Sounds like Drummond was equating it with server endpoint that client contacts.

Protocol we have used so far is client-server so possible peer-to-peer is still theoretical. Let's ask Drummond or Markus.

https://en.wikipedia.org/wiki/Communication_endpoint
"A communication endpoint is an interface exposed by a communicating party or by a communication channel. An example of the latter type of a communication endpoint is a publish-subscribe topic or a group in group communication systems.

A communication endpoint is a discoverable node of communication whose scope may be varied to narrow or broaden the discovery zone. Endpoints facilitate a standard programmable layer of abstraction whereby heterogeneous software systems and/or subsystems may communicate with each other and that the means of communication are decoupled from the communicating subsystems."

https://en.wikipedia.org/wiki/Service-oriented_architecture
"Rather than defining an API, SOA defines the interface in terms of protocols and functionality. An endpoint is the entry point for such a SOA implementation."

FWIW, in a spec I recently read, IETF defines an "endpoint" to be an entity participating in a protocol. An endpoint could act as either a server or a client.

To me, an XDI endpoint is the part of an XDI server that receives XDI messages. If we send XDI messages over HTTP, then the XDI endpoint has an HTTP URI. To me, an XDI client is not an XDI endpoint.

Currently we say "XDI messaging is peer-to-peer" - should this be deleted, or how should we explain it since we don't have peer-to-peer now?

In the Messaging definition itself I think I want to use terms as non-suggestive as possible, maybe just the sender and receiver of a message.

In the essay on XDI design goals we want to say a variety of configurations is possible.

ED: Discuss the truth/meaning of the statement: "XDI messaging is peer-to-peer between any XDI client and any XDI endpoint"