connect helper function does not allow definition of a retry policy

[malachantrio]

When a Client is created using the helper function connect the retry policy created for the client connector is a default policy with a with retries_per_node set to 0. This means that by default connections will attempt to re connect with a single endpoint forever. In the situation where some eventstore node in a cloud like setup vanishes (e.g. has been terminated either on purpose or by the cloud provider) this mean that the application would never recover it's connection. I would suggest one of two solutions

1. The default is set to something that will allow it to try other nodes in the cluster
   https://github.com/madedotcom/photon-pump/blob/master/photonpump/connection.py#L78

class Connector:
    def __init__(
        self,
        discovery,
        dispatcher,
        retry_policy=None,
        ctrl_queue=None,
        connect_timeout=5,
        name=None,
        loop=None,
    ):
        self.name = name
...
        self.retry_policy = retry_policy or DiscoveryRetryPolicy(retries_per_node=0)

2. We allow the retry policy to be passed in to connect