MSC3816: Clarify Thread Participation (#3816)

* Clarify the current_user_participated flag from MSC3440.

* Add a better link to the definition.

* Clarifications from review.

Co-authored-by: Richard van der Hoff <1389908+richvdh@users.noreply.github.com>

Co-authored-by: Richard van der Hoff <1389908+richvdh@users.noreply.github.com>

proposals/3440-threading-via-relations.md

@@ -57,8 +57,10 @@ would include additional information in the `unsigned` field:
   The latest event should be serialised in the same form as the event itself;
   this includes adding any bundled aggregations for the event (and applying edits).[^1]
 * `count`: An integer counting the number of `m.thread` events
-* `current_user_participated`: A flag set to `true` if the current logged in user
-  has participated in the thread
+* `current_user_participated`: A boolean flag, which is set to `true` if the
+  current logged in user has participated in the thread. The user has participated if:
+  * They created the current event.
+  * They created an event with a `m.thread` relation targeting the current event.
 
 #### Rich replies in a thread
 

proposals/3816-clarify-thread-participation.md

@@ -0,0 +1,73 @@
+# MSC3816: Clarify Thread Participation
+
+[MSC3440](https://github.com/matrix-org/matrix-doc/pull/3440) defines the `m.thread` relation
+type, and the format of the serverside aggregation for them. The definition of the aggregation includes a
+`current_user_participated` flag, which is not fully defined:
+
+> A flag set to `true` if the current logged in user has participated in the thread
+
+In particular, it is unclear whether sending the initial event (i.e., the event which is the
+target of the `m.thread` relation) counts as participating in the thread.
+
+Known implementations do *not* count the initial event in this way, and instead
+implement this as: "has the current user sent an event with an `m.thread` relation
+targeting the event", but this has found to give poor user experience in practice.
+
+For example, consider `A` as the root event in a thread from `@alice:example.com`, and `B`
+as a threaded reply from `@bob:example.com`. The bundled aggregations for `A`
+would include:
+
+| Requester            | `current_user_participated` |
+|----------------------|-----------------------------|
+| `@alice:example.com` | `false`                     |
+| `@bob:example.com`   | `true`                      |
+
+If `@alice:example.com` sends reply `C`, this would change:
+
+| Requester            | `current_user_participated` |
+|----------------------|-----------------------------|
+| `@alice:example.com` | `true`                      |
+| `@bob:example.com`   | `true`                      |
+
+The proposed clarification is that `@alice:example.com` should have always have
+participated in the thread (i.e. both tables would be `true` in the example above).
+
+## Proposal
+
+The [definition of the `current_user_participated` flag](https://github.com/matrix-org/matrix-spec-proposals/blob/main/proposals/3440-threading-via-relations.md#event-format)
+from the bundled aggregations for `m.thread` relations is updated:
+
+> A boolean flag, which is set to `true` if the current logged in user has
+> participated in the thread. The user has participated if:
+>
+> * They created the current event.
+> * They created an event with a `m.thread` relation targeting the current event.
+
+This better matches the intention of this flag, which is that a client is able to
+visually separate threads which might be of interest.
+
+## Potential issues
+
+The current implementations will need to be updated to take into account the
+sender of the current event when generating bundled aggregations. This should be
+trivial since all of the needed information is directly available.
+
+MSC3440 proposes using [new `filter` parameters](https://github.com/matrix-org/matrix-spec-proposals/blob/main/proposals/3440-threading-via-relations.md#fetch-all-threads-in-a-room)
+in order to list threads in a room that a user has participated in. There would
+now be an inconsistency that threads where the current user sent the root event
+but has not replied to the thread could not easily be fetched. A future MSC may
+solve this problem.
+
+## Alternatives
+
+Do not clarify [MSC3440](https://github.com/matrix-org/matrix-spec-proposals/pull/3440)
+and leave it up to implementations to define the behavior of the
+`current_user_participated` flag.
+
+## Security considerations
+
+None
+
+## Unstable prefix
+
+None, the changes above shouldn't dramatically change behavior for clients.