Reorder the events in process pending pointer capture #40
Conversation
Move the pointerleave/out events to be before the gotpointercapture and also relax the condition for sending those as the current spec does not consider the case that the capture is transferred from a node to another.
|
This looks good to me and seems like a logical simplification. This is still quite complicated, so I'd love to see some tests to go along with this change (but we don't necessarily need to block updating the spec). |
|
What was the status of this? Did we agree this was good to merge? If so, it'll need a rebase first |
|
No. I believe this is one of those issues that depends on whether we send boundary events while capturing or not. Do you want me to close this for now or is it better to leave it open? |
|
happy to leave it open for now until we figure out the bigger picture :) |
|
About the last 'if' of the modifications, in the case that the pointer capture target override is not set and set pending pointer capture to any node other than the hit test node, shall we still fire pointerover and pointerenter to the hit test node? |
|
I believe we will get rid of all boundary events line in "Process Pending Pointer Capture" section as the cleaner way of defining it would be to change the hit-testing in that case and the boundary events will follow the UI Events spec. So I need to get rid of this PR and upload another one for that proposal. |
|
With Chrome going forward with change in the definition of hit-testing while a pointer is captured, I will close this PR and create another one which includes that proposal in one PR. |
Following #39,
move the pointerleave/out events before the gotpointercapture
and also relax the condition for sending those as the current spec
does not consider the case that the capture is transferred from a
node to another.