Closed mustaqahmed closed 7 years ago
Can the transient activation actually be propagated by this kind of API? To know when the transient activation is ... active, the iframe needs to receive a message from the parent, but the postMessage
should probably propagate the gesture on its own, without needing the parent to opt in globally.
Yes, transient activation is currently tracked in Chrome through UserGestureToken
s, for which this kind of propagation perhaps doesn't make sense. But I was thinking ahead to our experimental user activation model which proposes to switch the token-based transient activation to a document-state based model, like what Chrome does for the sticky activation (for media playback) today. The goal there is to make the whole user activation model simple enough so that all browsers will be willing to implement.
So, yes, this issue can wait until we have the simple model implemented.
This proposal is meant to delegate the bit of "browsing context has been user activated" as the two features it is meant to be used with are using this information.
@domenic raised this question in a separate thread:
Do we need the distinction in the delegation API?
Alternatively, is it better to simplify the delegation API by substituting the list of APIs in
allowedActivationDelegation
by a list of "types of user activations to delegate"? Like "sticky" (for all media-like use cases) or "transient" (for all popup-like use cases) or "both"?