Over time features/base/util became a bucket where people seemed to dump
just about anything they couldn't think of a better place for. That's a
trend I don't like encouraging. Given that roomnameGenerator.js is
currently used in features/welcome only, I'm fine with moving it there
for the greater good.

Because timeUtil.js computes hours, minutes, and seconds out of a single
time/duration using three separate functions, I wouldn't recommend using
it, especially reusing it. That's why I'm even making the functions
private to their current use location.

I don't like the file/function name, I'm not excited about the
complexity of the logic it implements, and it's definitely a reusable
piece worthy of being called a utility.

Atlaskit at times will have localized styling for font-size and
sometimes will not. The button component will inherit its
font-size whereas selectors have localized font-size of 14px. For
consistency, the cancel/submit buttons on the atlaskit modals
will also have 14px. The atlaskit story book examples also use
buttons with 14px font-size.

Also move (native) avatar style to film-strip styles, since  that's where it
applies. This is analogous to how the large-view avatar is styled.

The Device Selection modal consists of:
- DeviceSelection, an overly smart component responsible for
  triggering stream creation and cleanup.
- DeviceSelector for selector elements.
- VideoInputPreview for displaying a video preview.
- AudioInputPreview for displaying a volume meter.
- AudioOutputPreview for a test sound output link.

Store changes include is primarily storing the list of
available devices in redux. Other app state has been left
alone for future refactoring.

This reverts commit 739298c782.

When a conference is to happen in a domain which is not the defaut, its config
is loaded and set. As part of this process, lib-jitsi-meet is disposed. Because
disposing is asynchronous, events happen in this sequence:

- set new config
- dispose lib (which effectively wipes the config)
- init lib

This results in the library to be initialized without the loaded config, which
was lost. This commit fixes that by delaying setting the config and
re-initializing the library until it was disposed.

JitsiConnection.initJitsiConference doesn't automatically pass the global config
options, so grab the config from the Redux store and pass it.

It's now natively supported:
https://github.com/oney/react-native-webrtc/pull/244

The functionality to use the react-native-webrtc custom API for fast switching
cameras was moved to JitsiLocalTrack in lib-jitsi-meet. Use that.

Ref: https://github.com/jitsi/lib-jitsi-meet/pull/444

The behavior can be triggered with the toggleAudioOnly action, which is
currently fired with a button.

The following aspects of the conference will change when in audio only mode:

- local video is muted
- last N is set to 0 (effectively muting remote video)
- full-screen mode is exited
- audio mode is set to "audio chat" (default output is the earpiece)
- the wake lock is disengaged

One aspect not handled in this patch is disabling the video mute button while in
audio only mode. The user should not be able to turn back video on in that case.

They better represent if a participant has video available or not. There are
cases when even a participant in the last N set would not have video because it
disconnected momentarily, for example.

Use the curstom _switchCamera API provided by react-native-webrtc to toggle the
camera instead of destroying the current track and creating a new one.

_switchCamera is implemented at a low level, so the track perceives no changes,
thus being a lot faster and less involved since the capturer doesn't need to be
destroyed and re-created.

In addition, don't mirror the video for the back camera.

Ref: https://github.com/oney/react-native-webrtc/pull/235

When a new local video track is created an associated video capturer is created
for it. The cause for the freezes seems to be creating mutliple tracks (which
come with a video capturer each). Fix this by first disposing of the previous
video track before creating the new one.

Ref:
https://github.com/oney/react-native-webrtc/issues/209#issuecomment-281482869

Lib-jitsi-meet does not really implement isScreenSharing. Besides,
getCameraFacingMode will already make sure that the video track does not
represent a desktop stream.

This is in preparation for an upcoming "audio only mode" feature. Setting last N
will also be required for it, so this patch factors out the action and makes it
public so other modules can reuse it.

In addition, if the value is set to undefined the configured default value (or
-1 if absent) is picked.

Replaces changeAvatarID, changeAvatarURL and changeEmail with
participantUpdated action.
participantUpdated can be fired for local user without id. This
fixes the problem with updating the local user before the user
join the conference which results in fix for failing to execute
commands for avatarID, avatarURL and email right after the iframe
api creates the iframe with Jitsi Meet.

Split language detectors to be web/native dependent. Take care of strings that contain html.

Adds translate function with default namespaces and options.

The error raised by JitsiMeetJS.init() is already in the state of
features/base/lib-jitsi-meet so it's not a good design to store the same
error in the state of features/unsupported-browser.