Why would a webpage reader even have an "indexdb"?