This PR changes the logic for connecting / disconnecting conferences. Instead of
doing it in mount / unmount events from the Conference component, it moves the
logic to the appNavigatee action.

This fixes a regression introduced in 774c5ecd when trying to make sure the
conference terminated event is always sent.

By moving the logic to appNavigate we no longer depend on side-effects for
connecting / disconnecting, and the code should be more maintainable moving
forward.

An improvement to this is the concept of sessions, which, while not tackled
here, was taken into consideration.

Dear reader, I'm not proud at all of what you are about to read, but sometimes
life just gives you lemons, so enjoy some lemonade!

Joining a conference implies first creating the XMPP connection and then joining
the MUC. It's very possible the XMPP connection was made but there was no chance
for the conference to be created.

This patch fixes this case by artificially genrating a conference terminated
event in such case. In order to have all the necessary knowledge for this event
to be sent the connection now keeps track of the conference that runs it.

In addition, there is an even more obscure corner case: it's not impossible to
try to disconnect when there is not even a connection. This was fixed by
creating a fake disconnect event. Alas the location URL is lost at this point,
but it's better than nothing I guess.

Use subscribers to detect state change and emit those
out to other participants. Use middleware to register
the command listener.

If tracks are created while the conference is in the 'joining' state
they will never be added.

🔥🔥🔥

Using anything non-serializable for action types is discouraged:
https://redux.js.org/faq/actions#actions

In fact, this is the Flow definition for dispatching actions:

declare export type DispatchAPI<A> = (action: A) => A;
declare export type Dispatch<A: { type: $Subtype<string> }> = DispatchAPI<A>;

Note how the `type` field is defined as a subtype of string, which Symbol isn’t.

Dominant speaker events can trigger local participant updates
without a display name. Do not update the name unless there
is an explicit update in the action.

Analytics is erroring when unpinning because the logged
event sends null for the objectId. The objectId should
be the id of the person getting unpinned.

Extracts methods which share the common logic. There are still some
leftovers on the web side left which are not used on RN. But this can be
a first step.

* feat(tile-view): initial implementation for tile view

- Modify the classname on the app root so layout can adjust
  depending on the desired layout mode--vertical filmstrip,
  horizontal filmstrip, and tile view.
- Create a button for toggling tile view.
- Add a StateListenerRegistry to automatically update the
  selected participant and max receiver frame height on tile
  view toggle.
- Rezise thumbnails when switching in and out of tile view.
- Move the local video when switching in and out of tile view.
- Update reactified pieces of thumbnails when switching in and
  out of tile view.
- Cap the max receiver video quality in tile view based on tile
  size.
- Use CSS to hide UI components that should not display in tile
  view.
- Signal follow me changes.

* change local video id for tests

* change approach: leverage more css

* squash: fix some formatting

* squash: prevent pinning, hide pin border in tile view

* squash: change logic for maxReceiverQuality due to sidestepping resizing logic

* squash: fix typo, columns configurable, remove unused constants

* squash: resize with js again

* squash: use yana's math for calculating tile size

Move away from middleware and instead update video quality
when the selected video quality updates in redux. This also
lead to removing of automatically exiting audio only because
with the change it's not so readily possible to tell if the
user switched off audio only by re-selecting the already
preferred video quality. Removing this automagic removed
some additional checking done for mobile.

The internal max will be used for tile view. Whatever the
user has set for preferred video quality, the internal
maximum will be respected. This allows for the case where
the user prefers high definition video, but in tile view
it only makes sense to send low definition; ux wise the
user is allowed to continue messing with the video quality
slider.

- "preferred" is being appended because in tile view there is a
  concept of what the user prefers to be the maximum video quality
  but there is also a maximum respected internall. For example,
  the user may prefer HD, but in tile view the tiles may be small
  so internall the preferred would be set to LD.
- "receive" is being renamed to "receiver" to be consistent with
  the naming in lib-jitsi-meet.

* Show subtitles when Jigasi sends transcription results in JSON

* fix: Import PropTypes from prop-types.

* apply feedback on initial PR

* Changed Object to Map, alphabetic ordering fixes ,css changes in transcription subtitles

* Sends Map of transcriptMessages as prop to Component

* Documentation fixes and uses config in redux state

* Minor doc fix

* rename feature 'transcription' to 'subtitles'

* Moves subtitles config to interfaceConfig and minor fixes

* minor lint fix

Unfortunately, as the Jitsi Meet development evolved the routing mechanism
became more complex and thre logic ended up spread across multiple parts of the
codebase, which made it hard to follow and extend.

This change aims to fix that by rewriting the routing logic and centralizing it
in (pretty much) a single place, with no implicit inter-dependencies.

In order to arrive there, however, some extra changes were needed, which were
not caught early enough and are thus part of this change:

- JitsiMeetJS initialization is now synchronous: there is nothing async about
  it, and the only async requirement (Temasys support) was lifted. See [0].
- WebRTC support can be detected early: building on top of the above, WebRTC
  support can now be detected immediately, so take advantage of this to simplify
  how we handle unsupported browsers. See [0].

The new router takes decissions based on the Redux state at the time of
invocation. A route can be represented by either a component or a URl reference,
with the latter taking precedence. On mobile, obviously, there is no concept of
URL reference so routing is based solely on components.

[0]: https://github.com/jitsi/lib-jitsi-meet/pull/779

* fix(connection): reload immediately on possible split-brain

There isn't an explicit way to know when a split brain
scenario has happened. It is assumed it arises when an
"item-not-found" connection error is encountered early
on in the conference. So, store when a connection has
happened so it be calculated how much time has
elapsed and if the threshold has not been exceeded
then do an immediate reload of the app instead of
showing the overlay with a reload timer.

* squash: rename isItemNotFoundError -> isShardChangedError

Up until now, mobile was oblivious to participants' presence state. Presence
state handling is required (probably, amongst other things) for "call flows".
So, let's add it! This is done by gathering the presence state when a
participant first joins, and handling subsequent changes.

* feat(settings): setting dialog

- Move device selection, profile edit, language select, moderator
  options, and server auth into one modal with tabs.
- Remove side panel profile and settings and logic used to update
  them.
- Pipe server auth status into redux to display in the settings
  dialog.
- Change filmstrip only device selection popup to use the new
  stateless settings dialog component.

* squash: do not show profile tab if not guest

* squash: profile button not clickable if no profile to show

* squash: nits

* ref: Settings dialog.

On web CONFERENCE_FAILED handlers are not setting the 'recoverable'
flag thus any middleware which rely on those should not execute on web.

Emits CONFERENCE_FAILED in response to CONNECTION_FAILED event
which then triggers JitsiConference.leave() through the middleware
processing. Also base/conference state will be adjusted. It is to have
a consistent redux state in which both connection and conference are
failed. It could happen that in a buggy environment the XMPP connection
is dropped, but the media is still flowing which would result in weird
user experience.

* Adds an option to disable features based on token data.

Reverts changes from b84e910086, removes disableDesktopSharing option and an interface_config option.

* Disable recording button based on token features data.

Hide recording if local participant isGuest and roles based on token.
When enableUserRolesBasedOnToken is enabled we were not hiding the record button for guests.

* Adds filtering of jibri iqs and rayo based on features.

Moves feature checking in separate utility function.
Renames utility method.

* Adds a footer text when outbound-call is not feature enabled.

* Fixes comments.

Try to use the getCurrentConference function wherever the indention is
to check for the current conference.

This includes recording agents, for example.

Simplify parts of the logic introduced in
11b7144ad0.

Specificaly, using all the state change avoiding functions doesn't give us much
since we need to copy the state for sure.

Makes sure that whenever a conference is left or switched, the local
participant's id will be equal to the default value.

The problem fixed by this commit is a situation where the local
participant may end up sharing the same ID with it's "ghost" when
rejoining a disconnected conference. The most important and easiest to
hit case is when the conference is left after the CONFERENCE_FAILED
event.

Another rare and harder to encounter in the real world issue is
where CONFERENCE_LEFT may come with the delay due to it's asynchronous
nature. The step by step scenario is as follows: trying to leave a
conference, but the network is not doing well, so it takes time,
requests are timing out. After getting back to the welcome page the
the CONFERENCE_LEFT has not arrived yet. The same conference is joined
again and the load config may timeout, but it will be read from the
cache. Now the network gets better and conference is joining which
results in our ghost participant added to the redux state. At this point
there's the root issue: two participants with the same id, because the
local one was neither cleared nor set to the new one yet
(PARTICIPANT_JOINED come, before CONFERENCE_JOINED where we adjust the
id). Then comes CONFERENCE_JOINED and we try to update our local id.
We're updating the ID of both ghost and local participant. It could be
also that the delayed CONFERENCE_LEFT comes for the old conference, but
it's too late and it would update the id for both participants.

The approach here reasons that the ID of the local participant
may be reset as soon as the local participant and, respectively, her ID
is no longer involved in a recoverable JitsiConference of interest to
the user and, consequently, the app.

Co-authored-by: Pawel Domas <pawel.domas@jitsi.org>
Co-authored-by: Lyubo Marinov <lmarinov@atlassian.com>

Pawel Domas said:
The conference state field is referring to the current conference in
progress, so it feels like this field should be cleared as soon as we
declare that the conference is being left and the asynchronous process
of leaving the conference starts (which happens on
CONFERENCE_WILL_LEAVE).

Lyubo Marinov said:
Merely setting conference to undefined is incomplete because there are
other redux state properties associated with conference.

What we're doing here is redefining the redux action
CONFERENCE_WILL_LEAVE:

The redux action CONFERENCE_WILL_LEAVE represents the order of the user
to leave a JitsiConference instance. From the user's perspective, there's
no going back (with respect to the instance itself). The app will perform
due clean-up like leaving the associated room, but the instance is no
longer the focus of the attention of the user and, consequently, the app.

Now CONFERENCE_LEFT and CONFERENCE_WILL_LEAVE have more in common than
before and we can do a much better job by expressing the common parts in
source code. Additionally, our source code can be rewritten to better
express the meaning of any of these two actions.

Co-authored-by: Pawel Domas <pawel.domas@jitsi.org>
Co-authored-by: Lyubo Marinov <lmarinov@atlassian.com>

The commit message of "Associate remote participant w/ JitsiConference
(_JOINED)" explains the motivation for this commit.

Practically, _JOINED and _LEFT combined with "Remove remote participants
who are no longer of interest" should alleviate the problem with
multiplying remote participants to an acceptable level of annoyance.

Technically though, a remote participant cannot be identified by an ID
only. The ID is (somewhat) "unique" in the context of a single
JitsiConference instance. So in order to not have to scratch our heads
over an obscure corner, racing case, it's better to always identify
remote participants by the pair id-conference. Unfortunately, that's a
bit of a high order given the existing source code. So I've implemented
the cases which are the easiest so that new source code written with
participantUpdated is more likely to identify a remote participant with
the pair id-conference.

Additionally, the commit "Reduce direct read access to the
features/base/participants redux state" brings more control back to the
functions of the feature base/participants so that one day we can (if we
choose to) do something like, for example:

If getParticipants is called with a conference, it returns the
participants from features/base/participants who are associated with the
specified conference. If no conference is specified in the function
call, then default to the conference which is the primary focus of the
app at the time of the function call. Added to the above, this should
allow us to further reduce the cases in which we're identifying remote
participants by id only and get us even closer to a more "predictable"
behavior in corner, racing cases.

The commit message of "Associate remote participant w/ JitsiConference
(_JOINED)" explains the motivation for this commit.

The plan set in motion here is to associate remote participants with the
JitsiConference instances that created them in order to be able to
remove remote participants when a JitsiConference is no longer the
primary focus of the jitsi-meet app. And that's supposed to alleviate a
problem with multiplying remote thumbnails.

Doing all of the above in a single commit is a bit of a high order. So
I'm splitting the whole into multiple successive commits for the
purposes of observability, comprehension. Each commit is supposed to be
safe even if subsequent commits are not accepted, are reverted,
whatever. Obviously, without the successive commits, a commit may be
"unused".

One of the important pieces of the multiplying remote thumbnails "fix"
offered is removing remote participants who are no longer of interest
i.e. PARTICIPANT_LEFT. But in order for _LEFT to be implemented, _JOINED
must be implemented first.

Because a conference can fail before or after it's joined it must be
"left" in order to release any allocated resources like peerconnections,
tracks and all the other things.

Zoltan Bettenbuk suggested the following:

         const state = getState();

          if (desiredTypes.length === 0) {
 -            const { audio, video } = state['features/base/media'];
 -
 -            audio.muted || desiredTypes.push(MEDIA_TYPE.AUDIO);
 -            video.muted || desiredTypes.push(MEDIA_TYPE.VIDEO);
 +            const startAudioOnly = getPropertyValue(state, 'startAudioOnly');
 +            const startWithAudioMuted
 +                = getPropertyValue(state, 'startWithAudioMuted');
 +            const startWithVideoMuted
 +                = getPropertyValue(state, 'startWithVideoMuted');
 +
 +            if (!startWithAudioMuted) {
 +                desiredTypes.push(MEDIA_TYPE.AUDIO);
 +            }
 +            if (!startAudioOnly && !startWithVideoMuted) {
 +                desiredTypes.push(MEDIA_TYPE.VIDEO);
 +            }
          }

          const availableTypes

The final commit is really a different implementation of the same idea
but takes into account that the state of base/media already contains the
intent of the URL and notices the delay in the realization of the
background app state.

Additionally, unbreaks one more case where setAudioOnly is incorrectly
dispatched on CONFERENCE_LEFT or CONFERENCE_FAILED and, consequently,
overrides the intent of the URL.