Back around March 21 2004, there was a long and interesting discussion on the SIGCSE list, reflecting the powerful promise and frustrations of using Java and OOP in CSI (see http://listserv.acm.org/archives/sigcse-members.html and also Bruce Kim's article this issue). For this column, I would like to use the history of the Nifty Assignments as a way to look at this complex situation.
/lp/association-for-computing-machinery/niftiness-4QiyPDEhnk