LOL. I just noticed that fsdump.py computes an MD5 digest for each data record: dig = md5.new(rec.data).hexdigest() It doesn't use the result, `dig` is simply ignored. Maybe fsdump was running too fast <wink>.