The switch to zst compressed raw image files broke our retriever workflow as it was not yet configured to deal with these files:
80b8d3a
We all can see that encrypting the files makes total sense as this would cut down on our data transfer size, yet this switch needs to be properly prepared (on both sides) and all inclusive (currently only AWS images where encoded).