Now that a file can only be in one dataset at a time it might be helpful for queries to have the dataset id on the file model as well