Common issues and how to fix them.
Problem: Tests fail with browser-related errors.
Fix:
uv run rfbrowser init chromiumThis downloads Chromium browser binaries (~250MB). It's a one-time download.
If it still fails:
# Install system dependencies (Linux)
npx playwright install-deps chromium
# Then retry
uv run rfbrowser init chromiumProblem: Test fails waiting for an element on the page.
Possible causes:
- Wrong selector — Double-check the CSS selector against the page
- Page not loaded — Add a wait or check navigation completed
- SauceDemo is down — Try opening https://www.saucedemo.com in your browser
Debug tip: Run with visible browser:
uv run robot --variable HEADLESS:false tests/your_test.robotProblem: Codespace or DevContainer won't build.
Possible causes:
- Docker not running (local setup) — Start Docker Desktop
- Insufficient disk space — Free up space or increase Docker disk allocation
- Network issues — Check internet connection
Fix for Codespace: Try deleting the Codespace and creating a new one.
Problem: rfbrowser init fails with permission errors.
Fix:
# In DevContainer, try:
uv run rfbrowser init chromium
# If that fails, use sudo for system deps:
sudo npx playwright install-deps chromium
uv run rfbrowser init chromiumPossible causes:
- Headless mode — CI runs headless. Make sure your tests work with
headless=${True} - Timing — CI machines may be slower. Avoid
Sleep; use Browser Library's auto-wait - Dependencies — Make sure
uv.lockis committed and up to date
Fix: Always test with headless mode before pushing:
uv run robot --variable HEADLESS:true tests/Fix:
curl -LsSf https://astral.sh/uv/install.sh | sh
source ~/.bashrc # or restart your terminalFix:
uv sync # Install dependencies firstAfter running tests, check the results/ directory:
report.html— High-level summary: pass/fail counts, statisticslog.html— Detailed execution log: every keyword, every argument, every screenshotoutput.xml— Machine-readable results (used by CI)
Open log.html in your browser for the most useful debugging information. You can:
- Expand test cases to see each keyword's execution
- Click on failed keywords to see error messages
- View screenshots taken during the test
Run this to diagnose all issues at once:
python scripts/check_environment.pyIt checks 11 prerequisites and shows exactly what's wrong and how to fix it.