11-QA-1 - Copy

Software Quality Assurance Overview

  • Course: Software Quality Assurance Verification and Validation (Part I)

  • Instructor: Lili Wei

  • Semester: Fall 2024

What is Software Quality?

  • Definition (IEEE):

    • The degree to which a system, component, or process meets specified requirements, customer needs, or user expectations.

  • Improvement:

    • Achieved through Software Quality Assurance (SQA).

Key Activities in Software Quality Assurance

Software Verification

  • Question Addressed: "Are we building the product right?"

  • Focus:

    • Ensuring the product meets specifications.

Software Validation

  • Question Addressed: "Are we building the right product?"

  • Focus:

    • Verifying that specifications meet user requirements.

Verification and Validation (V&V) Techniques

  • There are two varieties of V&V techniques:

Static V&V

  • Description: Analysis of code without execution.

Dynamic V&V

  • Description: Analysis conducted through code execution.

Static V&V

  • Objective:

    • Conduct V&V without running the software.

  • Methods:

    • Software inspections & reviews.

    • Static analysis.

Software Inspections & Reviews

  • Purpose: Discover problems using knowledge of the system and technologies.

  • Advantages over Dynamic V&V:

    • Can inspect incomplete versions.

    • Improves knowledge sharing and problem-solving.

    • Goes beyond simply finding bugs; it also identifies inefficiencies and style issues.

Static Analysis

  • Description:

    • Scans code without execution to provide quick analysis results.

  • Purpose:

    • Enables static bug detection.

Examples of Static Analysis Tools

  • Tools Used:

    • COVERITY by SYNOPSYS

    • clang

    • Lint

    • CodeRush

    • Fortify

    • Silent Code Cleanup

    • Infer

    • Intellij

  • Function: Identify and report issues by matching common patterns in code.

Static Analysis Pattern Example

  • Code Snippet: Illustrative of common static analysis issues.

    List<Integer> list = null;  
int length = random.nextInt();  
if (length > 3) {  
    list = new ArrayList<>();  
    for (int i = 0; i < length; i++) {  
        list.add(i);  
    }  
}  
list.add(-1);

Implementation of Static Analyzers

  • Course: ECSE 561 - Automated Program Analysis and Testing

  • Objective:

    • Foundation in software analysis and testing techniques, fault models, and test coverage models.

What is Software Testing?

  • Definition:

    • A formal process where software is examined by executing the programs to identify defects.

Software Faults, Errors, and Failures

Definitions

  • Software Fault:

    • A static defect, often due to human error.

  • Software Error:

    • An incorrect internal state resulting from a fault.

  • Software Failure:

    • Observable incorrect behavior relative to expected requirements.

Related Questions

  • Software Bug:

    • Can refer to faults or failures depending on context.

  • Software Testing Purpose:

    • Involves finding inputs that cause failures.

  • Debugging:

    • Process of finding a fault based on a failure.

Challenges in Testing

  • Quote by E. Dijkstra:

    • "Program testing can be a very effective way to show the presence of bugs, but it is hopelessly inadequate for showing their absence."

Real-world Testing Challenges

  • Difficulties in Triggering Failures:

    • Dependent on environment and edge cases.

  • Exhaustiveness:

    • It's nearly impossible to exhaustively test software due to large input and state spaces.

    • Example: Testing integer inputs in a given range.

Exercising Software Logic Paths

  • Illustration:

    • Number of paths increases exponentially with conditions in loops.

    • Path Complexity requires extensive testing exercises in complex algorithms.

User Behavior Prediction Challenges

  • Example Scenario:

    • Predicting user interactions can lead to unanticipated software behavior.

  • Testing Analogies:

    • Software testing can be likened to structural testing of other systems, such as a bridge, revealing the complex nature of testing software.

Quality Assurance and Risk Management

  • Objective:

    • Strike a balance between testing costs and the risk of undetected bugs.

Reality of Bugs in Software Development

  • Actual Practice:

    • Not all bugs are resolved due to constraints like time and risk of introducing new issues.

  • Perceptions of Bugs:

    • Circumstances and pressure may influence how bugs are categorized.

Upcoming Topics

  • Next Focus: Different test types in software quality assurance.