Hacker News Reader

Comment by tormeh | Hacker News Reader

I think this is more about the case for turning some universities into traditional grande ecoles. Engineering and business schools with a practical emphasis and a minimalist approach to theory.

I'm an engineering student and by now my brain has learned to tune out whenever the words "rigor", "proof" or similar are uttered. I really, sincerely don't care about the formal stuff. If I get the intuition then I'm satisfied, and I'm sick of the formal stuff. Someone's got to do it, but I don't want to be one of them. This doesn't mean I don't want to learn about advanced stuff; it just means that I'm fine with a shallower treatment with an emphasis on implementing and intuitive understanding rather than proving. I would like a school more to my tastes, but university is where all the smart kids go, so I go there as well. I would like an alternative, and I think I'm not alone.

View on HN

I had a similar frame of mind when I was a student. I didn't pay too much attention to the formal details; I didn't brush it off, to be fair, and I think I paid more attention to it than most of my colleagues, but I still massively underestimated how much I really needed it.

Reality kindda bitchslapped me later. Turns out the theoretical understanding of even the most basic and otherwise fundamental aspects of my slice of engineering (EE) is pretty much essential whenever you need to design something that is non-trivial and not previously implemented by someone else and available on the Internet. All those circuits I tweaked and combined were great study material, but when it came to doing something on my own that I couldn't quite find elsewhere, which had to respond to actual hard requirements (no more than this many mV of ripple on this signal and THD below this margin), it wasn't an enjoyable ride. And then there are whole branches of EE where you don't really get to prototype stuff on a breadboard; at one point I thought I wanted to get into microelectronics (programming got the best of me though). There are literally thick books about how transistors behave, and you literally need to understand that stuff if you want to make anything slightly more reliable than a 741.

To this day, I still value intuitive understanding more than the bland math. It's the idea percolator that makes you come up with clever designs and lean, elegant solutions, but it's the bland math that actually enables you to turn it into something that can be reliably manufactured and exploited.

Analog electronics is a unique, not representative case.I think it depends greatly on the field and application. Some applications are too complex for math, so you just have to create really good simulations, like for example in wireless communications.

In other applications , value comes through integration, managing complexity ,smart component selection, and rapid development and combining multiple knowledge domains.One such example is smart watches. Some of the theory is less valuable there.

> Some applications are too complex for math, so you just have to create really good simulations, like for example in wireless communications.

Understanding the math behind those simulations is absolutely crucial to not getting bit by them.

> One such example is smart watches. Some of the theory is less valuable there.

The experience of working on low-power devices with engineers that do not understand the theoretical fundamentals of the devices they work with is an experience I can only describe as very, very painful. Smart watches are an excellent illustration of this, yes. "Smart" component selection with no understanding of the underlying complexity of the technology is a sure way to ensure you end up with crap that's 30% more expensive and 50% more power hungry than it should be.

> "Smart" component selection with no understanding of the underlying complexity of the technology is a sure way to ensure you end up with crap that's 30% more expensive and 50% more power hungry than it should be.

Can you please expand on that ?

Actually, I almost feel, that in the czecho(-slovak) tradition, it almost is that way. There is a pattern, where often you have at least two universities, and one of them would be branded "technical".

I.e. Charls University and Czech Technical University in Prague, Masaryk University and Technical University in Brno, Kommenius University and Slovak Technical University in Bratislava, and I think I'd probably find more examples.

Usually, the technical ones brand themselves as being more "practical". For example, I studied CompSci at regular uni, my friend at technical and while we had a mandatory two semster algebra course and mandatory two semester discrete math course, they had these two combined in one course, first semester only. But it was more of a rule of a thumb, I know that their math-analysis course was as rigorous as ours. They still had better partnerships with local businesses.

I have to admit, that I was reasonably fond of all the theoretical stuff (although not very good at it :) The two semesters of group-theory certainly made my decision to do my masters in kryptology and information security much easier :P

In regards to your second paragraph, I think you would love starting a hardware startup. :) That's what I did. Although I did study CS in school, in terms of electrical engineering I'm almost entirely self-taught, learning by "implementing and intuitive understanding" as you put it. It works really well actually, my team and I have been able to pull off some pretty incredible feats of engineering, working this way.

I can imagine such "practical engineering" online degree being formed, which will give one the ability to create systems that combine multiple domains of knowledge.

Having people with such capabilities is highly useful in many startups, especially at the early stages, and startup access is percieved as a chance to wealth, so it seems such a degree would have significant marketing value and differentiation, maybe to a degree that would surpass it being online.

Reply on news.ycombinator.com