Fixed an issue where the IDLE_TIMEOUT was triggered even though one had been active in the interval.

evennia/server/portal/portalsessionhandler.py

@@ -382,9 +382,9 @@ class PortalSessionHandler(SessionHandler):
         #text = timetrace(text, "portalsessionhandler.data_in")
 
         if session:
+            now = time()
             if self.command_counter > _MAX_COMMAND_RATE:
                 # data throttle (anti DoS measure)
-                now = time()
                 dT = now - self.command_counter_reset
                 self.command_counter = 0
                 self.command_counter_reset = now
@@ -396,6 +396,7 @@ class PortalSessionHandler(SessionHandler):
                 return
             # relay data to Server
             self.command_counter += 1
+            session.cmd_last = now
             self.portal.amp_protocol.send_MsgPortal2Server(session,
                                                        text=text,
                                                        **kwargs)