Yeah, using the GUID works, but then it opens us up to someone coming in and just scraping every file we have on S3 which would be bad for us. I'm leaning towards encrypting/obfuscating the ids... any suggestions for python libraries suited to this task?