test_driver.send_keys() dispatches the key events synchronously, while the script that calls it runs: the keydown listener runs in the script entry point of that script (https://w3c.github.io/long-animation-frames/#create-script-entry-point records nothing while a script is pending), so no INPUT.onkeydown script is reported; a browser dispatches them from the task of the user interaction
the subtest forcedLayoutDuration and forcedStyleAndLayoutDuration are strictly positive after heavy forced layout work: the library does not measure the time it spends on the styles and the layout: styleDuration, layoutDuration and the forced*Duration of the scripts are 0 (with the virtual clock of the runner, which only advances when it is read, they would be 0 anyway)
the subtest forcedStyleDuration and forcedStyleAndLayoutDuration are strictly positive after heavy forced style work: the library does not measure the time it spends on the styles and the layout: styleDuration, layoutDuration and the forced*Duration of the scripts are 0 (with the virtual clock of the runner, which only advances when it is read, they would be 0 anyway)
the subtest Heavy render-phase layout produces strictly positive layoutDuration: the library does not measure the time it spends on the styles and the layout: styleDuration, layoutDuration and the forced*Duration of the scripts are 0 (with the virtual clock of the runner, which only advances when it is read, they would be 0 anyway)
a synchronous request of a window blocks the thread that advances the virtual clock: the runner answers delay.py?ms=180 without delay, so the synchronous XMLHttpRequest does not pause the script (pauseDuration is 0) and the script does not last 360 ms
the subtest of the module imported by import(): the engine evaluates the imported module in a reaction of the promise of the import, without a host hook, so its evaluation is reported in the resolve-promise script of the import (import.then), not as a module-script
the subtest Heavy render-phase style recalculation produces strictly positive styleDuration: the library does not measure the time it spends on the styles and the layout: styleDuration, layoutDuration and the forced*Duration of the scripts are 0 (with the virtual clock of the runner, which only advances when it is read, they would be 0 anyway)
the conditional user timing of the long animation frames (performance.markConditional(), performance.measureConditional() and userTimingEntries), a Chromium experiment that the specification does not define
the conditional user timing of the long animation frames (performance.markConditional(), performance.measureConditional() and userTimingEntries), a Chromium experiment that the specification does not define
the conditional user timing of the long animation frames (performance.markConditional(), performance.measureConditional() and userTimingEntries), a Chromium experiment that the specification does not define
the conditional user timing of the long animation frames (performance.markConditional(), performance.measureConditional() and userTimingEntries), a Chromium experiment that the specification does not define
the conditional user timing of the long animation frames (performance.markConditional(), performance.measureConditional() and userTimingEntries), a Chromium experiment that the specification does not define
the conditional user timing of the long animation frames (performance.markConditional(), performance.measureConditional() and userTimingEntries), a Chromium experiment that the specification does not define
the conditional user timing of the long animation frames (performance.markConditional(), performance.measureConditional() and userTimingEntries), a Chromium experiment that the specification does not define
the conditional user timing of the long animation frames (performance.markConditional(), performance.measureConditional() and userTimingEntries), a Chromium experiment that the specification does not define
the long animation frames of the dedicated workers (the congested moments), a Chromium experiment (LongAnimationFrameWorker) that the specification does not define
the promises of the platform have no creation-time attribution (optional, https://w3c.github.io/long-animation-frames/#webidl-monkey-patches): the reactions of the promises of the streams are reported as Promise.then, without source location, not as StreamPromise.resolve.then