Testcase: > document.body.addEventListener("webkitTransitionEnd", evt => console.log(evt)); > document.body.dispatchEvent(new TransitionEvent("transitionend")); Another testcase: > document.body.addEventListener("mozfullscreenchange", evt => console.log(evt)); > document.body.dispatchEvent(new Event("fullscreenchange")); This is probably not usually a big deal, but it's not quite spec-comform anyway. (This was inspired by bug 1268749 comment 17)

For the webkitTransitionEnd example, isn't this the desired behavior, per https://dom.spec.whatwg.org/#concept-event-listener-invoke? (Possibly I'm missing something, sorry.)

So even with events dispatched by the content, we need to try prefixed handler as well? This behavior seems weird. It seems Chrome does so as well, and given that's the desired behavior from the spec, I guess this is INVALID, then.

> This behavior seems weird. Yeah, the whole thing is weird IMO. ^_^