Bubbling model for SpeechSynthesisUtterance

[dontcallmedom]

The spec states that

[SpeechSynthesisUtterance Events] bubble up to SpeechSynthesis
https://wicg.github.io/speech-api/#utterance-events

Is that actually implemented anywhere? if so, the spec should describe the get the parent relationship between utterances on the parent SpeechSynthesis, and the relevant on… event handlers should be added to the SpeechSynthesis interface.

If not, the spec should remove that mention.

[foolip]

From quick testing in Chromium, the "start" event at least does not bubble, and additionally event.composedPath() is the empty array, so there must be no "get the parent" implementation for SpeechSynthesisUtterance.