STR is to scroll down carefully to move the dynamic toolbar you will see a red bar (NOTE: you don't need to use desktop mode) Now I am very confused. If the reason why youtube with desktop mode works on Chrome is just because Chrome's intersection observer accounts for the dynamic toolbar state, this attaching test case should work. But in fact it doesn't work. So there's definitely another reason why youtube with desktop mode works on Chrome. On the other hand, I've written an additional change upon D212920 to make youtube on desktop works by using viewport-ish metrics for our intersection observer implementation. It was simpler than I thought, thanks to D212920. So here are two options; a) Re-investigate why youtube with desktop mode works on Chrome and align our behavior match Chrome's behavior b) Fix this bug in a different way from what Chrome does I think a) is preferable, but it will take some more amount of time, given that investigating Chrome's behavior is harder than investigating our behavior (implementation). James, opinions?
Bug 1788504 Comment 24 Edit History
Note: The actual edited comment in the bug view page will always show the original commenter’s name and original timestamp.
STR is to scroll down carefully to move the dynamic toolbar you will see a red bar (NOTE: you don't need to use desktop mode) Now I am very confused. If the reason why youtube with desktop mode works on Chrome is just because Chrome's intersection observer accounts for the dynamic toolbar state, this attaching test case should work. But in fact it doesn't work. So there's definitely another reason why youtube with desktop mode works on Chrome. On the other hand, I've written an additional change upon D212920 to make youtube with desktop works by using viewport-ish metrics for our intersection observer implementation. It was simpler than I thought, thanks to D212920. So here are two options; a) Re-investigate why youtube with desktop mode works on Chrome and align our behavior match Chrome's behavior b) Fix this bug in a different way from what Chrome does I think a) is preferable, but it will take some more amount of time, given that investigating Chrome's behavior is harder than investigating our behavior (implementation). James, opinions?