Michael Godfrey

I am leaving Cornell. As of fall 1998, I will be an assistant professor at the University of Waterloo. My new home page is here. The remainder of this page is likely out of date.

Professional Info:

• I'm an assistant professor at Cornell University, Department of Computer Science.
  My office is Upson 308, and my office phone number is (607) 255-1051.
  My current office hours are Tues and Thurs 3:00-4:00 pm.
• I am associate director of the Computer Science Masters of Engineering Program.
• During the spring term of 1997-98, I taught CS211.
• During the fall term of 1997-98, I taught CS501 -- Programming Languages and Software Engineering. I also taught sections for CS100A.

Research and Teaching Interests:

My research interests include software engineering, configuration management, formal methods, and programming languages (but my interests in programming languages have little to do with lambda calculus or denotational semantics). In a nutshell, I am interested in exploring abstractions that may be derived automatically from source code. For a more detailed view, you can view my statement of research interests here. Some papers of mine can be found here.

In the spring of 1997, I gave a talk entitled On the Nature of Software Design as part of a Ph.D. professional seminar series at Cornell. Scribe notes from the talk can be found here and the slides from the talk (4-up Postscript) can be found here.

I have a healthy interest in (and strong opinions about) computer science education, particularly software engineering and CS1 courses such as CS100. I have taught CS1 and CS2 courses for several years, and been involved in courses that used C, C++, Java, and Object-Oriented Turing (see below). While obviosuly, there is a lot more to teaching programming than just picking the ``right'' language, my humble opinion is that pedigogically speaking, that list goes from worst to best. But then, I am biased. For the record,

• I think Java is a fabulous second language. It's also a pretty good first language.
• I have great trouble with the argument that beginning CS majors ought to learn C (or C++) first so they can get a summer job. My view is that first-year CS students should not be programming in C in the ``real world'' before they have a good handle on basic programming skills.
• While it is probably inevitable that an ``objects-first'' approach will become the way to teach CS1 in the near future, I think it's a more of a coping technique than a way of teaching real understanding and insight.
• Many CS1/CS2 instructors do not fully appreciate how hard it is to for their students to program with buggy tools, unchecked array bounds, and various moving targets. Part of dealing with complexity is minimizing distractions.
• You shouldn't teach recursion in much detail before your students know what a stack is (not everyone shares this point of view).
• CS undergraduates do need to learn program in C eventually or they will never develop an appropriate level of intuition/paranoia about what can go wrong and how. C is not just another programming language; its proper use requires a profound understanding of programming at a lower level of abstraction than most highl-level languages. Likely this issue will become less important over time, but it's still important now.

Academic Background:

My graduate supervisor was Prof. Ric Holt who got his Ph.D. from the Cornell Department of Computer Science around the beginning of time (in Unix terms). His thesis formalized the notion of deadlock; see any text on operating systems for details. He is also responsible for the Object-Oriented Turing programming language, which is used widely to teach introductory programming. If you think Java is a clean, well-designed language, you should have a close look at OOT.

Personal Info:

• Visiting Toronto? Here is a page of my favourite things to do.
• Je suis canadien errant newly arrived in Ithaca as of August 1996. I still say 'zed' and 'to-matt-owe', eh. For fun, I've started a page of coping tips for Canadians who live in the US here (contributions welcome!).
• My wife's name is Anita Buttemer. She teaches Suzuki violin and viola. There's a photo of her here.
• We have a vigorous toddler, named Trevor Cameron Buttemer Godfrey. He has red hair, unlike either of his parents, but like his grandfather and two uncles. I have a bunch of photos of him here.
• Here is a work in progress.
• In private life, I am a hockey maniac, folk music afficionado, amateur food snob, and connoisseur of fine British comedy.