The Case Against Data Lock-in

   Abstract