It would be neat if you could extract the desired data (text) from the page and hash it so that you could later fix the link (by pointing to a new valid link).