Outcomes

1. Differing versions of Maple
   The student system (Eden) and my home machine had more recent versions of Maple than the Math Department systems. There were syntactical differences and differences in the user interface. Many students in this course need to be taught that programs will not break if you don't type in the correct syntax, and get easily discouraged if things don't "work" immediately. So eliminating as many initial roadblocks as possible is a good idea. Always check programs and homework on the specific machines and systems students will be using. Don't hope or expect that things will work in the same way without checking for yourself.
2. Do the darn machines work?
   I naively expected after taking a training session that the machines in my "smart" classroom would behave like the ones I learned with. At the beginning of the second semester, the videotape machine wouldn 't "obey" me. I assumed the problem was mine, not the machines. Only weeks later I learned that something was broken. I always examine a classroom before I teach in it, to see how large I need to write, to see how loud I need to speak, to check it all out. Now I know that checking it all out means playing with the machines enough to make sure they behave the way I think they should.
3. A laptop that works and displays
   It would be wonderful to have a laptop working and displaying the appropriate software (Maple, Mathematica, Derive: whatever) in the classroom. I didn't have that but wanted it. Also access to "the web" during the class would be useful. Class webpages could be displayed, as could other web references.
4. Reporting on the course
   I spent much more time writing and constructing these web pages than one might imagine. I probably dedicated the equivalent of two solid weeks of work to this effort. I hope this is too much work.

TOP

Additional crypto topics

I did not discuss shift registers. These are fundamental to much of modern crypto. The text has good information and problems, and the theory of shift registers is complete, unlike, say, the "theory" of block ciphers. The mathematical attractiveness of unbridled theory probably made me very cautious, perhaps too much so, because nice problems could have been given. And shift registers are fast and easy, and do exist in the world. They would fit in nicely if there had been time to discuss more about bitstream generators.

Mr. Radomirovic observed correctly that we spent a great deal of time on the asymmetric public key systems -- they are seductive because they allow a great deal (the prison scenario, digital signatures, anonymous e-cash) that could scarcely be envisioned classically. But (to continue his observation) in fact symmetric systems (usually block ciphers) are much more commonly used. We should have paid more attention to them. I think this is correct, and in a future incarnation of the course would try to create some interesting assignments, in and out of class, as has already been discussed.

I wrote the following in the prospectus for the course.

TOP

Reference material for the course

• I've attended lectures and meetings. I'm fortunate to be at Rutgers where DIMACS holds most of its meetings, and also to be in an area where there are many industrial research labs in telecommunications and finance: many people interested in crypto are around here and travel through here.
• The obvious source for information on both the technical and policy aspects of the course is the web. The web seems, however, fundamentally ephemeral as a scholarly resource at this time. I wanted students to write papers about medical record privacy, crypto policies of governments, and DES during both semesters. When I returned to the web pages of references I constructed for these topics during the fall semester, I discovered that there were links no longer functioning on each of the pages. It is therefore useful to check links yourself. But note that students can contribute to this enterprise. Both semesters in every assignment I was given further links by students, and was able to add them to the list. Students know searching techniques and can be persistent and perhaps lucky!
• Mailing lists and web pages can be very useful to keep up with "the news". I had the luxury of working with N. Fefferman who knows lots about this and did a great deal of winnowing for me. Look at the cypherpunks mailing list; consider the slashdot web sit; keep an eye on the Electronic Freedom Foundation (EFF), the Electronic Privacy Information Center and similar organizations. For advances in the technical side of the subject, sites such asCounterpane and RSA might be useful. And just looking at the seminar lists ...
• I just started looking at traditional newspapers and magazines with some sensitization to the issues of the course. Almost every week (in fact, most days!) I was able to find articles about issues connected with this course. Frequently the articles are in the business sections of these serials because privacy and security are very important issues to business.
• Finding a text has some importance. I could have used the text I selected more aggressively. Instead I created a large amount of material. More books (and more web sites) about crypto are constantly appearing. A more suitable text might be found. Certainly instructors of the course should be familiar with some of the standard "upper-level" texts.
• Guest speakers, preferably working in "the real world", are a very useful source of current information. I was lucky to have some when I taught the course. If I taught the course again, I would invite people like Peter Winkler of Lucent and Rebecca Wright of AT&T. These and the people who spoke in my course are examples of individuals who can give elegant presentations with real technical content, and who can, when properly advised of the level of the audience, teach beautifully to such a class. I think it is important to have outside speakers whose appearance would confirm the importance of the subject matter of the class. Almost any school is likely to have someone managing computer systems who can discuss "hacking" and password validation. Or someone who can discuss what is done to keep student records (and payroll information!) secure from malicious folks.

TOP

Pedagogy

My attempt to run a "chat room" consisted of having a web page of questions and comments. It isn't clear to me how helpful this was. Certainly if the class were larger, this would have helped me decrease the amount of e-mail I answered. In a typical week, I think I received about 1 e-mail message per student in this course. This is an average: a few students never sent me e-mail outside of mandated homework, and some were far more "verbose" (e-verbose?). I know that students e-mailed each other quite a bit, and I certainly encouraged them to work together on several assignments. I also said they could work together on other assignments provided they indicated that the work was a joint enterprise. Again, some students were "lone wolves" and only worked with others when forced to. Other students were rather derelict in joint projects, and several times students spoke to me about how to compensate for uncooperative teammates. My lack of experience showed and I was not as helpful as more informed instructors might have been.

TOP

Attendance

This was all seemed correct until I contrasted the attendance in the course with that of students in calculus courses. In most of our calculus courses, attendance after the middle of the semester would tend to be 30% to 50% of nominal enrollment. In this course, attendance usually seemed to be about 80% to 90%.

I had told people at the beginning of the semester that attendance was important. I could have used 1 point quizzes to reward their attendance. It isn't clear what would have changed but at least I would be doing "something" and would have, in addition, a nice way to gauge how each class meeting went.

Again, correct student behavior should be rewarded. And stimulating behavior that is desired can be done by "rewards". I don't think students participated enough in the presentations. I think I should have made clear that I was somehow "scoring" participation. I think I also should have said that a question on the final exam would reflect some aspect of student presentations, and then followed though. That almost surely would have encouraged more attention be paid to presentations and the ensuing debates.

TOP

Scaling up?

TOP

A few orphan problems

TOP

Outcomes