Writing a (at least somewhat usable) compiler has always appealed. Would be tempting to try an implementation of PLOT -- http://users.rcn.com/david-moon/PLOT/

Writing a (at least somewhat usable) compiler has always appealed. Would be tempting to try an implementation of PLOT -- http://users.rcn.com/david-moon/PLOT/ http://users.rcn.com/david-moon/PLOT/ Alternatively, and rather more contact/resource dependent, would be some at-least-semi-lab-based biology.