New issue
Advanced search Search tips
Note: Color blocks (like or ) mean that a user may not be available. Tooltip shows the reason.

Issue 601865 link

Starred by 4 users

Issue metadata

Status: Duplicate
Merged: issue 689696
Owner: ----
Closed: Jun 2017
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: All
Pri: 3
Type: Feature
mus



Sign in to add a comment

*OutputSurface => components/display_compositor/surfaces

Project Member Reported by fsam...@chromium.org, Apr 8 2016

Issue description

Surfaces should live in a display_compositor subdirectory.

*OutputSurface files from content/browser/compositor might live in there too?
 
Hy

Comment 2 by danakj@chromium.org, Apr 14 2016

Cc: enne@chromium.org

Comment 3 by danakj@chromium.org, Apr 14 2016

We had a convo about MUS the other day and it came up that we might move raster to be after surface aggregation one day.. which is to say, moving tile manager and layers and stuff.. into the display compositor again.

So enne and I are thinking maybe it's not worth moving the contents of cc/surfaces to display_compositor as that's probably going to be some weird churn.

However all the embedder stuff (things in c/b/compositor/ for the display compositor) should probably continue to become a separate component like display_compositor, and that would just be the embedder of code in cc/surfaces/.


Status: WontFix (was: Untriaged)
sgtm Marking as WontFix

Comment 5 by danakj@chromium.org, Apr 14 2016

(The *OutputSurface part of the title still applies though, maybe that's elsewhere?)
Status: Available (was: WontFix)
Back to available. :-)
Summary: *OutputSurface => components/display_compositor/surfaces (was: cc/surfaces and *OutputSurface => components/display_compositor/surfaces)
Labels: mustash2 dcrefactor mustash mus
Components: MUS
Labels: mustash1 displaycompositor OS-All
Status: WontFix (was: Available)
Given we want to get rid of OutputSurface now, I'm marking as WontFix.
OutputSurface is going to continue to exist for display compositor (or the functionality inside it).

It's going away for LayerTreeHost.
Status: Available (was: WontFix)
So.. un-WontFix'd unless I'm misunderstanding.
Reopened then.. Hmm, this bug is a bit ambiguous then. I guess part of this bug is for whoever picks it up to figure out what all the *OutputSurfaces are used for in content/browser/compositor and which can and should move to display_compositor.
All of them can move except the one passed to ui::Compositor.
Cc: sadrul@chromium.org
I started looking at this, and it looks like: because of some of the dependencies of content/b/c/*output_surface*, the following also need to move into /components/display_compositor/: ImageTransportFactory (the pure interface), ReflectorTexture, ReflectorImpl, OwnedMailbox. Does that sound OK, or would it be necessary to keep them in content/b/c/?
Hmm, I don't think we want to use the current content/ version of reflector. We also don't want display_compositor to depend on ui/compositor.
Did y'all decide what you want reflector to look like (inside mus or inside the browser)?
Moving OwnedMailbox sounds okay fwiw.

ImageTransportFactory is used throughout content, and is what it is meant for. What pieces of it need to move to display_compositor? I think decoupling c/b/compositor from that interface and using GPTF directly is probably part of the pre-req work here.
re: Reflector: reimplement in terms of Surfaces. The interior details are TBD.
Components: Internals>MUS
Labels: Proj-Mustash
Components: Internals>Compositing
Labels: Type-Feature
Cc: weiliangc@chromium.org
Cc: varkha@chromium.org
Blocking: 732572
Mergedinto: 689696
Status: Duplicate (was: Available)
Blocking: -732572
Blocking: -601863
Components: -Internals>MUS Internals>Services>WindowService
Components: -MUS

Sign in to add a comment