[AussieWog93]

In defence of management, there is definitely a pattern of engineers (or employees in general, really) spending inordinate amounts on time doing something that they find personally gratifying but adds no real value to the product/business.

I've been guilty of this myself.

[veltas]

It's really easy to miss the value in Pointless Rewrite 5 of Core Module 4, but the grad who rewrites it will at least understand it afterwards and can fix the bugs they introduced and add new features to it easier. But the hard pill to swallow is they spent $200K of the engineering budget making the code worse.

The way to understand the value is to see the engineer as an asset. Engineers aren't interchangeable and no codebase can truly be documented and CI'd to the point where expertise doesn't matter.

[noisy_boy]

The pendulum has swung way too much to the other side. Now, people have to fight for just doing a reasonably basic job testing the code while the manager breathes down their neck with impatience fueled by unrealistic expectations of LLM magic (and this is even for those managers who are technically aware enough of LLM challenges so I can only attribute that to wilful ignorance).

The expectation of the code not blowing up in production still stands though. There is no arguing with this unreasonableness.

It basically sucks.