The list is fetched from the backend and the time estimate is shown.
The requests take progressively 3, 6, 9, ... seconds to test how the loading animation
handles the time overflow.
or control the load request manually:
basic, custom estimator,
reporting accuracy, low-level methods for custom actions