Since this is a serial communication specific pin, shouldn't this pin state be specified in the core implementation instead?

In my opinion the Harp protocol (which is just one part of the Harp Standard) should be as much agnostic as possible to the communication protocols.

This spec doesn't need to necessarily go into the protocol, I would probably but it in the Device. The same way that the baudrate of the synchronization protocol is not defined in its own document.

Feedback from SRM meeting:

• What does "connected" mean in the protocol specification?
  • @filcarv clarified the intention of "connected" as the host application opening a serial connection to the device.
  • DTR might open the opportunity for more devices to participate in the current client implementation
  • @glopesdev and @artursilva0 suggested that we leave the communication protocol open to any physical transport layer, but just specify that the transport layer needs to provide a way for devices to define "connected" and "not connected" states. In practice we can set DTR to true in the current C# client

This has been added to the new Bonsai version. I will make the draft into a PR to be reviewed.