A lot more diverse than that at my university. The professors basically had a four by four grid: strongly typed and object oriented (Java, covered in the introd

A lot more diverse than that at my university. The professors basically had a four by four grid: strongly typed and object oriented (Java, covered in the introductory programming classes), strongly typed and functional (Standard ML or Haskell, depending on what quarter you sit the class), weakly typed and functional (Scheme), and weakly typed yet object oriented (Ruby). Once recursion was sufficiently beaten into you, it was not particularly new material. The Stanford curriculum here seems a bit better designed: most people will not implement a language of their own or use more than one or two languages to any great extent (and a fairly small percentage will actually touch Standard ML or Scheme), and this class is better targeted to those two: it lets them see what the ideas were behind the language and how they can choose the right tool for a lot more jobs. In other words, the Stanford course is a bit better at showing the range of ideas and higher level languages to people who were up to that point Blub programmers (see http://www.paulgraham.com/avg.html http://www.paulgraham.com/avg.html , of course).