Is republishing the data "as data" a "fair" thing to do?