Comment 13 for bug 967036

It turns out this is due to str/unicode python2.x issues. The problem is that we get a utf8 encoded string and uses that as input for %s - then python will try to convert it to a string without a encoding and that fails. I added a explicit convertion back to unicode now. The alternative would be to return a unicode translated string in get_hw_missing_long_description() but that is also not great. Fortunately python3 will safe us at some point :)