Instead have VideoLayout reach into redux.

- Re-use the native redux pinning implementation for web
- Remove pinning logic from conference.js
- To the native pinning add a check for sharedVideo so
  youtube videos do not send a pin event
- Add shared videos as a participant to enable pinning and
  so they can eventually get added to the filmstrip
- Emit UIEvents.PINNED_ENDPOINT from middleware

* feat(quality-slider): initial implementation

- Add new menu button with an Inline Dialog slider for
  selecting received video quality.
- Place P2P status in redux store for the Inline Dialog
  to display a warning about not respecting video quality
  selection.
- Respond to data channel open events by setting receive
  video quality. This is for lonely call cases where a
  setting is set before the data channel is open.
- Remove dropdown menu from video status label and clean
  up related js and css.

* first pass at addressing feedback

- Move VideoStatusLabel to video-quality directory.
- Rename VideoStatusLabel to VideoQualityLabel.
- Open VideoQualitydialog from VideoQualityLabel.
- New CSS for making VideoQualityLabel display properly.
- Do not render VideoQualityLabel in filmstrip only instead of hiding with css.
- Remove tooltip from VideoQualityLabel.
- Show LD, SD, HD labels in VideoQualityLabel.
- Remove action SET_LARGE_VIDEO_HD_STATUS from conference.
- Create new action UPDATE_KNOWN_LARGE_VIDEO_RESOLUTION in large-video.
- Move VideoQualityButton into video-quality directory.
- General renaming (medium -> standard, menu -> dialog).
- Render P2P message between title and slider.
- Add padding to slider for displacement caused by P2P message's new placement.
- Fix display issue with VideoQualityButton displaying out of line in the
  primary toolbar.

* second pass at addressing feedback

- Fix p2p inline message color
- Force labels to break on words
- Resolve rebase issues, including only dispatching quality
  update on change. Before there was double calling of dispatch
  produced by an IE11 workaround. This breaks now when setting
  audio only mode to true twice.
- Rename some instances of quality to definition

* rename to data channel opened

* do not show p2p in audio only

* stop toggle audio only icon automatically

* remove fixme about toolbar button

* find closest resolution for label

* toggle dialog on button click

* redo last commit for both button and label

* ref(base/conference): add tracks before join

Sometimes it will be suboptimal to add local tracks to the conference,
after the room has been joined. It may slow down the session initiation
process by having to send unnecessary 'source-add' notifications.

* squash: fix typos/comments

Simplify the code by using a bitfied instead of a couple of boolean flags. This
allows us to mute the video from multiple places and only make the unmute
effective once they have all unmuted.

Alas, this cannot be applied to the web without a massive refactor, because it
uses the track muted state as the source of truth instead of the media state.

Currently lib-jitsi-meet looks there in case the `anonymousdomain` config option
was specified.

While this commit alone doesn't add support for authenticated deployments, it
avoids a failure if `anonymousdomain` was set, regardless of authentication being
turned on or not.

Fixes: https://github.com/jitsi/jitsi-meet/issues/1858

Refactors the previous "[RN] Cache avatars and provide a default in
case load fails" for the purposes of simplification but also modifies
its functionality at the same time. For example:

- Always displays the default avatar immediately which may be seen if
  the remote avatar needs to be downloaded.
- Does not use random colors.
- Uses a default avatar image which is not transparent and ugly but at
  least it's the same image that's used on Web. I've started talks to
  have images/avatar2.png replaced with a transparent and beautiful
  so that will land later on and we'll see the automagic colors in all
  their glory then.

Avatars are cached to the filesystem and loaded from there when requested again.
The cache is cleaned after a conference ends and on application startup
(defensive move).

In addition, implement a fully local avatar system, which is used as a fallback
when loading a remote avatar fails. It can also be forced using a prop.

The fully local avatars use a user icon as a mask and apply a background color
qhich is picked by hashing the URI passed to the avatar. If no URI is passed a
random color is chosen.

A grace period of 1 second is also implemented so a default local avatar will be
rendered if an Avatar component is mounted but has no URI. If a URI is specified
later on, it will be loaded and displayed. In case loading the remote avatar
fails, the locally generated one will be used.

* feat(presence): display status in thumbnail and large video

- Create a React Component for displaying presence. It currently
  connects to the store for participant updates but in the future
  should not be as smart once more reactification occurs.
- Modify filmstrip css so the presence status displays horizontal
  center and below the avatar.
- Modify videolayout css so the presence status displays horizontal
  centered and with a rounded background.
- Dispatch presence updates so the participant state can be update.
- Update message position on large video update to ensure message
  positioning is correct.

* squash: do not show presence message if connection message is displayed

Make toolbar's microphone button enabled whenever there are any
'audioinput' devices available and allow to add audio during
the conference even if microphone permissions were denied on startup.

Whatever I've tried so far ends up failing in really weird ways, so let's admit
defeat, for now. Destroy containers only on Android.

This shall be revisited when we update RN to version >= 0.43 and we have
"display: 'none'" available.

Web's ExternalAPI accepts an object with properties as one of its
constructor arguments and from which it generated a URL. Mobile's
JitsiMeetView.loadURLObject is supposed to accept pretty much the same
object.

Introduces loadURLObject in JitsiMeetView on Android and iOS which
accepts a Bundle and NSDictionary, respectively, similar in structure to
the JS object accepted by the constructor of Web's ExternalAPI. At this
time, only the property url of the bundle/dictionary is supported.
However, it allows the public API of loadURLObject to be consumed. The
property url will be made optional in the future and other properties
will be supported from which a URL will be constructed.

When a Container is not visible there is no need for it to react to touch
events, thus avoid wrapping it in a touch component.

In addition, simplify the style needed for hiding the component. Moving the view
out of the window boundaries no longer works on RN 0.42 on iOS. Seting the size
to 0 works well on both platforms, but in the future (when we upgrade to RN >=
0.43) we should switch to display: none:
4d69f4b2d1

Because on web video track is stored both in redux and in 'localVideo'
field, video is attempted to be unmuted twice when turning off the audio
only mode. This will crash the app with 'unmute operation is already in
progress'. This commit will prevent from taking action from the web
world if the video track already exists and will make the redux side
rollback unmuted status in case unmute fails.

In another installment of "how on earth?!", using a 1px border instead of a
0.2px border fixes view clipping on a device where it didn't work before
(Moto X Play).

Observations:

- When nothing was rendered, rotating the device made the buttons show up, this
  makes me think the Surface is not properly composited with the toolbar view
  for some reason. Why this happens in some devices and not in others remains a
  mistery.

Other approaches attempted:

- Setting View.collapsed to false so it will remain in the view hierarchy. It
  made no effect.

- Setting View.needsOffscreenAlphaCompositing to true. It made no effect.

Just like before, I came up with this workaround by accident, but couldn't find
another way, so here we go again.

This essentially reverts
023359b9d2

In the filmstrip we keep a container full of thumbnail views. Destroying these
every time we want we want to hide it is costly, as new renderers have to be
recreated, and they lack context, so there is an increased chance for "black
thumbnails".

When the 'startAudioOnly' config option is set to true the conference
will start in the audio only mode.

The speakerStatus field already exists as part of the objects
in the participant reducer. When the library updates the
connection status of a user, plumb that update through to redux.

navigator.product is read-only, so we don't run into the risk of it being
changed.

The end goal of this patch was to avoid opening the camera when there is no
welcome page.

In order to achieve this, the logic for creating the local tracks was
refactored:

Before this patch local tracks were created when lib-jitsi-meet was initialized,
and destroyed when it was deinitialized. As a side note, this meant that when a
conference in a non-default domain was joined, local tracks were destroyed and
recreated in quick succession.

Now, local trans are created and destroyed based on what the next route will be,
and this happens when the target room has been decided. This allows us to create
local tracks the moment we need to render any route, and destroy them when there
is no route to be rendered. As an interesting byproduct, this refactor also
avoids the destruction + recreation of local tracks when a conference in a
non-default domain was left.

The full-screen module uses the HIDE_DIALOG action to re-engage the full-screen
mode.

* feat(local-video): convert to react

- Create a VideoTrack component for displaying a video element.
  This mirrors native also having a VideoTrack component.
- The VideoTrack component does not let React update it to prevent
  the video element from re-rendering, which could cause flickers
  and would not work with temasys's overriding of the video element.
- VideoTrack extends AbstractVideoTrack to mirror native
  implementation and to get the dispatch of the onplaying event.
- Remove the onclick handler on the video element. Honestly, I
  didn't get it to work, and did not try, but it is also unnecessary
  because another handler already exists on the video wrapper.

* ref(device-selection): VideoInputPreview uses VideoTrack to show video

* squash into conversion: change css selectors

* squash into conversion: mix in abstract props

* squash into conversion: change shouldComponentUpdate check

* squash: update comment about why triggerOnPlayingUpdate is used

Listeners were set for when a track muted or changed its video
type, but the listeners were never removed. This would could
cause events to keep firing on the removed tracks, which would
cause redux to fire and error because the tracks were no longer
known. That the tracks still fire events after removal is
another issue...

Dynamically enables/disables the toolbar video button. Prior to that
commit if we would start with no video there would be no way to enable
it later on.

* Fix(AddPeopleDialog): Fixes error state and close dialog

* (to-squash) Addresses comments

Keep track of the connection and conference objects so we can leave and / or
disconnect early, before the connection is established or the conference joined.