The Scheming senior

Life has been busy once more, since school started. For some reason, I thought it would be a good idea to take three CS classes (software engineering, programming languages, and cryptography) as well as music theory. Music theory appears to have homework due every class, which sucks, even though it's pretty easy so far. In addition, being a senior now, I have to start worrying about what direction I want my life to go in next year. Of course, being me, I'm still rather indecisive over graduate school versus industry.

Research, to be honest, is so far basically what I made it out to be—boring. It didn't help that I picked a topic that I wasn't particularly interested in, but I'm doing my best to stay interested. Professor Mathur seems to be very passionate about what he does, which helps my motivation a bit. Even though it's not that exciting (calling it flat out boring is also an exaggeration), I guess this is a good experience for me to actually have a mentor that's been in the field for years. Hopefully I can get the motivation to finish this research project and publish a paper at the end of the semester, which would look nice on a resume for grad school. Weekly meetings should boost my motivation, at least a little bit.

Classes are kind of "meh" right now. Software engineering is definitely a lot less exciting than I thought it would be, and kind of wish I had taken something else instead. The project is some sort of grunt work from HP that has to do with finance. It definitely sucks compared to the project binder that we saw, which was a game of some sorts. From what the programming languages book is saying, the class will get more interesting as time goes on, but the current chapters are boring. Cryptography is pretty interesting; it's probably the most interesting thing right now. Music theory hasn't gotten past playing scales, so it's not very exciting. Playing on the keyboards on Friday is kind of fun despite the simplicity, since I haven't played piano in about seven years now.

Perhaps the most interesting thing that has happened lately (at least to me) is the arrival of my latest shipment from Amazon, consisting of:

• The Little Schemer, 4th ed.
• Serenity, Collectors Edition
• The Design of Sites

The Design of Sites is more to get some inspiration for web design, since I clearly suck at it. I guess this is necessary, since the web is the wave of the future. Or whatever. It seems to be very illustrated, but I haven't really looked too carefully at it, yet.

Serenity came out recently, which I couldn't resist buying, even though I already own the original edition. Firefly is one of my favorite series, and extended scenes and more bonus features makes an already awesome science fiction universe even more legen—wait for it—

Last, but certainly not least, I've already gone through two chapters of The Little Schemer. The book takes on a very peculiar format, based purely (at least so far) in Q&A format. They ask questions that you can hopefully answer and give explanations for each answer. For a fairly advanced programmer like me, this is amusing at first, but tiring in some spots. I was hoping for a book that I could read cover to cover, but this doesn't seem to be the book. Not that I'm disappointed with it at all, I just don't feel like reading through levels of recursion that are obvious to me.

This brings about an interesting pedagogical method; Friedman and Felleisen take an approach that depends on the human ability to recognize patterns. It seems like if you ask enough questions, the reader should be able to figure out what a certain function does. I had kind of an advantage of already knowing what some of them did, so playing a game of 20 questions was a bit overkill, but I wonder how this teaching method would apply to a more inexperienced student.

Actually, I already use this method somewhat with students. Whenever a student asks me a question that I deem should be obvious or already known from previous labs/lectures, I bombard them with questions until they figure it out. Unfortunately, this doesn't always work, possible due to a few reasons:

• The student is lazy. This might be a mean thing to say, but I think it's true. Graduate TAs that run consulting hours complain about students being unable to debug on their own, which I believe should be fairly intuitive. I've even had someone cry, possibly to get me to write the project instead.
• The student doesn't really understand the language. Compared the Scheme, Java is pretty complex, and some students never really properly learn some language constructs (e.g. the static keyword). This presents a huge problem when you get errors like, "non-static variable referenced in a static context." This is more of a lack of knowledge than reasoning ability, so this is where I throw my hands in the air and tell them to read the textbook.
• I'm not asking enough questions or I'm asking the wrong questions. This is quite possible; unlike the concepts presented so far in this book, so it's not always possible to ask an exhaustive list of questions to make the student reason through exactly what mess he/she has just made.

Regardless of the potential for failure, I think this could be applied to some specific topics for students to look like. The static keyword is probably a good target to start with. It is not as difficult to make an exhaustive list of questions for the students to "pattern match" through for one single concept, but I guess it's too late to use this kind of method when they're working on projects midway through the semester. Hopefully I can get some spare time to concoct an article for using the static keyword, and get some response from students on how beneficial it was.

Anyway, this has been a very random grouping of topics and I think I've poured out my heart and soul enough for tonight.

Dary.