[jimmywetnips]

So I was building a side hustle right before I had to switch to job hunting due to illness. The side hustle is sooo enjoyable to me since I get to finally solve problems I've heard about tangentially but never got to work on in my day jobs.

I also don't mind leetcode, strictly as a way to learn esoteric algorithms that don't show up in 99% of crud engineering. It's also a fun game. aka facebook quizes for nerds.

But you can't just do the 75 blind questions and be done. You have to scope out the questions that a company is known for asking and commit those to memory, since everyone else is doing it. Last night I had to re-learn matrix operations and memorize how to do them for a company. I loved linear algebra, ...back in the day.

And I just had to laugh last night at the absurdity. "In what world would anyone implement their own matrix operations? Why the fuck am I being judged on coding in a shitty web ide and ability to sight read code like im at some fucking nerd recital? I should be able to use an ide with a debugger dammit"

Instead of just following my passion and intense desire to build my project, which incidentaly should be the same skills a good hire should have, I have to pull myself away from "real" engineering, in order to focus on gaming the test like I was taking the SATs again.

[hansvm]

> In what world would anyone implement their own matrix operations?

In any world where their matrices are "special" -- big, small, scattered across the internet, high read/write ratio, informationally independent columns with 0.01% error tolerable, poorly conditioned, blockwise sparse, used as a sub-component in differentiating anything hairy (like a function of eigenvalues), ....

You can often get halfway decent results stringing together existing matrix primitives, but throw in a 50 clock cycle latency budget, a $5k per-job compute budget, or all sorts of other constraints and you're just as likely to find >>10x performance that you left lying on the table (assuming your matrices are at all special, which is common but not at all guaranteed in arbitrary domains).