spell out stream_id and fix todo

api/server-server/user_devices.yaml

@@ -57,10 +57,11 @@ paths:
               stream_id:
                 type: integer
                 description: |-
-                  An ID the requesting homeserver may use to detect changes in the
+                  A unique ID for a given user_id which describes the version of
-                  device list. This should increase as time goes on, and always
+                  the returned device list.  This is matched with the ``stream_id``
-                  produce the same ``devices`` list if not incremented.
+                  field in ``m.device_list_update`` EDUs in order to incrementally
-                example: 334608
+                  update the returned device_list.
+                example: 5
               devices:
                 type: array
                 description: The user's devices. May be empty.

specification/server_server_api.rst

@@ -1025,12 +1025,10 @@ This forms a simple directed acyclic graph of ``m.device_list_update`` EDUs, sho
 which EDUs a server needs to have received in order to apply an update to its local
 copy of the remote user's device list.  If a server receives an EDU which refers to
 a ``prev_id`` it does not recognise, it must resynchronise its list by calling the
-/user/keys/query API and resume the process.
+/user/keys/query API and resume the process.  The response contains a ``stream_id``
+which should be used to correlate with subsequent ``m.device_list_update`` EDUs.
 
-..TODO: how do you synchronise the result of the /query API with the stream_id
+.. TODO: this whole thing desperately feels like it should just be state in a room,
-  in the EDUs?
-
-..TODO: this whole thing desperately feels like it should just be state in a room,
   rather than inventing a whole different DAG.  The same room could be used for
   profiles, presence lists, etc.