Visualizing and Explaining Rust's Ownership Model
summary
Provides a concrete claim: embedding an auto-generated permissions model visualization (driven from the compiler MIR and a hooked MIR interpreter) into a Rust textbook improves ownership learning by about 9 percentage points (effect size ~0.56). The tool extracts borrow-checker facts (Polonius-compatible), shows flow-sensitive permission changes, and generates counterfactual heap/stack traces for code that fails to compile. Evaluation uses a data-derived ownership inventory from Stack Overflow examples and an A/B deployment with ~150 completers per condition.
AI-written summary · about
chapters (14)
- 0:00Introduction
- 1:24Why ownership is hard
- 3:07Concept inventory for ownership
- 8:43Misconceptions in studies
- 15:25New pedagogy for ownership
- 19:17Dynamic semantics visualization
- 21:44Static semantics with permissions
- 27:29Design decisions behind visualization
- 34:49Refining the visual model
- 38:43Limitations and open problems
- 44:38Diagrams in Rust textbook
- 46:24Evaluating pedagogy vs baseline
- 49:55Results and improvements
- 50:18Live demo and Q&A
related talks
19mA Grounded Conceptual Model for Ownership Types in Rust
16mPlace Capability Graphs: A General-Purpose Model of Rust’s Ownership and Borrowing(…)
28mYou Don't Own Me
40mAgainst Borrowing: Own the forest, not the trees!
40mAdvanced Rust Programming Techniques
36mThe Rust Borrow Checker: A Deep Dive
17mAdventure of a Lifetime: Extract Method Refactoring for Rust
59mRust for C++ developers - What you need to know to get rolling with crates