New issue
Advanced search Search tips

Issue 682471 link

Starred by 3 users

Issue metadata

Status: WontFix
Owner:
Closed: Apr 2017
Components:
EstimatedDays: ----
NextAction: ----
OS: Android
Pri: 2
Type: Bug
Proj-VR
Proj-XR
Proj-XR-VR

Blocking:
issue 668541



Sign in to add a comment

Daydream View capacitive nubs can abort WebVR presentation

Project Member Reported by bajones@chromium.org, Jan 19 2017

Issue description

From an external developer: "We are using the NFC triggered onvrdisplayactivate event to switch the user to WebVR mode. But sometimes the WebVR load process fails if a user inserts the phone in the viewer, closes the lid fast enough, and the cylindrical-button on the viewer touches the address bar section of the browser and interrupts the WebVR flow. To circumvent this, we are now switching to fullscreen mode first and then to WebVR mode (which would now require iframe embeds of the player to have both of these attributes)"

It's worth checking to see if there's anything we're doing to delay the time-to-DON after requestPresent is called. If it's not something we can control, maybe we could have a brief grace period after requestPresent during which the browser chrome ignores touches. (Would need to be limited to prevent abuse.)
 
This is one of the things I've been thinking about for  issue 668541 , which I haven't had time to work on yet.

We can probably get away with speculatively setting up the VR view hierarchy before the DON flow, and when returning from headset insertion that bypasses the DON flow and waits for the onvrdisplayactivate requestPresent.

Comment 2 by sko...@chromium.org, Jan 19 2017

Blocking: 668541
Labels: -Pri-3 M-58 Pri-2
Owner: mthiesse@chromium.org
Status: Assigned (was: Untriaged)
Michael, is this fixable on its own or do we need to fix all of  Bug 668541 ?  If not feel free to close this as a Dupe.
Well I guess  issue 668541  is more of an umbrella bug? This would be one aspect of the transition improvements we need to make, but there may be others.

Having this block 668541, but remain a separate issue makes sense to me.
Labels: -M-58 M-59
Status: WontFix (was: Assigned)
I think the removal of the fullscreen hacks with the compositor bypass fixed this. Can no longer reproduce.
Components: Blink>WebXR

Sign in to add a comment