Computer Science Curriculum in College

Ludwig Feuerbach writes "As it's back to school for university students, including Computer Science undergraduates like myself, I look at my course schedule for this semester and I have courses with titles like: Theory of Computation, Numerical Analysis, Artificial Intelligence, Machine Learning, and History of Economics from Plato to Keynes. The first 4 courses are required in my CS program. I had thought nothing of it until I read an opinion piece by Dan Zambonini, who stresses the type of courses I'm taking are, essentially, useless for getting a job. He lists several CS courses useful for a job. Is he right? I tend to think that an university education should stress scientific topics over vocational ones, but since I'm just planning to get a job after I grad, am I in the right program?"

Comments

Re:no

[WindBourne]

He is in the right program. A good Computer Science has a good to great theoritical underpinning. Look, the design paradigms will change. So will OSs. And the languages. and the DBs.

While 2 years ago, there were tons of CSers unemployed, so were the EEs and the CEs. Now, I do not know of any CSers that are unemployed. I do know of a LOT of CISers and vocational people who are unemployed. I also know a number of them have moved on to other professions because the industry has shrunk.

Basically, the CS/CE gives you the ability to do anything in the software world. The CIS/Vocational gives you the ability to do just what you learned. And back in the 90's, the CIS world was learning mainframes with Cobol, RPG, and PL1. Is that were growth is? nope. Has not been for sometime. Can these people move easily to Microsoft (where the most jobs are currently), or Linux (where all the growth is)? Nope. They do not have the underpinnings to make the jump.

Re:if you want more vocation, plus a better chance

[oc255]

I would agree with the usefulness of a Data Structures Course. It just so happens that I started a Jr Development (Java mostly) position recently and I was not a CIS/CS major (career change from SA/Unix).

I find myself trying to create a data structure/tree like a family tree or a directory structure. Each node has multiple children, a node can have no children, a fast way to find a path from a child to the root, etc. And in wondering how to create such a tree I find the usefulness of textbook knowledge, specifically the jargon. I'm reading like crazy about red/black trees, linked lists, doubly linked lists, binary trees, what a map is compared to a list, etc etc. It goes on and on. And I'm sure a course would have covered this or at least given me the knowledge to see quickly if the standard Java libraries have this structure already built.

Many of the Java books I'm reading have wording like, "if you remember from your CS class what a binary tree is, here's how to implement one in Java ..." The theory background would have made this a lot easier/faster, rather than jumping right into a specific language.

I definitely envy those who took any kind of data structures course.

Re:no

[danhirsch]

I couldn't agree more. Allot of software companies hire CIS grads...mainly because they are cheaper than full CS grads. What happens when a software company retools and moves on to new languages? Either the employees 1) Move on to other companies or professions out of frustration with having to learn a new language, 2) Get canned because they are no longer in need, or 3) Ante up and push through to learn a new language. Unfortunately, 1 and 2 are usually the direction things go. CS grads on the other hand have the theoretical knowledge to move between languages with a fair amount of ease. CS grads are the ones that don't rely on the fancy gui's or pre-built code snipits and controls to do all their work...they know how to do this in code....while many CIS's do not. For me...I can look at pretty much ANY modern language and tell what its doing...get a syntax reference book, and start editing. CIS grads that I know...would just freak...because they were taught more business skills instead of theory. I don't discount the CIS degrees, I just don't think they are very practical for the true-blue programmer or shall I say "engineer". I have recently been struggling on whether I should push through and get a masters in computer science or take the easy way out (for me at least) and get a masters in CIS. I keep having to remind me of these facts...over and over... CIS curriculum teaches "Java" or "VB", while CS curriculum teaches "Data Structures" and "Operating Systems". Sure..languages (c++) are taught in these classes..but the focus is on language structure...not the language in particular.

Re:The choice of degree matters less than attitude

[jwiegley]

Disclaimer: I am a university professor in CS. So, clearly, I believe in the value of a university education. But I'll try to explain why...

"I didn't learn C# in Comp Science but I could learn it in an afternoon.." I'm a young guy (22) and I've been programing professionally for nearly four years and I can tell you that this is simply false. Make no mistake about it, I'm still no coding grand-master and probably wont be for another ten years.

I think your statement here sums everything up nicely in favor of university degrees. You don't have such a degree, you can't learn new languages fast, can't recognize that ability in others and after four years you're still not an expert at the one language you do know and use daily.

Aside from the exaggeration of "[one] afternoon" which I agree is insufficient. You believe it's impossible because you yourself are unable to accomplish it due to your limited vocational training. Then you falsely project your own limitations on to others. As other posters have replied: yes. with a well grounded backround in theory and fundamentals it is possible to pick up yet another language in a very compressed period of time. (Though some of us benefit from an advantage in age over you.)

I have been proficient in the past with Fortran, Pascal, Modula-2, LISP and various assembly languages. I am currently proficient in Perl and shell and an expert in C, C++ and Java. (not trying to brag, a lot of /.ers have similar, or larger, skill sets and will relate to the rapid shifts in technologies that result in such sets.) The last job I took up required teaching advanced data structures in Java; a language I hadn't touched before the first day of class. Within one week I was productive in the language, within two proficient and within a month I was expert and using most of the advanced features of the language. I can't count the number of times my employment positions have put me in such a position where the programming requirements of the job have changed abruptly. I have always been ready to adapt to the challenge in a very, very short time frame and I believe this is due to my university-based education. I'm not afraid to change jobs or be fired because I know I can adapt and be valuable and productive in any new environment.

Here's two more examples:

1. Never during my education in CS did I expect to become a programmer and systems administrator for a Nortel phone system. But it did happen. I was ready for it and saved my company a lot of money in consulting fees because a dedicated technician didn't have to be called in to fix little issues.
2. Second example: One of my students obtained a job with an aerospace company and it was my responsibility to monitor them for a year to make sure everybody was happy with the arrangement. I asked what they were working on and they replied "debugging HPL programs. I've never even heard of HPL! How am I suppose to know this?". I said "neither have I. How are you doing at it". They said "Fine."

They were "fine" because they had the necessary theory fundamentals squared away. I would trust this student to pick up anything new and previously unknown in a short time period. In general, I would trust university educated people to have this adaptability more than vocationally trained peopl.

I'm really sorry for all the excellent, creative problem solvers you turned away because of your bias towards a single answer. "I don't know" is a perfectly acceptable answer (in any context) and doesn't, in itself, indicate an unworthy candidate. "I don't know; but I can learn it real fast" can indicate a truly flexible, useful person. Your loss; not the candidates.

Do you want a job or a career?

[LazyLawyer]

I got my CS degree in 1984. It's still useful, because they taught me theory, The languages they used (Pascal, PL/1 and LISP primarily) aren't.

My enthusiasm got me jobs. The degree only helped.

When I went to law school, almost everything I learned was theory. When I started the practice of law, I knew virtually nothing about actually running a trial. Now, I'm writing the book, and a publisher pays me for it.

Much of what I learned from the practice of CS and of law could have been taught at a trade school. 95% of the time, my work would be competent.

But that remaining 5% distinguishes between a tradesman and a professional. As a prosecutor, cross-examining the defence's psychologist or engineer, I have the advantage of knowing the basic theory behind their disciplines, because of the courses I took at university. I only tinker with writing software now, but I grok the new languages fast enough (when I get the chance to turn my mind to them).

I don't knock the trade schools. Enthusiasm to learn takes some people all the way through the theory they need to be pros. They don't need a university degree to be good.

And uninspired university graduates are so useless that should not be permitted to do anything important. I wouldn't hire them.

I remember that IBM used to hire only people with university degrees. Not just CS. Any degrees. IBM wasn't interested in what they learned at university. They wanted people who had the the enthusiasm/fortitude to slug their way through dry theory. A degree proved that the kid could work. Isn't that what an employer wants?

So what do you want? A job or a career? How much do you want it?

Re:if you want more vocation, plus a better chance

[alicenextdoor]

It may be because, unlike math or physics or chemistry... there's really nothing to "discover". There's lots to invent and improve, sure... but there's no real ingrained "laws of computers" that are woven into the natural fabric of the universe. There's no Plank's Constant or equation balancing or Realtivity. There's nothing that's tangibly abstract (???) that's just THERE to explore, discover and derrive.

On the contrary, there's a whole theory of computation that is far from fully understood. Godel's incompleteness theorem, Turing machines, the Halting problem, computational complexity (as opposed to trendy modern SFI-style complexity), issues surrounding self-replicating automata...some of the greatest minds in science, never mind computer science, have spent serious time on understanding what is computation and what are its abilities and limits. I'm sure there's room for one or two new thinkers.