CS UNIT 12

Program Development Lifecycle Overview

• The program development lifecycle is a phased approach to creating software, including analysis, design, coding, testing, and maintenance.
• It ensures a well-ordered and documented process for program development, making it understandable and usable by developers.
• It is a continuous process until the program is no longer in use, with alterations made to address new circumstances or errors.

Stages of Program Development Lifecycle

• Analysis: Clearly define the problem and gather requirements.
• Design: Use the program specification from the analysis stage to outline program development.
• Coding: Write the program using a suitable programming language.
• Testing: Run the program with various test data to ensure it meets requirements.
• Maintenance: Ensure the program's continued effectiveness, addressing problems, correcting errors, improving functionality, or meeting new requirements.

Analysis Stage

• The analysis stage focuses on clearly defining the problem and requirements.
• It often starts with a feasibility study, followed by fact-finding to determine program requirements.
• Requirements include detailing what the program should do and what data is needed.

Design Stage

• In the design stage the program specification forms how the program should be developed.
• This includes a clear understanding of all tasks, how they should be performed, and how they interact.
• Design documentation includes structure charts, state-transition diagrams, and pseudocode.

Coding Stage

• The coding stage consists of writing the program or set of programs in a programming language.

Testing Stage

• The testing stage consists of rigorous testing with different sets of data.
• The intention is to verify how the program conforms to the program design.

Maintenance Stage

• The maintenance stage ensures continued effective operation.
• This includes fixing errors that arise, improving features and adapting to new requirements.

Different Development Lifecycles

• Program development methodologies vary in strength, leading to different lifecycle models.
• The models mentioned are: waterfall model, iterative model, and rapid application development (RAD).
• The models can be divided into the five stages: analysis, design, coding, testing, and maintenance.

The Waterfall Model

• This model is a linear sequential approach where each stage completes before starting the next.
• It is suited for smaller projects with well-defined, unchanging requirements and a short timescale.
• Benefits include ease of management and understanding and stages that do not overlap.
• Drawbacks involve difficulty changing requirements later and the unsuitability for programs where change may occur.

The Iterative Model

• This model begins with a simple subset of requirements, expanded and enhanced through repeated cycles.
• It's appropriate for projects with known major requirements but evolving details.
• Principles consist of incremental development and high customer involvement.
• Benefits include early working programs, easy testing, and flexibility in altering requirements.
• Drawbacks include the need for the whole system to be defined at the start.

Rapid Application Development (RAD)

• This model develops different parts of the requirements in parallel, using prototyping for early user involvement in testing.
• RAD is runs program development cycles are run in parallel by different teams.
• Minimal planning and code reuse reduces overall development time.
• Drawbacks include the need for a modular system and strong, skilled teams.

Structure Charts

• Structure charts are modelling tools that break down a problem into sub-tasks.
• They illustrate a hierarchy of modules and their interactions.
• Modules are represented by boxes, and parameters are passed through arrows.
• Structure charts can represent selection using diamond shapes, and repetition by adding labelled semi-circular arrows.

Finite State Machine (FSM) & State-Transition Diagrams

• A finite state machine (FSM) is a mathematical model of a machine in a fixed set of states.
• State changes occur through external inputs, called transitions.
• A state-transition diagram shows the FSM's behaviour and the conditions for transitions.
• States are represented by nodes (circles), transitions by arrows, and events by labels on arrows.

Program Testing

• Testing includes avoiding and exposing faults in programs.
• A test strategy should be set out in the analysis stage of the program development lifecycle showing testing requirements.
• A test plan helps make clear what tests need to be performed.

Types of Test Data

• Normal test data: Should be accepted and shows the program works.
• Abnormal test data: Should be rejected due to being unsuitable.
• Extreme test data: On the limit of data accepted by a program.
• Boundary test data: Data that is just outside the limit of that rejected by a program.

Types of Errors

• Syntax errors: Errors in the grammar of a source program.
• Logic errors: Errors in the logic of a program, meaning the program does not do what it is supposed to do.
• Run-time errors: Errors that happen when the program is being executed.

Testing Types

• White-box testing: Detailed testing, testing the structure and logic of a procedure.
• Black-box testing: Tests a module's inputs and outputs.
• Integration testing: Testing of modules to ensure they work together.
• Stub testing: Using dummy modules for testing purposes.
• Alpha testing: In-house testing by the development team.
• Beta testing: Testing by a small group of users prior to release.
• Acceptance testing: Proving to the customer that the program works as required.

Program Maintenance

• Corrective maintenance: Corrects errors that appear during use.
• Perfective maintenance: Improves program performance.
• Adaptive maintenance: Alters a program to perform new tasks.