Clarify consumption of user activation at the initiation step

[mustaqahmed]

Closes #13.

[mustaqahmed]

@annevk PTAL if this provides enough clarification.

[annevk]

I guess that makes sense, but shouldn't it also be consumed in the document that is activated and wants to delegate?

[mustaqahmed]

I guess that makes sense, but shouldn't it also be consumed in the document that is activated and wants to delegate?

Hah, finally seeing your main concern! I agree: starting the consumption at the sender's Window makes the most sense here (even though user activation consumption "resets" the whole frame-tree, so consuming anywhere in the tree should be equivalent).

Wondering what's the right way to specify the sender's Window in this algorithm. Is this "relevant global object of current Window"?

[annevk]

I think you want incumbentSettings's global object for this particular algorithm. This is one of the few if not only case where we want to use the incumbent global.

[mustaqahmed]

I think you want incumbentSettings's global object for this particular algorithm. This is one of the few if not only case where we want to use the incumbent global.

Done.

[mustaqahmed]

@annevk PTAL, finally got a chance to incorporate your suggestions. Sorry for the delay!

[mustaqahmed]

As we promised in Chrome's intent to ship thread, we are planning to send out a PR to HTML. I think it makes sense if we merge this PR now and start a review on the HTML side for the whole "algorithm".

[annevk]

I wish some more work was done on flushing out #9 and #10 (see also w3ctag/design-reviews#655 (comment)). #15 doesn't seem touched at all.

I suppose we can work on HTML integration, but it's not totally clear we have a shared understanding of how this would all work.

[mustaqahmed]

I wish some more work was done on flushing out #9 and #10 (see also w3ctag/design-reviews#655 (comment)). #15 doesn't seem touched at all.

I suppose we can work on HTML integration, but it's not totally clear we have a shared understanding of how this would all work.

That's fair, thanks. Deferring the HTML integration plan until we have addressed them.

As for this PR, do you think it needs more work to close #13?

[annevk]

Yeah this looks okay. Thanks!