This patch adds an option to Scaffold to disable resizing the body to avoid the
window padding. This lets developers create layouts that are stable when the
keyboard overlays the app.

Fixes #3565

We had a math-o.

Fixes #3563

* add google analytics

* send in the run target type

* track device type targets

* use the real GA code

* review comments

* rev to usage 2.0

* rev to 2.2.0 of usage; add tests

* review comments

Wait until the end of the microtask to tell gesture recognizers that
they've won in the gesture arena. This lets recognizers dispose reject
themselves at arbitrary times without triggering gestures in awkward
call stacks.

Fixes #3183

* add a benchmarking mode to flutter analyze

* change arg names

Fixes #3186

Fixes #3289

Previously we were locking down the state even when calling layout in
LazyBlock. Now we lock only when building children. Making this work well
involved moving the catch out of lockState and into the few callers who
actually wanted it.

Fixes #3534

Handle the case of a null body in response.dart

We're supposed to return a null string when the HTTP response doesn't have a
body. Also handle the case of not having a headers map.

This is currently hidden for development use.  It will select artifacts from
the android_{Debug,Release}_Deploy output in a local engine build.

Improve docs for MediaQueryData.size

* improve text for devices

* nit

* tweak text

* exclude files from build/ when looking for asset variants

* exclude some dirs when creating flx files

This would have answered a question we go on flutter-dev.

Fixes #3517

Add Flow layout

A flow layout is optimized for reposition children with transformation
matrices. A flow layout can animate the position of its children very
efficiently.

This patch contains a few minor improvements to the framework that I ran
across while working on a new layout model.

Fixes #3510

Somehow I missed these in my previous patch.

Fixes #3503

Prior to this patch, we were pushing two frames each time the cursor blinked.
In turning the cursor on or off, the markNeedsPaint call was triggering another
frame to be scheduled because we cleared a bit in the scheduler at the
beginning of the frame instead of at the end of the frame.

To implement scheduling correctly, we actually need two bits: one for
ensureVisualUpdate, which just promises to get to the end of the pipeline soon,
and scheduleFrame, which promises to get to the beginning of the pipeline soon.

(Reland)

This adds in particular the ability to track the time at which the
framework boots up, and the time at which we are confident we have
completed the first useful frame.

Fixes #3501

We don't need this assert anyway.

This reverts commit 161f945e.

This patch caused a number of regressions.

Fixes #3497

The old names were getting silly and started stepping on valuable namespace.

The new names are consistent and clear.

* Adds NineSliceSprite and optimizes gallery front page

* Refactored MaterialApp

* Remove IconThemeData.clampedOpacity

This reorders some classes so that this file makes more sense, and adds
a bunch of docs. It also makes the following changes:

* Move allElements from Instrumentation to TestWidgets. (Instrumentation
  is going away.)

* Remove findElements.

* Rename byElement to byElementPredicate

* Rename byPredicate to byWidgetPredicate

* Implement _WidgetPredicateFinder so that byWidgetPredicate has good
  messages

* Fix one use of byElementPredicate to use byWidgetPredicate.

Roll engine to 4062813e7060666f26b89e77d3f58035523bfd6b

Previously the widgets layer only provided a concrete binding, which
makes it awkward to extend it compared to other bindings. This moves
widgets to the same style as the other layers.

In a subsequent patch I'll use this to make the tests layer saner.