- We need url for individual node as well as all nodes
- We need more extensive HTML output, now it's just nodes which have not
  succeeded in >1d.
- With JSON you can of course do what you want with the data