Hacker News Reader

Comment by qznc | Hacker News Reader

computer science = information theory + software engineering

My opinion is that education should split computer science into the information theory part and the engineering part. The information theory part might even merge back into the math departments where it came from.

View on HN

I think the integration of software engineering as the core of a computer science program essentially destroys the computer science program. Sure, many people with CS degrees wind up doing software engineering, but comparing CS to SE is like comparing a metallurgist to a blacksmith.

That's exactly the comparison I was looking for.

Sure the blacksmith and the SE can build the final products but without the CS or metallurgist to build the foundation they wouldn't have a job.

:-) I think another important thing to note is that both software engineers and computer scientists reinforce and add value to each other's work. CS can improve and verify the methods employed by those in SE and those in SE can provide direction (necessity is the mother of invention) to and implement the methods developed by CS. It's a very unrefined observation, but I think it gets the general idea across.

Information theory has a meaning that is (I think) distinct from the one you are giving it. https://en.wikipedia.org/wiki/Information_theory

Actually, I think information theory is the only branch of science that has something genuinely useful to say about computers. Unfortunately, it has few people working on it in the context of computer systems. The world is dying for useful models of (real) CPUs and complex networks based on information theory, but that's not really happening. Computer "scientists" rather spend their time building example systems that no one will ever use, or creating theoretical models that have no connection to reality. No idea what the purpose of either is.

education should split computer science into the information theory part and the engineering part.

Then the engineers will complain that "Software Engineering" doesn't have enough rigor to be called engineering. Actually, it's been happening for years.

http://lesswrong.com/lw/9sv/diseased_disciplines_the_strange...

The problem is that there are two kinds of "software engineering." There's the practice of software engineering, which is the (hopefully) disciplined process of designing and implementing software. We're not sure how to best do it yet, but we all agree that it's important to get right because it's being done every day. Then there's software engineering research, which tries to evaluate ways in which people build software - but it's goal is not itself to build software. I admit I have been unimpressed by some software engineering research, but at the same time, I have no problem saying that the practice of software engineering is extremely important.

Reply on news.ycombinator.com