I'm impressed Torvalds managed to not know what he was referring to (the Twitter firings).
The missing context whenever this comes up is the fact that it was a surprise one off.
If developers have no idea they're going to be graded by lines of code at some random future date that's a much different situation than saying you're going to give bonuses away every month based on how many lines of code were written.
Everyone knows the second is bad, it'll be gamed massively. The first one could be useful though.
And yes doing it as a one off is still problematic and you can think of all kinds of exceptions, but if you think the organization is full of dead weight in general and overhired massively, a crude stack ranking by lines of code is a pretty good metric for figuring out which (e.g.) 50% is the bottom.
> a crude stack ranking by lines of code is a pretty good metric for figuring out which (e.g.) 50% is the bottom.
I can write you an efficient algorithm in 2 lines or an inefficient one in 50. The metric is about as useful as a doctor checking how often someone picked up a bottle to figure out how much they drink.
> I'm impressed Torvalds managed to not know what he was referring to (the Twitter firings).
I mean, naughty old Mr Car didn't _invent_ this nonsense; IBM was fairly notorious for it in the 80s, say. He's probably the most prominent recent example.
> The first one could be useful though.
How?
> a crude stack ranking by lines of code is a pretty good metric for figuring out which (e.g.) 50% is the bottom.
No. It's really not. For a start, you probably lose basically everyone very senior by that mechanism. But also you lose the troubleshooters.
Lines of code written, in isolation, is a strange metric to determine if someone keeps their job or not. I simply don't think this was the only metric. People love jumping to conclusions about divisive characters.
Running is a very low bar, you could theoretically fire everyone except operations and keep it running. But software and networks rot. Twitter has been rotting, it's no secret. It pretty much sucks major balls.
How much smaller user base? Looking at some recent data, which may not be accurate (but they're required to publish user numbers in the EU at least), it looks like the user base may be only 0-20% smaller compared to 2022.
If you are writing HelloWorld-webscale daemon from scratch, then counting +lines is probably "ok", but considering some existing large project like Linux (for instance), you would be well off keeping people who has managed to retain functionality while removing lines. Old projects have a tendency to get a lot of old cruft in which tends to stick (chestertons fence and all that) but someone clever enough to rewrite and remove old useless code is a net win for you, so I agree that if you fire some percentage on most-committed-lines you either had a very recent project from scratch or the measurement is stupid.
"How many lines of code did you write in the past <timespan>" is not a good metric but replaced with a more generic "what did you actually _produce_ recently" I can understand the spirit.
He didn’t directly call Elon stupid. He said that a certain behavior is so stupid it indicates someone shouldn’t be at a tech company. The next step is causal inference. I’m sure Elon is familiar with the nuances of slander laws after he baselessly accused a hero rescue worker of a terrible crime and then hid behind his lawyers.
> I’m sure Elon is familiar with the nuances of slander laws after he baselessly accused a hero rescue worker of a terrible crime and then hid behind his lawyers.
Funny how people remember this.
Hero rescue worker he was not. He just was one of the hundreds divers in the area, while total 10,000 people were involved ("The rescue effort involved as many as 10,000 people, including more than 100 divers, scores of rescue workers, representatives from about 100 governmental agencies, 900 police officers and 2,000 soldiers." [0]). There were teams from 11 countries, Musk included.
No one was asked out as per his made-up story during interview. Essentially he was slandering Musk...
Stupid and smart are stupid terms. There are many, many dimensions to intelligence and a lot of tech elite are too "stupid" to not recognize that being a really "smart" at coding doesn't mean they're really "smart" at everything.
I'd guess Elon and Linus's character sheets are more closely aligned on the same dimensions of intelligence than Linus would like to admit.
Tomorrow: Grokipedia entry about how Linus Torvalds is a rapist and baby child eater and also used the word pizza a lot of times in his email.
You're doubting this now, but deep down you know I'll be right. Elon Musk wants to win the race not because he believes in AI, but because he wants to be in control of the perpetual present.
The idea of firing people based on the number of lines of code they wrote makes sense if you think like a psychopath.
On average, more productive developers write more lines of code. Of course, writing more lines of code doesn't mean you are actually more productive, but the trend is there.
Elon Musk wanted to lay off 3/4 of a workforce of thousands because he thought 1/4 was enough, it is going to be disruptive no matter what and no matter how you chose, it is hard to predict the outcome. So, the general idea is to pick people randomly. But you want to bias that randomness towards keeping the best and laying off the rest, and so he picked up the number of lines of code as a criteria. It is semi-random and likely to be biased towards better productivity. It is thinking in terms of statistics, not individual people.
He is likely to be the kind of person who would have no problem banning black people from communities if it wasn't illegal. Indeed, there is more crime where there are more black people, so to lower crime, eliminate black people. And it will probably work if you ignore the fact that they are people and not just points on a chart.
I mean, you also likely lose the troubleshooters. Many completely indispensable engineers don't write a lot of code; if a month-long investigation leads to a hundred lines of code being changed and a 10% cut in machine costs, well, is that not useful work?
This would be particularly on Torvalds' mind, I assume; people who fix bugs in the Linux kernel (generally not a lot of lines of code) are generally more valuable to the project than people who contribute device drivers for obscure hardware (many lines, much productive, wow!)
If only it stopped at taking credit. That's not half as ego gratifying as stamping your vision of the future <cough>cybertruck</cough> on the roads where monstrosities with deadly edges assault our eyes.
Much respect to Torvalds, but all available evidence points to the fact that people that stupid (and some even dumber) do in fact work at tech companies.
Like many people on both political extremes, Linus Torvalds confuses people who disagree with him with people who are stupid.
Elon Musk is the best engineering manager this century.
And a dickhead.
Edit:
Just re-read the story. Going by the quote there, the interviewer lied about what Musk did and the story lies about what Torvalds actually said about him.
> The best engineering manager in terms of getting things done.
Beyond the joke truck thing, his car company hasn't released a new car in almost seven years. Twitter doesn't appear to have done much beyond release a few previously gated features (longer tweets, tweet editing, and the birdwatch/community notes thing were all things they were previously testing) since acquisition.
The missing context whenever this comes up is the fact that it was a surprise one off.
If developers have no idea they're going to be graded by lines of code at some random future date that's a much different situation than saying you're going to give bonuses away every month based on how many lines of code were written.
Everyone knows the second is bad, it'll be gamed massively. The first one could be useful though.
And yes doing it as a one off is still problematic and you can think of all kinds of exceptions, but if you think the organization is full of dead weight in general and overhired massively, a crude stack ranking by lines of code is a pretty good metric for figuring out which (e.g.) 50% is the bottom.
reply