Don't monkey patch HTML's list of activation triggering input event

[marcoscaceres]

Related fix on the HTML side:
whatwg/html#6818

---

Preview | Diff

[mustaqahmed]

We can land this now without waiting for the HTML change (because HTML needs to address whatwg/html#3849 anyway).

[marcoscaceres]

@mustaqahmed, looks like "activation triggering input event" vanished from HTML. Any idea what that should be now?

[mustaqahmed]

@mustaqahmed, looks like "activation triggering input event" vanished from HTML. Any idea what that should be now?

Sorry trying to understand what you meant by "vanished"...I can still see it here. Perhaps you meant a missing <dfn export> instead, which is being added in whatwg/html#6818?

[marcoscaceres]

Ah, indeed! I thought it was exported already! Sorry for the confusion on my part.

[mustaqahmed]

@marcoscaceres I got a chance to look at it again after a while, and now it seems to me we can significantly clarify the current dependence to HTML if we simply drop the "engagement gesture" section and fix the existing 5 references by:

1. saying requestPointerLock() is a Transient activation-gated API, and
2. adding the non-normative exception currently mentioned in the "Note" about "no engagement gesture is required".

This change would make the text more precise, and as a bonus point, we will be able to land this PR independently from whatwg/html#6818.

What do you think?

[marcoscaceres]

sounds great, @mustaqahmed. The only thing would be getting "activation triggering input event" exported from HTML, but what you suggest makes sense.

[mustaqahmed]

Please consider the suggestion above without getting blocked by a holistic (and possibly involved) change needed on the HTML side. Landing the PointerLock change here before the HTML change would still be a big improvement because the current "engagement gesture" section is too incomplete anyway (it doesn't mention PointerEvents or TouchEvents, for example).