From mboxrd@z Thu Jan 1 00:00:00 1970 From: Dan Cross Message-Id: <200106091424.KAA17259@augusta.math.psu.edu> To: 9fans@cse.psu.edu Subject: Re: [9fans] could those of you who have students check this out for me? In-Reply-To: <200106091216.OAA08427@boris.cd.chalmers.se> Cc: Date: Sat, 9 Jun 2001 10:24:34 -0400 Topicbox-Message-UUID: b291c254-eac9-11e9-9e20-41e7f4b1d025 In article <200106091216.OAA08427@boris.cd.chalmers.se> you write: >I have discovered that today's computer science students, at least >some extremely good ones from here at any rate, think that algorithms >are a consumer good that you buy shrink-wrapped or something. At any >rate, the idea of going to the math library, reading a math journal, >and writing code of whatever strikes your fancy, is an utterly >bizarre and strange thing to do. We had better rock some minds then >if we want to have any cool algorithms to play with that we didn't make >ourselves. I agree, to some extent. Today's computer science programs are very much like sausage factories that turn out a mediocre product; computer scientists are the `spam' of the academic world. Why is that? Well, because the institutions let the students dictate the curriculum. Why is that? So they attract more students, so they can make more money. In turn, the majority of the students just want good jobs, and see a CS degree as a way of obtaining that. So, what ends up happening is that many students get vocational training in a four year undergraduate CS program, and end up coming out of school without any appreciation of even conception of the mathematical basis of comp sci. Without that understanding, they cannot build or even understand elegant systems; they are impressed by the thin veneer of the user interface and associate that wholy and completely with the underlying system. That's the only world they've known, and no one has taught them anything else. Of course, all of this is a sweeping generalization; there are obviously exceptions. Also, this isn't inherent in the students themselves; it's the Universities that do them a disservice by attempting to be ``relevant.'' But I've noticed a definate trend whereby students emerge with bachelor's degrees knowing more about Visual C++ than about, say, the basic sorting or graph traversal algorithms. It's a sad state of affairs, but I think that it will improve; I predict that within the next, say, 10 to 15 years, we'll start to see Software Engineering departments getting spun off from CS departments much as Electrical Engineering departments were once spun off from Physics. I predict that when this happens (if, indeed, it does at all) both departments will gain a new focus and retreat back to their core. I also predict that in industry, things like the IEEE Software Engineering certification will start to take off, leading to a greater focus on what's real versus what's hot today. We'll see. Oh, and no, I can't quite figure out how to teach Plan 9. Well, too things have worked so far; 1) As Kenji mentioned, the temporal filesystem is always a good demonstration; show them /n/dump. 2) I do a demo where I put an audio CD in my CD-ROM drive on my desktop machine, and then listen to it on my bitsy while walking around. That usually really impresses people, but assumes that you have a) an ``orinocco'' network, and b) a bitsy. Oh, and some audio CD's that people like. :-) - Dan C.