Software Life Cycle Models

Overview

Software Development Life Cycles are a series of phases

• which provide a common understanding of the software building process

• overseeing the flow of software as a project/product

  • from its inception as an idea through its development and testing
    • til its delivery and maintain its shelf-life

Software Engineer should have enough knowledge and experience to choose which SDLC model to use based on the project context and project requirements

Verification - are we buidling the product right Validation - are we building the right product

Requirements Analysis —> Specifications —> high level design ^Maintenance low level design ^Delivery <— testing <—- implementations unit testing integration testing system testing 2 ways of quality engineering

• quality assurance
  • designing a process for a schematic for product of how it should go
    • how its done, whos doing it, …
      • manual for raw to product
• quality control
  • matches the specs for product at the end
    • try to break the product
      • sample the product from the manuafacture and perform extreme testing
      • check if it operates within the boundaries

Waterfall model

• a linear sequential flow
  • for a phase to start, previous phases have to be finished
    • earliest model; introduced by Dr. Winston Royce in 1970
• suitable for project which do not change requirements
  • when a consumer has a very clear documented requirements

Advantages

• easy to follow
• well-defined and structured
• concise plan and schedule
• verifying each step ensures early detection of defects
• each phase has specific defined deliverables

Disadvantages

• assuming requirements will not change
• expensive to step-back through stages
• planning process is costly or infeasible

V model

• modified waterfall model
  • linearity is preserved but the process steps are bent upwards after the implementation phase
• early testing
• suitable for project which have clearly defined requirements
• SW dev tech stack and tools are well known

Advantages

• simple, easy
• each phase has deliverables
• high chance of success due to early testing
• verification and validation are done in early stages

Disadvantages

• adjustments are difficult and expensive
• no clear path for the problems during testing phase
• planning process is expenseive

Prototyping model

• creates incomplete prototypes of the final product
  • visualizes some components of the final product
    • to narrow the gap between requirements of the user
      • and understanding of the SW engineers
• flexible because requirements are built up as product is developed
• different types:
  • iterative prototyping
  • evolutionary prototyping
  • incremental prototyping
• can be used with any SDLC
• user interaction is crucial for success of this model

Advantages

• reduces time and cost as product validation is carried out early in the product lifecycle
• increases user involvement
  • more clarification for SW engineers and greater satisfaction for the user

Disadvantages

• user might confuse prototype to final product and give negative feedback
• negative feedback can lead to misunderstanding of user objectives
• prototypes implementation can be expenseive and exhaustive
• developers can get drifted to keep enhancing prototypes rather than focusing on the final product

Iterative prototyping model

• prototypes are built to easure the satisfaction of the user
  • and they are discarded after getting user's feedback
• prototypes are not functioning versions of the product/sub-product

Evolutionary prototyping model

• like iterative prototyping model but the prototypes represent evolutions of the final product and are not discarded

Incremental prototyping model

• the final product is an integration of separate prototypes

Spiral model

• combines both design and prototyping in its stages

  • Barry Boehm in 1986

• combines features of the waterfall and prototyping models

• difference between this model and waterfall

  • is that it is planning and prototyping at each stage with a risk assessment

• used for large projects associated with much more risk

• suits projects that can be built in small phases and segments

Advantages

• estimates become more realistic as work progresses because of the early discovery of issues while studying risks and planning for it
• early involvement of developers

Disadvantages

• high cost for planning and risk management
• needs special skills and domain experts for risk analysis
• highly customized, which limits re-usability

Agile model

• based on iterative, incremental development
• requirement and solutions evolve through the collaboration between cross-functional teams
• engages the user with the developing teams
• can be used with any type of project
• can deliver some of the requirement before the final product is finished

Advantages

• development can go side by side with the requirement analysis
• face-to-face communications and continuous input from user limits the gap of misunderstanding
• final product is high quality software in the least time possible and customer satisfaction

Disadvantages

• scalability; big project might be a challenge for this model
• needs spontaneous high communications skills
• documentations might also be a challenge
  • leaving the specfication and conditions ambiguous
    • which leads to delivery delays and conflicts