Hacker News Reader

Comment by jnwatson | Hacker News Reader

I guess this is a meta question: do typical US undergraduate university Computer Science curricula not have Formal Methods as a class?

I'm just wondering because of the number of fairly basic questions here. I went to a middling midwestern state school 20ish years ago and that was part of the curriculum. Perhaps we have lots of non-CS-graduates here?

This is not to denigrate the discussion. I'm just surprised that this is news.

View on HN

While it seems anecdotally true that the mathematical foundations of computer science are being downplayed in universities in favour of giving students a more practical education in software engineering, which is an entire contentious discussion in and of itself. I think that a big factor in influencing your observation could be that CS is no longer the main entry point into a career in software development. In the organisation I work in, many of the younger hires do not come from CS backgrounds, and even many of the older hires come from other scientific disciplines.

if you want to get into software development, just go to any so called "boot-camp" class for few couple months.

Then you know how to write some webpages with new shiny JS framework, congrats you are a "qualified" software guy. Now go to write those sh*ty code ............

Math? what math? CS background? algorithm? if you are not going to challenge FAAG why bothered?

if you just want to code few pages, writing some sql query, there is no need for that. and for 80% of enterprises, they need only CHEAP engineers.

sorry for the rant, just another morning with lots of idiots bothered me.

I agree with you. Unfortunately there's numerous perverse incentives within the business sector that are promoting this. I'm glad I won't be alive in 2070 to fly intercontinental in an aeroplane with avionics written in Node.js...

I don't think NodeJS is an issue, the problem always lies in PEOPLE who wrote code.

If you want me to said what's the terrible thing? the code that very FEW people actually use, no one EVER VERIFIED it.

Do you dare to install these piece of crap to the plane you sit on?

Speaking as someone who's graduating this year, I wasn't exposed to it at all in my undergrad. I might be covered in one or two upper levels, but I hadn't heard of it until taking a graduate course where we talked about verification.

Here at the University of Queensland formal methods are in optional classes for computing degrees. I'm in the last six months of a part-time postgrad coursework degree, during which I did both a formal specification and a concurrency validation course. For my thesis I did a formal model using a proof assistant to describe event sourcing.

Neither formal methods class was particularly full, but neither was OS architecture, advanced algorithms, compiler design... machine learning and AI were both packed to the rafters though.

Kids entering university are picking based on current trends, and formal methods are not trending.

(Proof assistants are fun, though.)

I never completed my degree, but even at a very math-centric CS program I don’t remember seeing a formal methods class.

OTOH my recall of a curriculum from 15-20 years ago should not be relied upon.

There is huge variation between institutions and between individuals in terms of the electives they take.

My alma mater had everyone takes are an intro programming sequence, discrete math, and algorithms. Beyond that it was up to you. It could be all proofs, all systems programming, or a mix of the two. (I went with systems programming).

> Perhaps we have lots of non-CS-graduates here?

I think so. I'm an EE.

I can't speak to all of them, but most save that till graduate studies if it's covered at all.

Reply on news.ycombinator.com