Blurring the lines between screenshot tests and real devices

Paparazzi has established automated screenshot testing as a standard industry practice, catching countless regressions early. However, at Cash App, we've recognized that traditional screenshot tests still fall short when validating high-fidelity, touch based interactive experiences or user interfaces that depend on Android's system UI, such as navigation bars or soft keyboard insets. These limitations often force us back into manual testing, allowing subtle yet impactful bugs to slip through.

To bridge this gap, we've been exploring "interactive" animated paparazzi tests, which automate real-time gestures like taps, swipes, and even custom drawings to simulate how real customers interact with our UIs. Additionally, we've built a realistic fake version of Android's system UI that renders the navigation bar, status bar and soft keyboard within our paparazzi tests. This system dispatches accurate insets, allowing us to verify that our UIs react appropriately.

In this talk, I'll demonstrate how these improvements are significantly reducing our manual testing overhead, catching complex UI regressions earlier, and bringing automated paparazzi testing environments closer to the fidelity of real devices.