1.2.3 Software Development Methedologies

• Waterfall Model

  A traditional, linear approach to software development. Consists of sequential phases, where each phase must be completed before moving onto the next. Easy to understand & manage, limited flexibility.
• Waterfall processes:

  1. Requirement Analysis
  2. System Design
  3. Implementation
  4. Testing
  5. Integration
  6. Deployment
  7. Maintenance
• The waterfall model offers:

  • A clear structure and documentation
  • Defined deliverables
  • Suitable with well-defined requirements, with its disciplined approach.

  However, negatives are:

  • lack of flexibility
  • limited customer involvement
  • sequential
  • inefficient resource usage
• The waterfall model is more rigid and sequential, less emphasis on customer collaboration and iterative development, limited need for flexibility. Can be justified for project where disciplined approach needed, making changes later could be expensive. This may include: government contracts or safety-critical systems. If a change needs to be made within a project being developed using the waterfall model, programmers must revisit all levels between the current stage and the stage at which a change needs to be made.
• Spiral Model

  Risk driven, combining the waterfall and iterative development. 4 quadrants, cyclically progressing through each. Continuous evaluation, risk management and refinement.

  1. Determine objectives
  2. Identify and resolve risks
  3. Develop & test
  4. Plan the next iteration
• The spiral is good for:

  • risk management
  • flexibility
  • incremental development
  • customer involvement

  But is bad for:

  • it's complexity
  • time-consuming nature
  • Overemphasis on risk
  • Less suitable for small projects: added complexity may not be justified.
• Spiral model


  • Shares some aspects with the waterfall model, such as sequential plans
  • Adds some iterative development
  • Adds risk management for added flexibility
• Rapid application development


  • Focusses on iterative development
  • Prioritises quick prototyping over risk management
• More suitable for


  • Large, complex projects with significant risks
  • Projects with refinement and adaption
  • Projects where proactive risk management is crucial
  • Projects where benefits outweigh complexity
• Suitable projects


  • Research and development projects
  • Mission-critical systems
• RAD

  Emphasises rapid prototyping, iterative development and close collaboration with end-users. Quickly deliver working software while accommodating user feedback throughout the process.

  
  

  1. Requirement gathering
  2. Prototyping
  3. Iterative development
  4. Testing and integration
  5. Deployment
• RAD is good for:

  • Speed - faster development and delivery
  • Flexible - accommodates changes and user feedback during the process.
  • Customer involvement - encourages close collaboration
  • Reduced risk via early identification.
  • Not good for:
  • Less focus on planning
  • Requires skilled team members
  • Not suitable for large-scale projects due to pace
  • Potential for scope creep - never-ending feedback
• Agile methodology


  • Focuses on quick prototyping and iterative development
  • Flexible, adaptive
  • Incorporates iterative development but places greater emphasis on risk analysis and management
• Agile methodologies


  Share principles of iterative development and customer collaboration (different practices – extreme programming)
• Agile methodology

  • Most suitable for small to medium sized projects
  • With a rapid and flexible need with high use of feedback
• Agile methodology

  Justified for projects where speed and adaptability are critical, benefits of user collaboration
• Suitable for

  • Software prototypes
  • Internal business applications
  • Fast changing market environment
• Compared to waterfall, agile methodology focuses on quick prototyping and iterative development
• Extreme programming (XP)

  Agile, emphasis on communication, collaboration, feedback and simplicity. Deliver high quality software, meeting changing customer needs, promoting efficiency and adaptability + continuous improvement. Frequent releases in short development cycles.

  • Planning
  • Small Releases
  • Continuous integration
  • Test driven development
  • Refactoring
  • Pair programming
  • Collective ownership
  • Customer collaboration
• XP is good for:

  • High quality software via testing & improvement
  • Adaptability
  • Efficient development
  • Team morale

  Drawbacks:

  • Requires strong communication
  • Not suitable for remote teams
  • High initial investment
  • Limited documentation
• Like other agile methodologies

  XP focusses on iterative development, customer collaboration and adaptability
• RAD

  Focuses on rapid prototyping and iterative feedback but not specific practices
• Waterfall and spiral models

  More structured and sequential compared to XP, with less emphasis on customer collaboration and iterative development
• XP

  • Most suited to small to medium-sized projects with changing requirements
  • Strong emphasis on collaboration and communication
  • Justified for projects where high-quality software, adaptability and supportive environment are critical success factors
  • Could include customer-facing software, or projects with higher degree of uncertainty
• Agile

  A set of principles and practices that emphasise flexibility, adaptability and close collaboration between the development and the customer or end-users. They prioritise rapid, iterative development and the delivery of working software in small, incremental cycles. Focus on continuous improvement and prototyping, adapting to changing requirements and fostering a high degree of customer involvement throughout the development process.
• Agile development methodologies

  • Less focus on documentation
  • More priority given to user satisfaction
  • Different sections of software can be at different stages of development
• Non-agile (traditional):

  Waterfall: linear, sequential approach to development, each phase needing to be completed to move on to the next. More rigid, less customer focussed.

   

  Agile (or agile-inspired):

  Spiral: combines elements of the waterfall model with iterative development and strong focus on risk management. Does incorporate flexibility and iterative development.
• Black box testing - testing with no knowledge of the code inside and system architecture, testing to see the output given an input.

  White box testing - knowledge of the code and system architecture, trying to test faults and bugs to produce errors.
• Alpha testing: carried in house, bugs pinpointed and fixed.
• Beta testing: Carried out by end users after alpha, feedback from users used to form next stage of development
• TELOS - Feasibility Study

  • Technical
  • Economic
  • Legal
  • Operational
  • Scheduling
• Regardless of the problem, all good algorithms have certain key qualities which are highlighted below:

  • Inputs must be clearly defined - what is valid and what is invalid?
  • Must always produce a valid output for any defined input
  • Must be able to deal with invalid inputs
  • Must always reach a stopping condition
  • Must be well-documented for reference
  • Must be well-commented so modifications can easily be made