1. 4 generations of programming languages
3 paradigms
(in)feasible: early 60s, MIT: computer vision as a course project, security system: replace guards: compare human face vs id photo
iris: violating, fingerprint: dirty false negative
specification report: legal doc: many readers
estimation and planning: $$+people+pl+…
design: 2 levels data struct+ alg design
implementation: coding+debugging (self)
testing: different people avoid subjectivity
alpha vesion: insiders, beta version: outsiders
maintenance and evolution: diff version

2. CASE: smart editor, version control: git, gitlab(linux), make (
ver1, ver1.01, ver1.02…..ver8.01 ver8.02
current version, keep deltas, collaborate
Software engineers!= coders, analyst, design, verification
writers
3 Subjects/parties: clients $; dev: O; end-users/customers
Software development process:
life-cycle models: many choices
tools: PL(matlab, python, R, java, C#, Ruby), DB, OS, IDE/CASE;
personnel. Intelligence: moving target? 88/4Turing’s test
Life cycle: series of steps thru which the development progresses.

3. Water-fall: traditional, >80%, ord eng.
req analysis (>30%): specification; needs/constraints/$$/deadline/reliability/size
investigation: concept exploration,
meetings, feasibility, finance
30 cities, travel once minimal Hamilton circle, NP-hard
mutual misunderstanding: replace(put it back) the handle in gas station; “I have a lamb”:
I have a lamb. 1 food; 2: pet; 3: nice husband/guy
solutions:
0. more NL
1. use logic description, math: universal language
2. diagram/graph/visualization tools;
3. prototype: simulation sw/hw, simulink, python/matlab prototyping pl, sql