In the following I will explain an overview how data is saved into Amazon S3 when using this program. You can examine it by using a simple S3 browsing application like JetS3t Cockpit to view and download data that was previously uploaded using JaBaS3.
When starting a backup, JaBaS3 first lists all files in the given "directory" on Amazon S3 to see which files and versions of these files are already there. Then it scans through the files you have chosen to backup to see which have changed. These are uploaded. When there are too many old versions for a single file, they are removed until the number of files remains, that are configured in the program's settings.
The filenames of the files you want to backup are used as keynames for Amazon S3. In a future version you will probably be able to use encrypted filenames, but for now only the content of the files is encrypted.
Each keyname is appended the information of the file's date, size and md5 hash. This allows to quickly see which files were changed locally since the last backup.
Files that were locally removed are not yet removed from Amazon S3, they will simply remain, but can be deleted using the "Delete" button in the restore window.
For details about the used encryption, please see the encryption page.