> -----Original Message----- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On > Behalf Of Arthur > When students have something substantial enough, and intricate enough, > where > unit testing is a real solution to a real problem - yes. > > But that's probably a few years in.
Just as I believe students looking to construct a Triangle should *not* be thinking about API issues. You are making them make-believe, solve problems *they* don't have. Let them solve the problems they *do* have. Like understanding a triangle, units of measure, hundreds of things, before they are worried about exposing their Triangle's API. I feel strongly about this in part from undertaking the study of mathematics as well as from undertaking the study of programming. Learning mathematics from a mathematician/teacher who starts from the premise that each mathematical idea must be rigorously stated, is - in my experience - next to impossible. I am not yet up to appreciating the abstraction behind that rigorous statement of things, and unless someone comes down to talk to *me*, I never will. Luckily for me, enough of the greater mathematicians understood who they were, and did not feel obliged to communicate in ways that other of their colleagues apparently felt themselves to be professionally bound. Art _______________________________________________ Edu-sig mailing list [email protected] http://mail.python.org/mailman/listinfo/edu-sig
