebXR) |
| Systems / engine | C++ (Unreal, Raylib) or Rust (Bevy, WASM) |
- C#: Microsoft Learn
- TypeScript: MDN — JavaScript Guide
- C++: learncpp.com
- Rust: The Rust Book
- Godot scripting: GDScript
Data structures, time/space complexity, composition over inheritance, Entity Component System (ECS), finite and hierarchical state machines (FSM/HSM), memory allocation and cache locality.
Rendering pipeline (vertex processing → rasterization → fragment shading → framebuffer), shaders (GLSL/HLSL/WGSL), materials and lighting.
Reference: LearnOpenGL
Git fundamentals and Git LFS for binary assets.
Engine-focused production tracks.
| Track | Stack |
|---|---|
| Unity | Unity 6.3 LTS, C#, URP / Render Graph, Input System, Cinemachine, Netcode for GameObjects, DOTS/ECS |
| Unreal Engine | UE 5.6, C++ & Blueprints, Nanite, Lumen & MegaLights, Chaos Physics, Niagara, GAS, PCG |
| Godot | Godot 4.5, GDScript 2.0 / C# / GDExtension, Forward+ / Mobile / Compatibility renderers |
| Web | WebGL 2, WebGPU (Baseline 2026), Three.js / Babylon.js, Phaser / PixiJS, WASM (Rust/Bevy) |
| Low-Level / FOSS | Raylib (C), Bevy (Rust), Blender 4.x, LDtk/Tiled, Audacity |
Standards, hardware and spatial computing.
| Track | Stack |
|---|---|
| Core & OpenXR | OpenXR 1.1, 6DoF tracking, motion-to-photon latency, ergonomics |
| Virtual Reality | Unity XRI 3.x + XR Hands, Meta XR SDK (Interaction/Haptics), Horizon OS on Quest 3/3S |
| Augmented Reality | Unity AR Foundation 6 (ARKit/ARCore), Vuforia (Model/Area Targets), Lightship VPS |
| Mixed Reality | Passthrough, Scene Understanding, Spatial Anchors, Meta MRUK, Apple visionOS 26 (RealityKit, SwiftUI) |
| WebXR | WebXR Device API, Three.js / Babylon.js WebXR, Wonderland Engine |
README.md — This overview and prerequisites
GameDev.md — Game Development tracks
XR.md — Extended Reality tracks
CONTRIBUTING.md — Contribution guidelines
1. Track large binaries with Git LFS before first push:
git lfs install
git lfs track "*.psd" "*.fbx" "*.obj" "*.blend" "*.wav" "*.mp4" "*.tga"
git add .gitattributes2. Never commit generated artifacts:
Library/, Temp/, obj/, Binaries/, DerivedDataCache/, Intermediate/, Saved/, .godot/, build/
Use the canonical gitignore for your engine:
3. Branching:
main— stable, reviewed, playabledev— integrationfeature/<name>— isolated work
This is a curated lab curriculum. See CONTRIBUTING.md before proposing changes — please open an issue first for non-trivial edits. Small, version-verified PRs with a clear learning rationale are merged fastest.
Maintained by Pausch Lab @ Next Tech Lab.