Keynote 2: Robust Python and Why It Matters
summary
Claims that code often outlives authors and must survive decades of change then drives three concrete tactics: intention, extension, protection. Examples show the measurable payoff — choosing Counter or dataclasses reduces inspection time, replacing physical deps with a pub/sub layer (pi PubSub style) makes adding features cheap, and property testing with Hypothesis plus mypy pylint Bandit finds subtle bugs in milliseconds. A demo pizza‑maker system shows the approach actually prevents night‑long rollbacks.
AI-written summary · about
chapters (13)
- 0:00Robust code for future teams
- 2:30Defining robust Python
- 8:47Intention in robust code
- 15:28Expressive Python abstractions
- 18:23Reducing cognitive burden with types
- 23:59Data classes and invariants
- 33:10Extensible code for changes
- 37:53Protecting developers from mistakes
- 47:53Static analysis and testing
- 53:34Acceptance and property testing
- 1:02:22Software that lasts
- 1:06:16Robustness in early stage
- 1:08:44Performance vs readability
