chapter 9

program evolution dynamics

useful systems keep changing, and complexity grows unless teams refactor and re-architect deliberately

software maintainance

delivers corrective, adaptive, perfective and preventive changes.

dominates lifecycle costs

complexity and process

track complexity and process metrics, like change frequency lead time and defects to forecast risk, target refactoring and keep evolution predictable

evolution processes

effevtive evo process triages requests, perfoms impact analysis implements changes, runs runs regression tests etc.

software change

inevitable

new requirements emerge, business environment changes, errors need repair, new equipment

key problem for orgs: implementing and managing chage to existing software systems.

importance of evolution

software are investments, need to maintain value by changing and updating when needed. consumes the most of the budget

lehman’s laws

continuing change: real world programs must change or they will become obselete.

increasing complexity: as programs evolve, structure becomes more complex. more resources needed to preserve and simplify it.

large program evolution: self-regulating. system attributes like size, time btween releases, and number of errors is constant for each system release.

organizational stability: rate of development is constant and independent of resources devoted to development

conservation of familiarity: incremental change in each release should be constant. no big changes to the system in one release.

continuing growth: increasing functionality offered each release for user satisfaction.

these laws are generally applicable to large tailored systems.

software maitainance