Discipline of Software Engineering needs overriding theory.
Suggests
- Higher level hierarchical layers of modules or objects
- Incremental development ensuring that the system always works.
Harel (1992, “Toward a Brighter Future for System Development”) offers some optimism
Brooks argues that he was even “Too Optimistic
Harel’s Analysis (evidence of 3 “glooms” ) in NSB
- Sharp separation of essence and accident
- Treatment of each candidate silver bullet in isolation
- Predicting for only 10 years
Brooks argues that hardware performance/price ratio has improved 100 fold over 10 years, but not software.