Why agile
summary
Argues that half a century of software engineering fails to fix three real problems — internal complexity, human error and project variability — and backs that with a concrete disaster the Polar Lander lost signal and crashed because of a missing conditional costing 165 million dollars. Shows how abstraction and formal methods repeatedly fall short and how industrial process models overpromise. Proposes an empirical alternative: observe form hypotheses run short experiments and iterate — examples include pragmatic ORM and DSL work like the ActiveRecord pattern that get tried and either survive or die.
AI-written summary · about
chapters (14)
- 0:00Intro
- 2:00The Softer Crisis
- 4:05The Software Crisis
- 5:40Source of Problems
- 11:10Formal Methods
- 16:15Project variability
- 19:35Project repeatable
- 22:20Fundamental misunderstanding
- 23:30Industrial engineering
- 25:10Project and build
- 27:05Rational Unified Process
- 29:10Software is not defined
- 31:40A less loaded example
- 35:00Outtakes
related talks
38mCognitive biases and failure of Agile
01h 22mAgile - The Road Towards Organizational Agility - Progress & Challenges Examined
49mAgile Methods are Dangerous
34mHow to screw Agile and digital transformation
47mAgile Is Not the Easy Way Out
03mAgile is a disaster today!
46mPeople, Principles and Pragmatism
44mBehavior and attitude, the new Agile approach