How should (scholarly) responses to artefacts be modelled?

[phochste]

This is related to #6 (how should responses to activities be modeled).

In scholarly communication progress can be made by an (open) discussion on topics and response papers. Artefacts in the network are not created ex nihilo or as a new finding (an Announce) or a response to a Service request (an Offer).

When Bob creates a artefact about a new finding, Alice has the possibility to create an new artefact reaction without having received an Offer from Bob "please write a scholarly response on my paper".

Alice will publish this artefact in her Pod and will trigger an Announce. But in what way she can express this is a response on something Bob wrote. Is it an inReplyTo (to an artefact? to an activity?) or an context (to an artefact?). Or should this be modelled in an other way? E.g. the artefact (metadata) itself states in some way that the contents is a scholarly discussion with Bob's artefact as subject.

[mielvds]

Alice will publish this artefact in her Pod and will trigger an Announce. But in what way she can express this is a response on something Bob wrote. Is it an inReplyTo (to an artefact? to an activity?) or an context (to an artefact?).

Definitely not to an activity, because there is no preceeding notification.
I think this is pretty straightforward actually: this is the inReplyTo in the object description. This is also how AS applies it. Example payload (sorry for butchering the actor/origin uris):

{
  "@context": "https://www.w3.org/ns/activitystreams/",
  "summary": "Koninklijke Bibliotheek announces Alba dataset",
  "type": "Announce",
  "actor": "Alice",
  "object": {
    "id": "http://example.org/alice/1",
    "type": "Document",
    "isReplyTo": "http://example.org/bob/1"
    "name": "Response to Bob's paper"  
  },
  "origin": "Orchestrator",
  "target": "Bob",
  "updated": "2021-01-28T19:56:20.114Z"
}

[phochste]

Good that there is a solution. But should we write something about that in the spec document?

The spec is from Mellon viewpoint currently about the Scholarly artefacts and Lifecycle Event Data that records the "fulfilment of the functions of scholarly communication for the researcher’s personal contributions" (from the Mellon proposal).

In this light we have maintainers that create artefacts and can have network communication patterns with Service Hubs that will eventually generate the Lifecycle Event Data.

We don't talk about the other things that is in the Mellon proposal:

• Interaction artefacts (e.g. Alice annotates Bob's artefact)
• Peer contributions (e.g. The world talks about Alice artefact)
• Social interactions (e.g. the Republic of Letters)

This is not part of the patterns that are now listed in Network communication patterns.

Is this something that is currently out of scope (bit will be described in later versions)?

Should we experiment at least what we have now and to see that these use cases don't clash with what we already have?