Initial draft hack for typing notification docs

drafts/typing_notifications.rst

@@ -0,0 +1,55 @@
+Typing Notifications
+====================
+
+Client APIs
+-----------
+
+To set "I am typing for the next N msec"::
+  PUT .../rooms/:room_id/typing/:user_id
+  Content:  { "typing": true, "timeout": N }
+    # timeout is in msec; I suggest no more than 20 or 30 seconds
+
+This should be re-sent by the client to continue informing the server the user
+is still typing; I suggest a safety margin of 5 seconds before the expected
+timeout runs out. Just keep declaring a new timeout, it will replace the old
+one.
+
+To set "I am no longer typing"::
+  PUT ../rooms/:room_id/typing/:user_id
+  Content: { "typing": false }
+
+Client Events
+-------------
+
+All room members will receive an event on the event stream::
+
+  {
+    "type": "m.typing",
+    "room_id": "!room-id-here:matrix.org",
+    "typing": ["list of", "every user", "id who", "is currently", "typing"]
+  }
+
+The client must use this list to *REPLACE* its knowledge of every user who is
+currently typing. The reason for this is that the server DOES NOT remember
+users who are not currently typing, as that list gets big quickly. The client
+should mark as not typing, any user ID who is not in that list.
+
+Server APIs
+-----------
+
+Servers will emit EDUs in the following form::
+
+  {
+    "type": "m.typing",
+    "content": {
+      "room_id": "!room-id-here:matrix.org",
+      "user_id": "@user-id-here:matrix.org",
+      "typing": true/false,
+    }
+  }
+
+Server EDUs don't (currently) contain timing information; it is up to
+originating HSes to ensure they eventually send "stop" notifications.
+
+((This will eventually need addressing, as part of the wider typing/presence
+timer addition work))