Unit 20 · Language valence: Python under policy · q27b

**Gut Reaction:** This is a recipe for a maintenance nightmare and a performance bottleneck. Python is fantastic for the server modules, data stores, and plugin logic, but forcing it to drive a real-time 3D viewport and complex async UI on mobile/desktop is fighting the language’s fundamental design. You’re asking for C++/Rust-level concurrency and rendering performance from a GIL-bound, interpreted language. It *can* be done, but it will require heavy reliance on C-extensions (like PyOpenGL, PySide/PyQt, or Kivy) and careful architectural isolation to prevent the UI from freezing during async operations. **Top 3 Technical Risks:** 1. **The GIL & Real-Time UI/3D Performance:** Python’s Global Interpreter Lock (GIL) will serialize your async tasks, making true parallelism impossible in pure Python. On mobile, this means janky frame rates in the 3D viewport and unresponsive UI during heavy async calls. You’ll need to offload rendering and heavy computation to C-extensions or separate (greedy generation)

**What Breaks First:** The **3D Viewport on Android**. Python cannot hit 120 FPS (8.33ms per frame) on mid-range Android hardware in a single-threaded, GIL-bound process. The overhead of the Python interpreter, garbage collection pauses, and the cost of crossing the Python-to-C boundary for every OpenGL/Vulkan call will tank your frame rate. The UI thread will also freeze during GC spikes, causing dropped frames and input lag. **What You Do About It (The "Python-Only" Compromise):** You don’t write the renderer in Python. You write a **C-extension bridge** that Python *imports*. This is still "100% Python" from the developer’s perspective (the app is a `.py` file importing a `.so`/`.pyd`), but the hot path is native. ### 1. The 3D Viewport: Offload to a Native C-Extension * **Problem:** Pure Python OpenGL/Vulkan bindings (like PyOpenGL) are (greedy generation)