[chairs]

'Barely any developers who can't develop for shit.' That's a frustrating sentence to unpack.

Is the claim that the majority of developers are proficient because of Leetcode as a filter?

I don't think Leetcode is a useful interview tool when the actual role deviates from the style of questions that are expected to be completed in an interview scenario. And this is a prevalent issue.

Clearly it can be gamified and I don't think this makes for good engineers. It makes for people that are good at solving Leetcodes. Jira tickets aren't Leetcodes.

Whatever technical filter is implemented (pairing/take home/in house white boarding etc.) I care much more about observability, logging, approach to difficult problems etc. and general attitude.

Run me through the architecture for previous solutions that have been worked on. Did you own it? Were there any outages? What happened when requirements change? How can you ensure what you've created is fit for purpose for the organisation?

There's so much more to engineering rather than just code. And in my experience many engineers are worse at this than they think.

[ahartmetz]

Germany is kind of a developing country regarding software. There are good companies and many not so good ones. The things you are talking about, that comes after being able to write code, good code even. Architecture is not going to save you if you cannot write 200 lines of good code. Now, Leetcode doesn't test for good code but for "working" code AFAIU. And I'll be the first to say that code is written for humans to understand, but part of that is problem-solving ability.

Apart from that, I'm just passing on what I have heard from people. It's not first hand, but I believe it.