Maybe taking a step back here...could someone kindly recap (here or in the ticket) why we're suggesting a rather speculative "let's try a few times" approach. Is there no way to more definitively identify whether we need to retry or not?
Or is there simply no way to separate "rectifiable 401s" (which should trigger a retry) from "permanent 401s" (which should be allowed to fail? --- Reply to this email directly or view it on GitHub: https://github.com/jclouds/jclouds/pull/104#issuecomment-22303742
