This behaviour is deprecated and will be removed in future versions
of Python. Technically, it could be considered a bug (invalid usage
causing different behaviour from the expected one), and in practice
it should not break much code (because .get_event_loop() would likely
be the same event loop anyway).
This will help unifying the retry loop and reconnecting
if the server disconnects us during auth key generation
which will be done in a follow-up commit.
Which leads to "RuntimeError: readexactly() called while another
coroutine is already waiting for incoming data" being raised,
and causing everything to break or halt.
This way it is easy and cheap to copy the two required values
to all incoming updates in case we need to getDifference since
the previous pts/date to fetch entities.
This is still a work in progress.
It's the only way to properly clean all background tasks,
which the library makes heavy use for in MTProto/Connection
send and receive loops.
Some parts of the code even relied on the fact that it was
asynchronous (it used to return a future so you could await
it and not be breaking).
It's automatically syncified to reduce the damage of being
a breaking change.
* Create `_NOT_A_REQUEST` when needed. Currently, modifications
in the raised exception would be "global".
* `retries` parameters were actually attempts. This has been fixed
to actually be the amount of retries, so 0 now means don't retry.
* Helper function to deal with retries instead of using a range with
different styles every time.
The idea behind distinguishing between the user being connected and
the actual connection having been made was to support retries while
the user decided to connect, even if the connection wasn't made yet.
The problem is that _user_connected is used directly to tell whether
the sender is connected or not which will be wrong if an exception
occurs on the initial connection. Since the logic retry isn't used
there we can simply mark as connected once a successfull connection
is made.
This prevents us from locking forever on any task that doesn't
rely on cancellation tokens, in this case, Connection.recv()'s
_recv_queue.get() would never complete after the server closed
the connection.
Additionally, working with cancellation tokens in asyncio is
somewhat annoying to do.
Last but not least removing the Connection._disconnected future
avoids the need to use its state (if an exception was set it
should be retrieved) to prevent asyncio from complaining, which
it was before.
