An oldie, but a goodie
He did hurt the mission. Plenty of kernel maintainers have left, and those were people who had been with the project for years. Losing experienced people to toxicity 1000% harms both the project and the product.
Did that demonstrably happen?
If there’s a surplus of talent (sounds like Mauro was dead weight) then at most he was just rude on Mauro’s way out the door.
I’m not saying it’s cool to be rude, but if it’s Linus’ review then you get what you get. To be butthurt about someone being rude to you should motivate you to learn your code interactions better. (In this case error handling)
You are just going to die on this hill, aren’t you? Even Linus recognized that his attitude was toxic, eventually, and that it was having a negative impact on the kernel development community. Yes, people left. Talented people decided it wasn’t worth the abuse.
You’re making judgements on people’s utility and ability based on the volatile reactions of man who admits to having issues. That creates toxic environments where people are not encouraged to do better, but any amount of change is due to fear of repercussions. This does not promote growth or new ideas that would genuinely improve something, but rather a fear of failure if they attempt something new. This also isn’t useful programming criticism because the actual useful criticism is buried in an emotional slurry that’s going to make something less receptive to the useful information.
Fear in the weak, those that can’t handle stern words.
Something line Kernel dev is not the space for fast and loose, and people need to be held accountable. Not coddled
And yeah, I’m making utility judgements based on Linus Torvalds. I’d say he has a pretty good eye for utility.
Yes people left and were not heard at the time:
No, was the product reduced or damaged, not did people leave. No one cares about individuals, if they can be replaced without blocking the progress of the project.
Those articles are very whiney. They chose to work on that project, with a singular leader. It’s his house, his rules, his standards.
In almost instances of Linus going off on one in public it is because maintainers weren’t doing their jobs (to act as quality gatekeepers), or particular developers thinking they could steam roll road changes through if they kept submitting them, or not listening to what Linus was saying. I remember Linus used to ream out Hans Reiser a lot (the guy who was subsequently imprisoned for murdering his wife) because he constantly tried to get ReiserFS into the kernel despite serious issues Linus had with it.
So generally when you see a rant, there is a history behind it and the rant itself is directed with a point. I also think it’s self evident that the kernel has benefited from this “benevolent dictator” model. I’m sure some people have gotten all precious over their feelings being hurt. The rest raised their game and the result has been a code quality standard you’ll probably never see anywhere else.
I can understand Linus getting frustrated at people who consistently push him (i e. Lennart) and I agree that there’s a reason he’s stayed at the helm of kernel maintenance and development all this time; however, that doesn’t denigrate that this is an unacceptable way to treat someone which Linus himself acknowledges! If this were about ReiserFS going into the kernel, I would understand that. But a poorly made commit should not be met with this vitriol. I’m not saying there shouldn’t be consequences for poor work, but this is not it.
