Linus Torvalds Shows His New Polite Side While Pointing Out Bad Kernel Code

[timrichardson]

With language like this his lessons will be picked up ten times less effectively, so he will be plowing through ten times more shit. Anyone knows what that will result in.

Well, apparently the old way didn't work very well, evidence is the big collection of rude Linus posts, so kudos for trying a different approach after 20 years. In my opinion the messages are just as devastating, particularly the second example about the zero testing, but without the distraction of overly-emotional language.

[PreparationH67]

Glad to see the edgy teenagers have come out in force to comment /s

To expand on this further, I bet I could blow your minds with the revelation that being an asshole is not a trademark of intelligence and that its possible to be harsh but fair without talking about who sucks who's dick. Shocking I know, but true.

[PreparationH67]

Glad to see the edgy teenagers have come out in force to comment /s

[varikonniemi]

With language like this his lessons will be picked up ten times less effectively, so he will be plowing through ten times more shit. Anyone knows what that will result in.

[hoohoo]

It's good to see the sarcasm has not gone away. Seriously.

[Djhg2000]

I wonder how many people there are complaining about this, that have actual developer jobs in the real world that they enjoy? This is a fairly standard expectation of behavior for developers being paid at most of these big companies people complain about not contributing enough to the kernel.

That's kind of the thing here though, Linus doesn't really have to answer to anyone. He's still working on his hobby project but enough people like it for him to get paid.

Usually I'd agree with some of the criticism but to many this isn't an optional codebase. Plenty of platforms rely on the Linux kernel being stable with no real alternative. Therefore, if you're contributing to the kernel, you better know exactly what you're doing. Ignoring basic guidelines and a frequently repeated warning shows a level of ignorance that ultimately is harmful to the kernel.

Let me demonstrate by another example. Let's say you're working at a major car electronics supplier. In the ABS controller firmware (responsible for ensuring the brakes don't lock up on slippery roads) you add support for a new wheel speed sensor but your code breaks every other sensor. Now, do you think the maintainer should simply reject the patch or do you think you deserve a bit of verbal punishment for not testing such a basic use case? Keep in mind it's fairly critical code; with a bug of this magnitude you'd be sure to at least temporarily lose some customers. Also keep in mind you'd probably be looking at unemployment and/or being a target in a law suit if the bug was subtle enough to make it into the production branch.

Linus can't fire contributors. All he can do to deter further offenders is to show his discontent in an appropriately colorful language.

[Geopirate]

I wonder how many people there are complaining about this, that have actual developer jobs in the real world that they enjoy? This is a fairly standard expectation of behavior for developers being paid at most of these big companies people complain about not contributing enough to the kernel.

[Djhg2000]

this sounds like an interesting project for a weekend.

Most of the hard work has already been done in the XKCD fork of the "cloud to butt" browser extension. All you'd need to do is replace the word list and you have yourself a profanity filter (not to be confused with an anti-profanity filter).

[euler271]

It should be fairly easy to implement an online-translator from new-PC-Linus speak to old-non-PC-Linus speak - just detect detect some negative words (like "*not*"), then insert a few random profanities at the found positions.

this sounds like an interesting project for a weekend.

[bison]

I have the feeling this is going to end like an episode of Faulty Towers.