Notes from shipping software that works.
Testing strategy, tooling and the human side of an always-on release pipeline, written from what we actually ship.
Playwright 1.65 Changes How Page Object Fields Initialize
Playwright 1.65 follows useDefineForClassFields, so page object fields initialized from this.page can throw. It also adds Bun support and failures-only reports.
ReadWhich Test Data Isolation Strategy Fits a Shared Database?
Test data isolation: rollback, per-worker databases or Testcontainers? Three questions: same connection as the app, workers or shards, fidelity vs CI time.
Playwright Test Agents: Run the Healer Without Losing CI Signal
Playwright Test Agents: run the healer on a branch a person reviews, keep the gating CI job deterministic, and sort each heal diff by cause before it merges.
Exploratory Testing Tours: Whittaker's Six Districts
James Whittaker's book groups exploratory testing tours into six districts. This post adds its own selection rule for tours that fit one charter's mission.
Playwright Visual Tests That Pass in CI for the Right Reasons
Fix failing Playwright visual regression checks in CI: a four-step triage with tolerance last, and CI-generated baselines approved by named code owners.
Playwright in GitHub Actions: Docker, Failure Capture, and Worker Parallelism
Match the Playwright Docker image tag to your installed version, pick screenshot and video capture on purpose, and set CI workers instead of the local default.
HTSM: Inside James Bach's Heuristic Test Strategy Model
HTSM is James Bach's four-branch model for test strategy, currently v6.3: test techniques, project elements, product factors, and quality criteria categories.
HICCUPPS-F: Nine Testing Oracles for Defending Your Bug Calls
HICCUPPS-F: nine testing oracles beyond "the spec says so" for recognizing and defending a bug call before it gets waved off as works as designed in triage.
Session-Based Test Management (SBTM): Definition, Sessions and Metrics
Session-based test management (SBTM) explained: charters, 90-minute sessions, session sheets, TBS metrics, PROOF debriefs, and where the method breaks.
Precision in PROOF Debriefs: The Debriefer's 60-Second Test
Run the debriefer's 60-second test on every field of a PROOF debrief before the report ships: five checks that separate action-ready fields from filler.