This is just a proof of concept. If we like this direction, it will move out of the examples directory (likely re-written) and be committed in smaller pieces with unit tests and formal reviews.

TBR=abarth
BUG=

Review URL: https://codereview.chromium.org/971183002

Previously, we weren't scheduling a visual update when our surface was created,
so we'd just continue to show black until the author requested a new animation
frame. After this CL, we schedule a visual update as soon as our surface is
created.

As part of this change, I've removed knowledge of the GPU delegate from
PlatformView. Now, all the calls from PlatformView to the GPU system bounce
through the UI delegate, which serializes the commands with other commands from
the UI engine to the GPU rasterizer.

R=eseidel@chromium.org

Review URL: https://codereview.chromium.org/974483004

-Remove AbsoluteClipRects. It's never used.
-Inline some static functions. In the process,
noticed that the clip rects for CSS clip are broken.
In theory, this patch fixes that, but I didn't test it
since we probably want to just remove the feature anyways.

R=abarth@chromium.org

Review URL: https://codereview.chromium.org/964723004

These instructions don't work yet, but they're something to aim for.

R=eseidel@chromium.org

Review URL: https://codereview.chromium.org/968213005

This broke with https://codereview.chromium.org/965013003.
Before that patch, updateTransform was only called on the
RenderLayer itself. Now it's called on the RenderBox,
so we need to null check the RenderLayer.

R=abarth@chromium.org

Review URL: https://codereview.chromium.org/969893002

We now register the sensor service from SkyDemo, which shows that embedders of
SkyShell can register their own Java services.

R=eseidel@chromium.org

Review URL: https://codereview.chromium.org/969753003

double is a built-in type in Dart. float isn't...

TBR=eseidel@chromium.org

Review URL: https://codereview.chromium.org/968293003

RenderImage needs to notify its base class unconditionally so that it will
schedule a visual update.

This broke in https://codereview.chromium.org/791023006.

R=ojan@chromium.org
BUG=https://github.com/domokit/mojo/issues/55

Review URL: https://codereview.chromium.org/969863002

The touch demo needs some basic smoothing, right now
it looks rather jittery after this change.  But the pressure
code is definitely working!

R=abarth@chromium.org
BUG=

Review URL: https://codereview.chromium.org/970493003

The only thing that uses it is RenderTreeAsText.

R=abarth@chromium.org

Review URL: https://codereview.chromium.org/964023002

This patch makes it so that querystrings are no longer considered part of a mojo application's identity.

So if you connect to http://a?foo, then http://a?bar, you'll end up talking to the same application, just two different connections.

The URL of each connection is sent to the app via a new param in AcceptConnection().

This patch also adds some app tests that exercise http app loading, which wasn't tested before.

R=qsr@chromium.org

Review URL: https://codereview.chromium.org/943053003

This CL attempts to fix clean builds by adding a missing dependency from
service_registry to service_provider.

Also, I've attempted to fix the sky tests by updating the reference to
service_registry.mojom.dart.

TBR=hansmuller@chromium.org

Review URL: https://codereview.chromium.org/968553002

This is more member data on RenderLayer that is now only
used by RenderBox.

R=eseidel@chromium.org

Review URL: https://codereview.chromium.org/965013003

This is not complete gesture support by far, but it's a start.
MojoShell and Chrome use a C++ GestureDetector, this code attempts
to use the Android (Java) GestureDetector instead.

We probably should not be sending gesturetap until we've decided
it's not a scroll, etc, but this implementation does not go that far.

I had to fix a bug whereby we were assuming the InputEvent.time_stamp
was in TimeDelta's internal format, which was wrong.  When we get
time_stamp from Android its in ms since boot.

R=abarth@chromium.org

Review URL: https://codereview.chromium.org/969493002

We need to create a new service provider when we navigate to a new page
otherwise the new page is sad that the old page took its service provider.

R=eseidel@chromium.org

Review URL: https://codereview.chromium.org/962383003

It needs to be in a conventional public location to be available to consumer repos.

R=tonyg@chromium.org

Review URL: https://codereview.chromium.org/968493004

Apps can import this library if they wish to reload on shake.

R=eseidel@chromium.org

Review URL: https://codereview.chromium.org/963303004

This CL adds a sensor_service to sky/services and wires it into SkyShell
The plan is to eventually use this data to implement shake-to-refresh.

R=eseidel@chromium.org

Review URL: https://codereview.chromium.org/962043002

It doesn't do anything RenderLayer specific anymore. This
does mean adding another pointer to RenderBox. We might want
to do something about that eventually, but for now it seems fine
to make forward progress in terms of getting rid of the RenderLayer
tree.

R=eseidel@chromium.org

Review URL: https://codereview.chromium.org/963253002

ServiceRegistry enables a chain of Mojo applications to accumulate
services without wrapping and forwarding the incoming ServiceProvider
request.

Sky's DocumentView adds ViewManagerClient to the ServiceRegistry (if any)
that's provided to it via ConnectToApplication(). Sky applications can add
additional services with embedder.serviceRegistry.addServices().

R=abarth@chromium.org

Review URL: https://codereview.chromium.org/958673002

-Remove RenderLayerFilterInfo. It was just a static map for
FilterEffectRenderers. Instead, put the FilterEffectRenderer
directly on RenderLayer.
-Make FilterEffectRenderer not be RefCounted. This involved
deleting a ton of dead code around ReferenceFilters since they
were the other subclass of Filter. As best I can tell,
reference filters are already don't parse in Sky, so this should
just be removing dead code.

R=abarth@chromium.org

Review URL: https://codereview.chromium.org/962543003

This change causes the generated abstract class having the same name as the interface to contain only the interface's method declarations. The generated Proxy class then implements the base class. In addition to implementing the interface methods, the generated Proxy class has one private field _proxyImpl, which is a MojoEventStreamListener and manages sending messages and receiving responses. Operations on the ProxyImpl (close, bind, etc.) are exposed through generated utility functions.

The generated Stub is largely as before with the difference that a class providing a service will implement the mojo interface and *have* a Stub rather than be a Stub. Where appropriate, this change also calls listen() immediately where a Stub is constructed.

BUG=
R=hansmuller@google.com, sky@chromium.org

Review URL: https://codereview.chromium.org/959993002

This makes debugging that case much easier.

R=abarth@chromium.org
BUG=

Review URL: https://codereview.chromium.org/958203002

-Move perspectiveOrigin to RenderBox. It's only used in
one place and called on a RenderBox.
-Inline some methods that were just calling out to the renderer.
-Inline updateStackingNode() into the constructor and remove
the branch since requiresStackingNode is always true.
-Delete a bunch of dead code.

R=abarth@chromium.org

Review URL: https://codereview.chromium.org/961053002

Before this patch it crashed.

Also updated the touch-demo to support multi-touch.

R=abarth@chromium.org
BUG=

Review URL: https://codereview.chromium.org/961483004

This CL renames SkyShell.apk to SkyDemo.apk and adds a Stocks.apk which
defaults to the stocks demo app. Sky Shell is now a library that can be
packaged by different APKs.

R=eseidel@chromium.org

Review URL: https://codereview.chromium.org/965493002

Inline a couple methods that are only called from one place
and remove a bunch of dead code.

R=abarth@chromium.org

Review URL: https://codereview.chromium.org/961553003

This code is unused. It's not even built.

R=eseidel@chromium.org

Review URL: https://codereview.chromium.org/963513004

This add the missing RequestNavigateHistory method on sky::Engine.

It is currently not implemented.

TBR=abarth@chromium.org

Review URL: https://codereview.chromium.org/957973003

-m_hasSelfPaintingLayerDescendant is never read.
-remove showLayerTree. The layer tree is going away. We won't need this.

R=abarth@chromium.org

Review URL: https://codereview.chromium.org/952343003

TBR=eseidel@chromium.org

Review URL: https://codereview.chromium.org/952213005

However WebView::close() no longer crashes.  close() is never called
in MojoShell since mojo can't shutdown yet.

I tried closing the existing WebView and replacing it
but somehow that caused it to only draw red.  After a while
of looking at this with abarth we decided to just load into
the same WebView for now.

Eventually we should do something smarter where we start the
provisional load and only replace the webview once the new one is
ready, but that's a later CL.

R=abarth@chromium.org
BUG=

Review URL: https://codereview.chromium.org/952273003

This improves consistency with the convention that the dart: scheme is
only for things from the Dart standard library.

BUG=
R=erg@chromium.org

Review URL: https://codereview.chromium.org/951783004

This should let SkyShell load images.

R=eseidel@chromium.org
BUG=https://github.com/domokit/mojo/issues/52

Review URL: https://codereview.chromium.org/959773005

This is the only remaining caller.

R=abarth@chromium.org

Review URL: https://codereview.chromium.org/954913002

R=abarth@chromium.org

Review URL: https://codereview.chromium.org/943333005

https://codereview.chromium.org/899753003 incorrectly passed down the
RenderView's clipping rect all the down through transforms instead of
using the one from the LayerPaintingInfo, which is adjusted for transforms
as you recurse down the tree.

TBR=abarth@chromium.org

Review URL: https://codereview.chromium.org/957643002

Originally reviewed for the branch on:
https://codereview.chromium.org/945083002

TBR=abarth@chromium.org

Review URL: https://codereview.chromium.org/953933004

This was landed on the skyshell branch and I forgot to pull it over:
https://codereview.chromium.org/946583003

TBR=abarth@chromium.org

Review URL: https://codereview.chromium.org/954893002

-Consolidate the beginTransparencyLayers calls. The call
on layer()->parent() was just not needed as best I can tell
and the filter one could just be shared with the non-filter
codepath. FWIW, blink has already made a similar set of changes.
-Don't walk up the tree calling beginTransparencyLayer.
We're doing this in a recursive call down, so we've already
created the transparency layer for any transparent ancestors.
-Remove unused TransparencyClipBoxMode argument.
-Inline the paintingExtent method.
-Removed unused REVEAL_TRANSPARENCY_LAYERS #define.

R=abarth@chromium.org

Review URL: https://codereview.chromium.org/959543002