Handle janus server being down

[iccanobif]

also check cases like the server going down after a stream has already started and clearStream() fails

[iccanobif]

• add await to all detach() calls and check all the other janus-client functions that might return a promise.
• use my janus-client fork that handles rejected promises properly
• add some sort of semaphore mechanism so that operations that affect janus handles are not performed in parallel, at least for a single user (this might have been the cause for a site-wide stream malfunction that apparently happened in Feb 1 12:16:53 after a user called user-want-to-drop-stream and user-want-to-take-stream very quickly one after the other)