The dystopian world of software engineering interviews

That's what every bitter HN comment on the topic forgets. If Larry Page could sit down and talk to each candidate, I'm sure he'd be able to properly assess these kinds of corner cases. The challenge isn't being able to do that at an individual level; it's being able to do it in a way that's somewhat consistent and scalable across tens of thousands of interviewers and millions of candidates. Any large system is going to have false positives and negatives, and hiring is broken to begin with at most companies (assessing whether someone will be a good employee for the next few years based on a few short meetings is really hard)

> But it is an undeniable fact that some people pass this interview process; software companies do fill positions with this process.

This argument is devoid of logic.

We could have an interview process where everyone runs a 100 meter dash and the fastest get hired. Some people would pass this interview process and get hired.

It's obvious that running fast is not a great predictor of success at software development. It's not obvious that passing l33tcode interview questions is a bad predictor of success; but it's not obvious that it's a good one, either.

You haven't made an argument for either case, here. Personally, I don't think it tests for much that Raven's Progressive Matrices wouldn't cover; but that's illegal, because it's too obvious.

Right. Coding tests are designed to find people who are happy in a job where their only responsibility is: "Here's your coding assignment. Go do it." Technology companies are overfull of engineers who want more responsibility than that. They don't need any more.

I feel the need to nitpick your example.

> does the candidate want to code it up or understands that they could do: cut | sort| head

Piping together commands is undoubtedly programming, just in ancient shell script. So in a sense you're really testing for bash expertise. Which is maybe really relevant to you but I wouldn't say you're really "avoiding" coding by knowing to use those commands together, you just decided to code it with obscure shell commands. I might fail your test by choosing to write that sort of thing in python, because I could write it much more reliably in 10 minutes than deal with all the incredibly weird things that can happen in bash. I mean the final product in bash might be slightly more elegant, but your terminal history is probably littered with "man cut" and various attempts at it.

But for your example, my answer might be: if you're querying this file once, you may be querying it again, it's probably just way simpler to load the thing into sqlite instead of trying to imitate sql with some janky unix commands.

> everyone with real work experience has a story about sorting.

I am 34 and an experienced coder and I literally have no stories about sorting, and I've never once wanted to sort a CSV file on the terminal.

A few months ago I interviewed with Major CDN Company for a front-end dev position. They sent me a take-home React/NextJS project stub with dependencies and such already defined, and instructions to finish building out the full app. "Perfect!", I thought. No stage pressure, plenty of opportunities for going an extra mile. They encouraged me to get creative and I did; it met all the requirements and then some. I proudly submitted it.

A few days later I got an email saying, "Sorry, we're going to pass. The feedback from the person who reviewed it said that, 'It crashed with res.flat() is not defined when we tried to run it'".

"That's weird", I thought. I assumed they were running it in a different browser that lacked Array.flat(). Annoying, but maybe browser compatibility was part of the test (it hadn't been stated as such). So I did some digging just to be sure; I asked what version of NodeJS they were using. Version 10. Turns out that version of Node is somewhat old and doesn't have flat(). Huh. Dug some more.

.flat() wasn't even called in my code.

The stack trace went down into NextJS itself. They had given me a project with a particular dependency declared and then run it in an environment which was incompatible with that dependency, and then immediately punted it without any further debugging. I tried to engage my contact via email, presenting the proof that it wasn't my fault. I got an icy "Thanks for your feedback, we'll forward it to our hiring team", followed by silence.