Attach WebContents to HeadlessWindowTreeHost [chromium/src : main]
Viktor Semeniuk (Gerrit)
Viktor Semeniuk added 1 comment![]( "Open in Gerrit")
Hi Erik and Peter,
I would like to get your opinion on the following implementation. The CL is not ready to be submitted, I'm just looking for an early feedback.
Related details
- Erik Chen
- Peter Kvitek
Code-Coverage
Code-Owners
Code-Review
Review-Enforcement
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Erik Chen (Gerrit)
Erik Chen added 2 comments![]( "Open in Gerrit")
Hi Erik and Peter,
I would like to get your opinion on the following implementation. The CL is not ready to be submitted, I'm just looking for an early feedback.
conceptually, the idea of a making a custom window tree host to pump beginframe events makes sense. I'm not a graphics expert so I don't want to comment on the correct of the details of what you've written. Feel free to ping blundell to find an appropriate owner.
What is the plan for macOS support?
This approach seems simpler than the "hidden window" approach discussed over email. Can someone provide context about why we're moving in the hidden window approach vs this approach?
Related details
- Peter Kvitek
- Viktor Semeniuk
Code-Coverage
Code-Owners
Code-Review
No-Unresolved-Comments
Review-Enforcement
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Viktor Semeniuk (Gerrit)
Viktor Semeniuk added 2 comments![]( "Open in Gerrit")
Erik ChenHi Erik and Peter,
I would like to get your opinion on the following implementation. The CL is not ready to be submitted, I'm just looking for an early feedback.
conceptually, the idea of a making a custom window tree host to pump beginframe events makes sense. I'm not a graphics expert so I don't want to comment on the correct of the details of what you've written. Feel free to ping blundell to find an appropriate owner.
What is the plan for macOS support?
Adding @blun...@chromium.org as a reviewer.
What is the plan for macOS support?
There are no problems with rAF on MacOS. This caused plenty of confusion during the early development since testing the same site can lead to a different result depending on the OS.
This approach seems simpler than the "hidden window" approach discussed over email. Can someone provide context about why we're moving in the hidden window approach vs this approach?
Oh, hidden window approach just came up because @kvi...@chromium.org is working on it to solve b/479056794. Originally it has nothing to do with APC or rAF problem we are tying to solve here. I just contacted Peter since he works on a headless mode.
Related details
- Colin Blundell
- Erik Chen
- Peter Kvitek
Code-Coverage
Code-Owners
Code-Review
No-Unresolved-Comments
Review-Enforcement
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Colin Blundell (Gerrit)
Colin Blundell added 1 comment![]( "Open in Gerrit")
Erik ChenHi Erik and Peter,
I would like to get your opinion on the following implementation. The CL is not ready to be submitted, I'm just looking for an early feedback.
Viktor Semeniukconceptually, the idea of a making a custom window tree host to pump beginframe events makes sense. I'm not a graphics expert so I don't want to comment on the correct of the details of what you've written. Feel free to ping blundell to find an appropriate owner.
What is the plan for macOS support?
Adding @blun...@chromium.org as a reviewer.
What is the plan for macOS support?
There are no problems with rAF on MacOS. This caused plenty of confusion during the early development since testing the same site can lead to a different result depending on the OS.
Thanks! Viktor and I chatted offline, and he's going to send an email to chrome-compositor@ with the context and link to solicit advice from the most knowledgeable folks.
Related details
- Erik Chen
- Peter Kvitek
- Viktor Semeniuk
Code-Coverage
Code-Owners
Code-Review
No-Unresolved-Comments
Review-Enforcement
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Peter Kvitek (Gerrit)
Peter Kvitek added 2 comments![]( "Open in Gerrit")
Viktor SemeniukThis approach seems simpler than the "hidden window" approach discussed over email. Can someone provide context about why we're moving in the hidden window approach vs this approach?
Oh, hidden window approach just came up because @kvi...@chromium.org is working on it to solve b/479056794. Originally it has nothing to do with APC or rAF problem we are tying to solve here. I just contacted Peter since he works on a headless mode.
This minimalist WTH implementation looks fine to me, however, there is no guarantee that it will be enough to run all the existing and future password change pages. As you found out, some happen to need rAF, others may need something else.
The 'hidden browser window' approach runs pages in exactly the same way they would have run normally, now and in the future, so it seems like a more reliable in the long run.
class HeadlessWindowTreeHost : public aura::WindowTreeHost {Could we rename this to something not having `Headless`? `DetachedWindowTreeHost` or something like that.
Reason: Headless is already overloaded in Chrome source tree so it would be good to avoid adding another overload.
Related details
- Erik Chen
- Viktor Semeniuk
Code-Coverage
Code-Owners
Code-Review
No-Unresolved-Comments
Review-Enforcement
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |