Fixes #1

signal emission apparently does not work in _finalize()

Fixes some critical warnings from remote-endpoint.c because disconnection
from the destroyed endpoints did not happen

This avoids being able to find the same object on the globals list
from code that is connected to the signal.

avoids criticals and crashes when the core is destroyed earlier

+ add a method to retrieve the core
+ change the register function to not require the core as a parameter

This is a cleaner way to interface with the remote pipewire daemon.
The WpRemote base class can be subclassed also for interfacing
with other daemons (hardware-specific managers, etc)

This can be used by code that depends on policy to re-evaluate
choices they have made in the past that could now change based
on the new policy.

The core may be the last object holding a reference to the endpoint,
so the unregister call may destroy it.

This allows loading of config and modules when deployed in
the target system without setting environment variables.

and also implement
* global-added/removed signals
* a foreach function to iterate through globals
* registering modules and factories with the same key

g_weak_ref_clear destroys the weak ref, so it will crash
when _finalize() tries to destroy it again

This changes the registration point to be the endpoint and not the
session manager, as it seems easier to implement (the implementation
will not need to keep a pointer to the session manager around)

* Make streams a GVariant array, for future-proofness
* Add API for controls (volume, mute, brightness, contrast, etc...)
* Remove API for profiles (it's not well-thought; may be re-added
  in the future)

After discussing things at the AGL May 2019 F2F meeting
and reflecting on the initial design of WirePlumber,
it became clear that it needed a fresh start.