If ICE config object gets modified while there's any native
PeerConnection instance holding a reference, it will cause
an exception about final object being modified.

It happens when trying to reconnect XMPP Websocket on RN and
is caused by the fact that new TURN config is fetched after
the connection is re-established.

If user experiences any failed reconnect during the entire
session, even if followed by a success, then
the "connectionFailed" flag will trigger
JitsiConnectionEvents.CONNECTION_FAILED on hang up while
processing the disconnected transition.

Get rid of the experimental flag for simulcast support, update js-utils for detecting Edge on Android

Trigger renegotiation only for plan-b when tracks are muted or unmuted

The 'searchParams' are not implemented on react-native.

The LipSyncHack is currently broken in Jicofo and
will lead to unstable behavior.

Switch to using ontrack for remote tracks, re-use mids/transceivers when remote sources are removed and added.

Prevents from calling disconnect on the Strophe
connection if it already has been called or
not connected.

When modifying the room protection (setting or unsetting a password), the muc#roomconfig_passwordprotectedroom should always be defined. This fixes issue #512.

Co-Authored-By: Saúl Ibarra Corretgé <s@saghul.net>

Trigger renegotation only when negotiationneeded event is fired on the browser.
Do not disable simulcast for screensharing in unified plan.
Fix the order of the simulcast streams for Firefox

Use unique session id per TPC.
Translate a=rid to Jingle only when simulcast through RIDs, i.e., ssrc-less simulcast is supported

Send conference.join event only after the peer connection is created.
This way we won't send events for lonely calls.

Copy/paste mistake - there's no such thing as this.connection in
the XmppConnection. Hard to track down because no error is logged.

Either connect or attach can be used to initiate the connection.

Use the rest argument expression to not miss any.

When Jiconop is used to attach to a warmed up BOSH session Strophe goes
to ATTACHED instead of the CONNECTED state.

The LastSuccessTracker has direct access to the Strophe.Connection.

The XMPP stream resumption will enable automatically if supported by
Prosody on Websocket connections. Pass the 'enableWebsocketResume'
config option set to 'false' to disable.

Websocket keep alive is HTTP GET request sent on the service URL passed
to Strophe. It is used to keep the stick tables alive. The feature comes
in enabled by default whenever Websocket are used. Pass
'websocketKeepAlive' config option set to -1 to disable.

See the code for more details about the default interval value.

If Strophe goes to the ERROR state it will pass the error message there
and then transition to the DISCONNECTED state, but without any error
msg.

Given that the XMPP connection is able to reconnect after a drop, now
the jingle session needs to hold off any updates that have to be sent
while the signalling is down. Send the update upon reconnect to sync
Jicofo and JVB with the client's source description.

The connected flag is flipped to true as soon as transport connection is
established by Strophe(such as WebSocket TCP), but it doesn't mean that
the connection is ready to send stanzas yet.

All calls to connection.send should go through the wrapper in order to
prevent breaking the connection by calling it in the not connected
state.

This prevents from accessing Strophe internals directly and allows to
block 'send' calls while disconnected. The latter breaks the stream
resume logic by putting stanzas onto the send queue.

TODO make jitsi strophe plugins use the XmppConnection wrapper whenever
possible instead of interacting with the Strophe.Connection directly.

Websocket sends queued presence 'unavailable' and the stream close
message correctly when the connection is being closed.

This way it's easier to figure out which parts of the disconnect method are about
cleaning up the connection and the reasons on why it's important can be put in
the JDoc of this method.

...it make it easier to process the logs in automated way.

It can either be a websocket or bosh URL, so the current naming is
confusing.