It's written by AI. I have made LLMs adopt the same tone by prompting to be engaging. Short sentences. Every point intended to land with impact. Artificial gravitas. I consider that a failed experiment and rewrote it, rather than posting to Medium

I was half expecting it to end with "And Kabir's real name was Anthropic's Claude Code."

This is basically the average rustacean's wet dream. But I can probably point out half a dozen reasons (without even thinking) why this wouldn't work in a realistic environment. Unless that one guy is literally a savant and the others are negative net contribution random monkey devs who were twiddling their thumbs all day at best. And in that case the end result would probably be the same, it has nothing to do with the specifics of this story.

If it's true.

A rust-factor of 1 too

It's a bullshit slop story.

But even if true, that's trivial to change hiring another dev (or Rust dev specifically) and giving it a couple of months to understand the architecture.

They literally do clap at some point in the story!

This is not just LLM drivel — It is drive by LLM drivel.

Add https://archive.ph/ before the URL and it will strip the paywall. But better to shun people who use Medium altogether.

> what the fuck were all of you doing?

Hotfixing fires caused by brittle software, causing more brittle software.

> How was he able to replace an entire stack in less than a month

There weren't hundreds of thousands of lines of code.

The conceptual model was already mature and well-understood.

So there was little new understanding to be gained, only fixing bugs in a second iteration.

> without introducing huge amounts of logical errors?

Supposedly by being a good engineer who tests things against the existing solution and new tests.

> The "healthcheck" that is provided is just junk

Yes, the whole story seems made up.

A moral of the story may be:

Build huts with mud, and skyscrapers with steel.

There are so many historical precedents to suggest that rewriting anything from scratch is extremely risky.

This story, while it may be fake, suggests that you shouldn't altogether abandon the idea and live in mediocrity.

As a person without a huge amount of legacy risk on my shoulders, I like that.

I didn’t read the whole thing because login but yes, firing the whole team thanks to one rockstar is the most braindead bit of management I have ever heard

Let’s say it’s easy to replace those people with Rust engineers (it isn’t)

Lets say there were no critical bugs introduced in a total rewrite (there almost certainly were)

Let’s say that morale among the rest of the org won’t instantly nosedive because a whole team got canned over architecture decisions they had no part in (it will)

Are we to believe that this unicorn 10x Rust engineer wont instantly be bored out of their skull because now that the team is gone there is no one left to maintain the devops pipelines, write all the docs and runbooks, and all that tedious stuff that any regular team handles?

How long until this rockstar unicorn ninja bails for another fun Rust from scratch project elsewhere?