[mas-ev]

I have still yet to see evidence that people were fired for their "lines of code added" as everyone is implying.

Would it not be reasonable to take into consideration a employees LOC as a metric?

If you're an engineer and made very few code contributions without a solid justification of why, I'd expect you to be on the chopping block...

These headlines are misleading and people are jumping the gun making assumptions as always.

[isthiseasymode]

A simple IDE assisted refactor could inflate your line count by a few thousands.

You can also just write a ton and not actually need it if you’re trying to game the system or if you’re inexperienced.

Commit count would be a little more accurate, but that also is just an analog for who saves the most. It doesn’t really tell you who made any impact with or without code.

Finally, a lot of senior work is just convincing other teams to allow you to do what you’re trying to do. Most code features are trivial to write, but a nightmare to get approved.

[Hermitian909]

> If you're an engineer and made very few code contributions without a solid justification of why, I'd expect you to be on the chopping block...

At a company of Twitter's size you generally expect the junior and mid-level engineers to be writing most of the code and the more senior engineers to be implementing the trickiest bits, managing stakeholders, and writing design docs.

At the extreme end, engineers working to optimize performance might spend months painstakingly optimizing a single hot loop to save the company a few million a year (I've seen this happen).

Given this reality LOC, is so wrong for more senior engineers as to be a waste of time to even consider.

[watwut]

Juniors do most code, seniors less. Those who work on easy tasks produce most code, those who work on hard ones less.

And then there are usually senior positions where producing code is just part of your responsibility - compliance, analysis, negotiation, architecture, operations, troubleshooting shooting.

[dom96]

So you're doubting that LOC was used but also completely happy for LOC to be used as the criteria for firing?

[mas-ev]

No. I DO believe lines of code was used as metric. It's a fast way to protect the top code contributors. I just don't think it's the only thing taken into consideration like people are implying...

There are countless teams and examples for it being a horrible metric to solely fire on.

I just think it's not right to jump to conclusions like Dictator Elon fires engineers with low LOC.

All we have to go off of is they had review meetings where they had to show off the code they contributed and Elon "suggested to query for lines committed."

[poszlem]

Is believing in both of those things is somehow a contradiction? "I haven't seen a good proof that he fired based on LOCs and even if he did that would not be stupid anyway"?

[isitmadeofglass]

Just because you take something into consideration does not mean it becomes a criteria. But I think we can all agree that it would be a mistake in a mass layoff to accidentally fire the top 10 contributors to a given project if the 11th contributor only ever made 3 contributions all just updating comments right?

Well that sort of logic requires you review the logs. It doesn’t have to mean you set up some arbitrary “anyone under ckloc per week is fired”

[Ar-Curunir]

lmao, and what about teams that focus on, say, security, or privacy, or devops, or numerous other areas that require reading and auditing code/systems, as opposed to writing new code?

[mas-ev]

I agree. I viewed those as "a solid justification of why". Also, I'd expect there to be some grouping or consideration of job titles and organizations.

No body knows unless real leaks come out. Did they do it strategically? Or by numbers across the board like people are implying?

[MichaelCollins]

In these threads, we're all pretending that finding evidence of non-performance in VCS history isn't common practice across the software industry. Because fuck Elon Musk, he's just the worst.

Whether your counting lines changed, number of commits, etc, you can easily catch out the people who do jack shit. Cross reference that metric with their other duties and it is easy to find a lot of slackers. The people who should be coding a lot, but just don't have any work to show.

[ShamelessC]

Are you yourself a software engineer?