Maybe the test should simply disabled (or at least have it's failure made non-fatal), a test without reproducable results is of rather limited value.
-- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected]

