* Implement the current Free Haven design.

* Figuring out the designs for coupling a Free Haven backend onto
  something like Freenet or Mojo Nation, to provide persistent storage
  for some items.

* Implement a generic mix-net simulator.

* Integrating strong anonymity into Freenet or Mojo Nation design, and
  make sure it's done right.

* Implement attacks of systems such as Freenet, to point out
  limitations from accountability/anonymity perspective.  Are the
  attacks we suggest actually feasible?  How easy are they?  How
  sufficient is this whole caching idea?  Can we get some upper bounds
  for traffic analysis?

* Do something with reputation systems for Ibal project at Harvard.
	-NOTE: don't put this one in until after I run it by the Ibal people.
	-dmolnar

* Further design/implementation for distributed append-only ledger as
  described in mix-acc paper.

* Work on design/analysis/implementation of Red Rover.  It sounds like
  a good idea for grass-roots end-users, but may currently lack
  technically-savvy designers.