After a google search I've only found this: https://groups.google.com/forum/m/#!msg/jclouds-dev/hBzkDyT_XiY/mJXEq8l4fP4J
If you read Adrian's comment, it seems the encoding 'tricks' are intentional due to some S3 issue. I also agree with you @diwakergupta that the encoding might be broken, but I'd prefer to have some feedback from @adriancole before merging, so we can understand why current encoding works like and make sure we don't introduce regressions. --- Reply to this email directly or view it on GitHub: https://github.com/jclouds/jclouds/pull/82#issuecomment-21990829
