I had this tool saved as a bookmark, and didn't find any existing HN discussion, so I decided to post myself.
At first glance this seems to offer similar functionality to PAR2, with the disadvantage that errors cannot be corrected, as parity data is not stored in the digest.