content removal is supposed to succed if the content was already not present

P2P/Annex.hs

@@ -91,10 +91,14 @@ runLocal runmode runner a = case a of
 			Left e -> return (Left (show e))
 			Right result -> runner (next result)
 	RemoveContent k next -> do
-		v <- tryNonAsync $ lockContentForRemoval k $ \contentlock -> do
-				removeAnnex contentlock
-				logStatus k InfoMissing
-				return True
+		v <- tryNonAsync $
+			ifM (Annex.Content.inAnnex key)
+				( lockContentForRemoval k $ \contentlock -> do
+					removeAnnex contentlock
+					logStatus k InfoMissing
+					return True
+				, return True
+				)
 		case v of
 			Left e -> return (Left (show e))
 			Right result -> runner (next result)

doc/todo/tor.mdwn

@@ -12,9 +12,6 @@ Current todo list:
   object is already in progress, the message about this is output by the
   remotedaemon --debug, but not forwarded to the peer, which shows
   "Connection reset by peer"
-* git annex drop --from peer1, when the peer does not have a file,
-  shows "<socket: 18>: hGetLine: end of file"; the peer has closed the
-  connection.
 * Think about locking some more. What happens if the connection to the peer
   is dropped while we think we're locking content there from being dropped?