improve update interval computation

Don't do anything if a downloaded diff is empty after all
(may be happening when an empty diff was published upstream).

Correctly compute the waiting interval before checking for new
data. As the interval is now computed based on the date of the
newest object in the database, the configured intervals need
to be adjusted slightly to take into account the time it takes
for the server to publish a diff.

utils/osm_file_date.py

@@ -28,6 +28,7 @@ if __name__ == '__main__':
 
     h.apply_file(sys.argv[1])
 
+    if h.filedate is None:
+        exit(5)
+
     print(h.filedate)
-
-