Weekly digest

Averages are comforting, but they are also excellent at hiding problems. When a test case has a long history of passing, a sudden failure this week barely moves the needle on a twelve-month average. We end up looking at a healthy dashboard while the pipeline is quietly struggling. This week, we changed how test case metrics are scoped to eliminate that blind spot.

The summary metrics on a test case page now respond to a date range picker. You can filter the view to the last 24 hours, 7 days, 30 days, or 12 months, or set a custom range. The real power here is not just in the filtering, but in the shareability. The selected timeframe lives directly in the URL. When you share a link to a test case, the recipient sees the exact window you were investigating without having to guess your intent.

This adjustment to the test case metrics timeframes shifts our relationship with test data. It forces us to look at the immediate present rather than the smoothed-out past. Narrowing the window is exactly how you surface a test that passed for months and started failing on Tuesday.

We often build tools to give people more data, assuming more visibility leads to better decisions. But visibility without context just creates noise. By letting developers isolate the exact timeframe of a regression, we give them the precise context they need to act. The focus now turns to how this isolated view changes the way we triage failures across the Once and Tuist testing environments.

It is a small interface change, but it changes the conversation.