1. Why software development difficult MIS for bcny: system DB/func conform: follow, no optimality, change, lack of visualization: moving targets, between se. communications, specification: ph1 report Y2k hype: cobol year: 99-00 2 levels: Architecture: modules/classes and detail design: alg+data struct modules + comm; ds + logic/alg: each module alg, math (closed-form, x+N/x), simulation Coding/implementation: coding+debugging understandable, clean logic, swimming and integration: deliver prod Testing: by different people: alpha version(internal), beta(outsiders) Maintenance: customer SQA: system quality assurance team toyota Sys an  >plan  >design…. pros and cons: quality, documentation, strict discipline; Cons: time consuming; lack of comm; big overall risk

2. Rapid prototyping: quickly build summarizations reflect main func in order to communicate with clients/peers embedded in wf model pros: better comm; better response, inputs from clients cons: invite more moving targets, Xwaste time/efforts: No no reuse of prototypes Wf+rp Evolution: simple  complex: system having ind. components partition system into many builds: Ccny mis: std + fac + lib + funding 2yrs std(water fall+ proto): 8 months stop reduce risk Pros: reduce overall risk, Cons: fatal blow—hard to integrate, have to redo design/coding Spiral model: military huge system water fall + risk analysis at the end of each phase Pros: safety ; cons: very expensive Pros and cons of these models

3. Software requirements feasibility (NP, $$) study, trouble of natural languages: ambiguity View-point oriented analysis: all angles data src, repr, reciever of services use-case diagrams: UML unified modeling language (CASE)

4. Finite state machine: definition: place, transition, alphabet examples: binary number property pattern recognition variable parsor; pattern id State chart/diagram: variances from FSM: shape, initial/based place, recurrent nature