The term "course notes", as used on materials bound and sold to students at the UW Bookstore, is usually a misnomer; for most CS courses, they aren't notes, but copies of the transparencies used in lectures. I am ambivalent about their use, and even more ambivalent about handing them out to students.
It's not that I want to withhold information from students, or somehow control how it is metered out. My worry is that when the transparencies or presentations are made available, students become more passive. I don't understand how people learn -- I don't think anyone really does -- but I think that the act of writing uses a different part of the brain than the acts of reading, watching, and listening, and that writing somehow helps in fixing information in long-term memory. I don't have any hard data to back that up, except that's the way it works for me.
Having copies of the transparencies means that students don't have to scribble madly to get information down. This, in theory, frees them to write additional notes -- the real course notes -- on what the instructor is saying, or on insights they have gleaned from the presentation. I don't often see such additional notes when students ask me questions just after class, and in sitting at the back of classes observing other instructors, I don't see a lot of writing going on. I worry about the "notes" giving students a false sense of security.
Some course notes are really extensive written materials -- this is the case for CS 241, which doesn't have a textbook. CS 251 has a great textbook, but I have had students complain that the notes were "incomplete", as if they were supposed to be something more than an exact version of the slides.
When I first taught CS 251, Steve Mann had prepared a set of transparencies for himself that I could also use. Even better, he had designed a system which created slides, notes for the instructor to use in class, and "notes" for students (basically the transparencies compressed to remove whitespace) from a single source text. While my previous experience with transparencies prepared by other people was that they were frustrating and confining, these were fine to use.
They also acted as a proper definition for the course. In recent years more and more courses have definitions of this sort. When I first started working at UW, most courses had no definition beyond the one-page description in the undergraduate handbook. Instructors chose their favourite text, put emphasis on their own favourite material, and the contents of offerings could vary considerably from term to term. That still happens in some cases, but more often instructors are cooperating to maintain a central set of "notes".
Each course has a course coordinator who is supposed to make sure the course does not drift, but some coordinators are more active than others. Steve, who coordinates 251, has been very good at making sure changes are gradual and sensible, even in terms when he is not teaching the course.
Knowing all this, when I revised the course in 2000, I tried to create a set of transparencies with a similar style and spirit. I did the first draft in the summer of 2000; Steve read it and offered comments, but we both knew the real test would come in the first offering, in the fall of 2000. There were a lot of small glitches, and some discoveries about what worked and what didn't. We created some material on the fly as we realized that certain topics needed more coverage, and added sections to the notes after the fact. But overall it worked nicely.
I adopted Steve's idea when I designed CS 135 in 2004, except that I built my own system instead of using his, so I didn't have to keep asking him how to make changes. --PR
(Adapted from a blog posting originally written July 14, 2003, in Paris.)