Answers from testers who’ve been there.
Open threads and solved fixes from software testers, SDETs and automation engineers. Search by problem, filter by tool, or jump straight to what's solved.
60 threads
Is Playwright better than Selenium in 2026?
I'm planning to learn automation testing and I'm confused between Playwright and Selenium. Which one is more in demand in 2026? What are the advantages of Playwright over Selenium,…
Best approach for API testing in Playwright vs Postman?
I'm currently learning Playwright and have been using Postman for API testing. I noticed Playwright also supports API testing. In real-world projects, do teams completely replace P…
Playwright tests pass locally but fail in CI — where do I start?
Every single Playwright spec is green on my laptop. Push to GitHub Actions and 3–4 of them fail on the first run, then usually pass on retry. The failures aren't consistent — somet…
What are the actual root causes of Appium real-device flake?
Running Appium + XCUITest on a BrowserStack real device pool. Flake rate ~12% on iOS, ~6% on Android. Screenshots on failure look identical to passing runs. I've tried longer waits…
How do you actually debug intermittent Playwright flakes?
We have one Playwright spec that fails ~1 in 15 runs. The rest of the suite is rock solid. I've read the trace viewer docs but every time I open the trace of a flaky run, it looks…
India: moving from services QA to a product / remote QA role — how?
5 years in a services company in Pune, mostly manual + some Selenium. Comp is fine but the work is Jira-ticket-shaped and I want to grow. Trying to move into a product company or a…
Selenium vs Playwright for a complete beginner in 2026?
I'm switching from manual QA into automation. I've been doing manual testing for 3 years and I know Java a bit, some Python. Every tutorial I open is either "Selenium is the indust…
Who owns accessibility testing — devs, QA, or a specialist?
We just failed a WCAG 2.2 audit. Leadership response: "QA should have caught this". Devs response: "we didn't have a11y requirements in the ticket". Design says "we followed the de…
Should we switch our team from Selenium to Playwright?
~1400 Selenium (Java + TestNG) UI tests, 4 QA engineers, 3 years of investment. Suite takes 90 minutes on Selenium Grid and flakes at ~4% per run. Leadership just asked "should we…
How do you actually get engineering to adopt Pact contract testing?
Sold Pact to leadership. Set up the broker. Wrote example consumer + provider tests. Two teams tried it, three teams politely ignored it. The teams that ignored it are exactly the…
Roadmap from manual QA to Playwright automation in 90 days?
4 years of manual QA (mostly exploratory, some API testing in Postman). Zero coding experience beyond copy-pasting snippets. My company will let me shift to automation if I can dem…
How do I actually fix flaky Selenium tests without rewriting everything?
Inherited a Selenium + Java + TestNG suite. ~600 tests, roughly 8% flake rate. Rewriting in Playwright isn't on the table this quarter. Every "fix flaky tests" article says "use ex…
Are test retries an anti-pattern or a pragmatic fix?
Half my team says "retries hide real bugs, never enable them". The other half says "if we don't retry, one flaky test blocks every PR". We're on Playwright with `retries: 2` in CI.…
Which QA metrics actually help vs which ones become theatre?
New VP asked for a "QA metrics dashboard". I've been in QA long enough to know that whatever I put on that dashboard is what my team will optimize for — for better or worse. Test c…
Postman vs REST Assured for API test automation?
Team of 6 QAs, half Java-heavy, half JS-heavy. Currently split: some tests in Postman/Newman, some in REST Assured. Reporting is a mess. Consolidating to one tool this quarter. Whi…
Is the testing pyramid still valid in 2026?
Everyone I follow on LinkedIn is either "the pyramid is dead, use the honeycomb" or "the pyramid was always right, you were just doing it wrong". For a modern SaaS with microservic…
Cypress → Playwright migration — what do you regret?
About to start migrating ~700 Cypress specs to Playwright. Read all the marketing "why Playwright > Cypress" posts. Now I want the other side. For teams that've completed this migr…
Playwright file upload works locally, times out in CI
Test uploads a 2MB PDF via `setInputFiles`. Locally: 100% pass. GitHub Actions: fails ~30% with "Timed out waiting for upload response". The file is checked into the repo under `fi…
What's a locator strategy that actually stays stable across UI refactors?
Devs refactor our React app roughly monthly. Every refactor breaks 20–30 tests because someone renamed a class or restructured a component. We've tried: - CSS selectors → break on…
How much security testing should sit with SDETs vs a security team?
We have a small AppSec team (2 people) and 40+ engineers. AppSec can't keep up with review. SDETs are being asked to "shift left on security". Where's the reasonable line? Am I sup…
How do you isolate test data when running tests in parallel?
Cranked Playwright workers from 4 to 16. Suite time dropped 60%, but flake rate went from 0.5% to 6%. Every flake is "user already exists" or "record not found". Shared DB, shared…
What are companies actually asking in SDET interviews right now?
Interviewing for SDET roles at 4 companies (2 fintech, 1 SaaS, 1 e-commerce). Prep resources online feel dated — half of them are still asking about Selenium XPath tricks. For folk…
When should we NOT automate a test in our regression suite?
Leadership pushed a "100% automated regression" mandate. I've been in QA long enough to know that's usually a bad idea, but I need a concrete list of "don't automate this" cases to…
k6 vs JMeter for API performance testing in 2026?
Standing up performance testing for a REST + gRPC backend. Small team (2 engineers), TypeScript-friendly, need Grafana integration and results in CI. Everyone on Reddit says k6. Ol…
Appium 2: desiredCapabilities vs W3C options — what changed?
Old scripts use desiredCapabilities and Appium 2 keeps warning. What is the correct 2026 shape?
Can AI actually replace QA automation engineers?
Every vendor is pitching "AI-powered testing". My CTO forwarded me one of these pitches with the note "why do we need a QA team then?". I want to give an honest answer — where AI a…
How do you handle iframes in Cypress?
Payment iframe from Stripe. cy.get inside the iframe returns undefined. What is the current pattern?
k6: scenarios vs stages — what is the difference?
Stages felt simple. Scenarios look powerful. When do I actually need scenarios?
QA culture red flags to look for BEFORE accepting a job offer?
Two offers on the table. On paper both are similar (comp, title, stack). I want to avoid walking into a "QA is a testing department that runs regression 2 weeks before release" sit…
GitHub Actions: secrets vs vars — when do I use which?
Both are org-level. When should something be a var vs a secret?
Postman: pre-request script vs Tests tab — what runs when?
I keep putting logic in the wrong tab and chaining requests behaves weirdly.
Why must JMeter run in non-GUI mode for real load tests?
GUI works fine on my laptop. Why do all guides insist on -n -t for CI?
When should QA use Testcontainers instead of docker-compose?
Both spin containers. What is the actual decision?
How does Playwright auto-waiting actually work?
Coming from Selenium where I sprinkled WebDriverWait everywhere. Playwright claims to auto-wait — what exactly does it wait for before each action?
k6 thresholds — how do I fail a build on p95 latency?
I want the CI job to fail if p95 > 500ms or error rate > 1%. What does the threshold syntax look like?
Playwright storageState expires mid-suite on long runs
Auth token has a 30 min TTL. Suite runs ~45 min. Half the specs fail on redirect to login. Refreshing storageState in globalSetup only helps for the first shard. What is your patte…
Can you run Cypress in parallel without Cypress Cloud?
Cloud is pricey. Team asks if we can shard purely on GitHub Actions.
Appium: real device vs emulator/simulator — which for CI?
Emulators are cheap but I keep hearing "test on real devices." What is the right split?
What are the most common Cucumber / BDD anti-patterns to avoid?
Our feature files read like UI scripts. Where did we go wrong?
Postman collection passes locally, fails in Newman on CI
Same collection, same environment file. Locally green. In GitLab CI Newman fails on a chained request that reads an id from the previous response. Feels like variable scoping or ti…
GitHub Actions matrix for Playwright shards - reports do not merge
Each shard produces its own HTML report but I only get 1 in the final artifact. I want a single merged HTML report from all shards. Using playwright@1.45, actions/upload-artifact@v…
How do I run tests in parallel with GitHub Actions matrix?
Suite takes 40 min. Want to shard across 4 runners. What does the workflow look like?
Which Selenium locator strategy is most reliable?
id vs css vs xpath vs data-testid — team keeps arguing. What actually holds up on a modern SPA?
How do I extract a value from a JSON response with REST Assured JsonPath?
Need to chain: POST /login, grab token, use in next request. Cleanest way?
How to safely reset DB state between Cypress e2e specs
Every spec assumes a fresh DB. Right now we hit a /test/reset endpoint in before() which mutates a shared staging DB - flaky as expected. What patterns are people using? Per-worker…
Playwright: when should I run headed instead of headless?
CI is headless by default. Locally I keep flipping. Is there a rule for when headed matters?
Cucumber hooks: what is the execution order of Before, BeforeStep, After, AfterStep?
Multiple hooks in different classes are firing in a weird order.
cy.intercept not matching a specific GraphQL mutation
I have 8 GraphQL calls hitting /graphql. cy.intercept('POST','/graphql') catches all of them but I only want to stub the CreateOrder mutation. Matching by req.body.operationName do…
How do I cache node_modules / npm in GitHub Actions?
CI installs deps from scratch every run. Slow. What is the current best pattern?
Cypress vs Playwright — the honest 2026 take?
Existing Cypress suite, ~800 specs. Worth migrating or double down?
How do I use Appium Inspector to find mobile locators?
New to mobile. Docs assume you already know the workflow.
Appium Android: single button consistently misses touch
One button in a bottom sheet ignores ~30% of clicks. Element is found, isDisplayed=true, but nothing happens. Adb tap on the same coordinates works. Chromedriver? Overlapping view?…
Postman: environment variables vs global variables vs collection variables?
Three overlapping variable scopes and I never remember which wins.
How do I simulate realistic users (not just RPS) in JMeter?
Marketing says "10k users." Ops says "50 RPS." What thread group + timers combo actually models that?
JMeter Throughput Shaping Timer vs standard timers for realistic ramp
We are simulating a launch spike: 0 -> 2000 users over 5 min, hold 30 min, drop to 200. Constant/Uniform timers plus stepping thread group feels hacky. Is Throughput Shaping Timer…
How do I spin up a full test environment with docker-compose?
Need app + Postgres + Selenium Grid for E2E on CI. What does a clean compose look like?
Selenium: implicit wait vs explicit wait — when to use each?
Docs warn against mixing them. Why? And what do you actually run in prod?
k6 vs JMeter — which for a modern perf pipeline?
Migrating off JMeter. Team knows JS. Is k6 mature enough to fully replace?
What is Selenium WebDriver and how is it different from Selenium IDE?
Interviewer asked me this and I fumbled. Looking for the crisp 2026 answer.
REST Assured vs Karate for API testing in 2026?
Java shop, considering Karate for its DSL. Is REST Assured still the safer bet?