Apparently the wizkids at Google haven’t yet figured out how to decode the elusive “HTML document” format… (I encounter this bug quite frequently, in fact)