Additive Bias: Why Software Keeps Growing More Complex
Additive bias makes teams add features, code and test cases rather than remove any. Complexity and debt grow, and subtraction is the overlooked fix.
Non-functional Quality
Software can meet every functional requirement and still fail because it collapses under load, shuts people out, or cannot be run securely. This hub collects contributions on the quality characteristics that tend to be missing from requirements and become very visible in production.
Accessibility is a major focus, from testing with people with disabilities to accessible PDFs and the question of how AI can support accessibility. Alongside it are performance and load testing, quality in technical domains such as electric cars, wind turbines, and e-government systems, and quality by design. Retrofitting this kind of quality just before go-live is expensive, so a recurring question here is how to plan for it from the start.
Articles
Good Software: What Users Need Beyond Passing Tests
Good software supports its users, and matching the spec is only part of that. A camper van shows how software lets people down and which questions help.
Software Security Awareness: Security Is Everyone's Business
Software security awareness decides whether users trust digital services with sensitive data. Security requirements and tests alone won't get you there.
Test Dimensions for Integration Testing: Identifying Gaps
Test dimensions structure integration testing: objectives, test objects, level, test basis and typical errors. Mapping them reveals the gaps.
Object-Oriented Technology: Promises Only Partly Kept
Object-oriented technology promised big productivity gains and near-zero maintenance. Studies show partial success, with gains in reuse and portability.
Sensory Impressions in Software Design
Software design gains depth when user wishes are captured with all senses. The Walt Disney strategy, VAKOG and the Dilts pyramid help along the way.
Testing Tools: Why Candidates Drop Off the Shortlist
Testing tools drop off a shortlist for simple reasons: no hands-on evaluation, unstable runs, poor process fit, pushy sales or clunky usability.
Learning from Apps: Testers Should Speak Up for Usability
Testers can learn from apps. Instead of only checking requirements, they speak up for the focus and ease of use that business software so often lacks.
Be Good, Get Better: Customer View and Creative Test Methods
Software testing is an established profession, yet testers get better by seeing the app through the customer's eyes and using test methods creatively.
E-Government Platform: How Saxony Runs Acceptance Tests
Saxony's e-government platform relies on acceptance tests in a shared test center, run by test specialists and business testers, plus release tests.
Application Value Assessment: Metrics for IT Strategy
An application value assessment measures the size, complexity and quality of legacy systems to guide maintenance, migration and IT strategy.
Testing Geoportals: Services and Clients per ISTQB Process
Testing geoportals means system and acceptance tests for services and clients. The ISTQB test process adds structure, from test concept to security.
Podcast Episodes
Dark Side of Testing: The Failures We Cause Ourselves
A test can leak data, expose code or bring down production. The dark side of testing begins with the setup, which needs a risk assessment of its own.
Kubernetes Load Testing: Start with a Risk Question
CPU and RAM are rarely the bottleneck in Kubernetes load testing, caching or an external login more often. A test pays off once a decision follows.
Quality Hunting: Testing for Information Instead of Bugs
In quality hunting, teams of about four people explore a system for half a day and report what stakeholders need to know about its quality.
Agentic Systems in Testing: Why Humans Stay in the Loop
Agentic systems are not just agents working alone. Between 8 and 12 agents can cover testing end to end, but humans still judge what AI cannot.
Performance Monitoring: Set a 70% CPU Alert Before Failure
Performance monitoring catches slowdowns users abandon after five seconds. Alerts, AI log analysis and a small replica find the exact query at fault.
Vibecoding and the Lack of System Understanding in Practice
Vibe coding risks grow where system understanding is missing. The future belongs to Pi-shaped developers who combine breadth with real depth.
Children's Book About Testing That Inspired a Future Tester
A children's book explains software testing through dragons and knights, and a 12-year-old at a book fair said she wants it as her profession.
Automatically Reviewing Documentation with AI Personas
Documentation testing with AI personas: a simulated junior developer burned 1.8 million tokens where a senior needed 300,000, exposing gaps in the docs.
Architecture Fitness Functions: Making Decisions Testable
Architecture fitness functions turn design decisions into testable KPIs, so hard-to-reverse Type 1 decisions defend themselves against violations.
Blue Angel for Software: Transparency Instead of Top Marks
The Blue Angel for software asks for energy data, support for five-year-old hardware and free security updates. Certification takes about three months.
How Stiftung Warentest Keeps Its Product Tests Objective
Stiftung Warentest buys products undercover, runs blind tests with reference devices and double-checks every measurement. Testers can borrow a lot.
Formal Methods: Why Is Functional Programming Helpful?
Formal methods sound like an ivory tower, but they provide what testing alone can never do: mathematical proof of correct behavior.
Design Patterns in Test Automation: Less Spaghetti Code
Test automation design patterns such as Page Object, Builder and Facade keep test code readable. DRY and YAGNI stop the framework from bloating.
Thermal Management Testing: Earlier with HiL and Virtual ECU
Thermal management testing for electric vehicles moves from the road to HiL and a virtual ECU. Faults show up earlier, yet a test run still takes days.
Testing in 2025: Realism Wins, Accessibility Grows
Testing in 2025 got faster with AI for test cases, scripts and test data, and accessibility became a legal duty. Next up: orchestrating AI agents.
Performance Testing vs. Load Testing: Not the Same Thing
Performance testing vs. load testing: capacity tests every sprint waste cloud budget. The real question is how a system behaves and what it costs.
Testing Wind Turbines: Lightning Only Strikes in the Field
Testing wind turbines means covering about 164 subsystems with simulators, hardware in the loop and field data, because lightning cannot be provoked.
Dynamic Analysis: Embedded Coverage on Real Hardware
Dynamic analysis on real hardware: embedded trace measures code coverage in integration testing without the timing side effects of instrumentation.
Software Testing as Art and Science: Logic and Empathy
Software testing needs logic and imagination alike. What da Vinci's curiosity and love of simplicity teach testers about reports people understand.
Software Testing Value: Frame Quality in Business Terms
Software testing value gets lost when testers pitch defects. Revenue and user satisfaction make a stronger case, and domain knowledge matters more.
Accessible PDFs: Let People with Disabilities Test Them
Accessible PDFs need clean headings, alt text and labeled tables. For contracts, invoices and archives, converting old files is a migration project.
Teaching Test Automation: Hands-On Labs Before Theory
Teaching test automation in Vilnius starts with scripts, APIs and performance tests. Test plans come later, once students have seen systems break.
When Property-Based Testing Pays Off, and When It Doesn't
Property-based testing generates test cases from a specification and found a 17-step LevelDB bug. It gets costly when every test run is billed.
Open-Source Software: Four Requirements for Vendors
Open source compliance starts once you ship software. Vendors need license checks and an S-BOM, which the Cyber Resilience Act makes mandatory.
Holistic Testing: Quality Work at Every Stage of the Loop
Holistic testing makes the whole team own quality, with testing at every stage of the DevOps loop. Shared understanding early prevents costly rework.
Nudging for Security: Making the Secure Choice the Easy One
Nudging for security starts where most attacks start: human error. Secure defaults, opt-out 2FA and password meters guide users without forcing them.
AI Audit: How the AI Assessment Matrix Structures Testing
An AI audit needs a yardstick. The TÜV AI Lab's AI Assessment Matrix maps test dimensions against the lifecycle, so assessors pick the relevant fields.
How to Pick a Test Design Technique Without Learning 30
25 to 30 test design techniques exist, yet about ten from four groups cover most cases in practice. How to pick one by problem type, risk and skills.
Accessibility Act: Blind Testers Find What Tools Miss
Accessibility testing can't be fully automated. Accessible components can still break once combined, and blind testers catch it with screen readers.
Software Quality Metrics: Trends Say More Than Single Values
Software quality metrics only help when they fit the context and feed real decisions. Team comparisons mislead, and lead time deserves a closer look.
Cyber Resilience Act: Connected Machines Are Affected Too
The Cyber Resilience Act covers connected machines, with risk assessments and reporting duties. Starting early secures consultants before rates rise.
Secure by Design: Security Starts at the Beginning
Secure by design means building security in from day one, not bolting it on later. What defense in depth, secure defaults and threat modeling look like.
Accessibility Testing: AI Checks Linguistic WCAG Criteria
Accessibility testing runs on a jumble of Word, Excel and browser tabs. AI checks linguistic WCAG criteria, and clustering picks representative pages.
Accessibility Under the BFSG: What B2C Companies Must Do
Since June 2025, the BFSG makes accessibility mandatory for B2C companies, with fines up to 100,000 euros. WCAG-based tasks help teams comply.
Sustainable Software: Performance Cuts the Carbon Footprint
Sustainable software starts with performance: less CPU, memory and data transfer cut carbon and cloud costs. Three image tweaks alone save 97 percent.
Tests from Rules: Generated Cases for Distributed Systems
Tests from rules replace hand-written cases: describe a system property once and a tool generates thousands. A shrinker traces failures to the cause.
Mobile App Testing: Second Attempt with Robot Framework
Mobile app testing, second attempt: HUK Coburg scrapped four years of automation. Now manual testers write keyword-driven tests with Robot Framework.
Sustainability in Testing: Going Green Often Saves Money, Too
Sustainability in testing is already hidden among ISO 25010 quality criteria. Dropping redundant test runs saves computing power, and with it money.
Accessibility in Apps: Why Do We Perform Manual Testing?
Mobile accessibility testing is mostly manual: VoiceOver, TalkBack and switch control can block each other, so the test matrix has to log every setting.
Digital Accessibility: 57 Percent Can Be Tested Automatically
Automated tests catch only about 57 percent of digital accessibility issues. The rest needs manual testing and an early start in the style guide.
Evolutionary Quality Goals: What Quality Matters and When
Evolutionary quality goals show which quality characteristics count in each system phase. Maintainability pays off once growth brings new developers.
Green IT: Fewer Features, Less Data, No Idle Machines
Green IT makes software draw less power, green by IT uses software for sustainability. How the Pink Index, caching and smaller cloud nodes help.
C++ Core Guidelines: Ownership Is the Most Important Rule
The C++ Core Guidelines by Stroustrup and Sutter put ownership first: does code own a resource or only borrow it? Getting it wrong crashes programs.
User Feedback at MOIA: When Colleagues Are the Users
How MOIA collects user feedback from its own drivers: monthly surveys, mockup interviews, test drives and slow rollouts to 20% of the fleet.
Prompt Engineering for Test Cases: Review Remains Mandatory
Prompt engineering for test cases gets better results with few-shot examples and embedded specs. Still, two of eleven generated tests remained wrong.
Sustainable Software Development: How Testing Saves Energy
Sustainable software development starts with the requirements, long before deployment. Performance, maintainability and lean test pipelines all count.
Whole-Team Quality: The Three Factors Good Teams Share
Quality as an attitude makes quality the whole team's job. Good test cases, static analysis as a hard gate and test automation back it up.