back to article Linus Torvalds pens vintage 'f*cking' rant at kernel dev's 'utter BS'

Linux overlord Linus Torvalds has fired off an expletive-laden rant of the sort that only he seems to find acceptable. His post to the Linux Kernel mailing list takes aim at a chap named Kees Cook, who The Register believes to be a Google employee working on security for the company's Pixel phones. Cook appears to have earned …

This post has been deleted by a moderator

Silver badge

Decades of human factors research shows that you're wrong. If you have a loudmouth on the team then others tend not to question them even when they make huge mistakes.

45
7

+1 to that Adam

4
4
Anonymous Coward

Can you point me to an instance of Linus making a mistake?

18
7
Anonymous Coward

Every time he opens his mouth/responds to a mailing list

4
46

Grow a pair then. Part of being an adult is not picking up and running away when others might get hurt. There are no "safe spaces" in the real world.

25
22
Silver badge

Except when they are right 80+% of the time.

7
2
Silver badge

"Every time he opens his mouth/responds to a mailing list"

How can I explain this? I know.

Have you ever seen a newspaper headline that said "Liner crosses Atlantic without hitting iceberg"?

23
2

Yes. The way git handles renames. Every time I refactor something in a way that involves renaming of files, I cry and wish to go back to clearcase. Filenames are important metadata, despite Mr Torvalds rant on the subject.

10
1
Silver badge

"Except when they are right 80+% of the time."

No. The landmark incident was this. Captain Veldhuyzen van Zanten had an exemplary reputation, right up until he killed lots of people.

https://en.m.wikipedia.org/wiki/Tenerife_airport_disaster

4
2
Bronze badge

I am gon'na build a great kernel

...and make the Mexicans pay for it.

3
2

Interesting. Mercurial handles the rename nicely ("hg rename"), but it doesn't go in and edit all the places where it's referred to. But (using PHP) the autoinclude system handles that if you use a filename that matches the pattern for the class name. The autoinclude system works out the filename from the class name. So rename the class in the source file and wherever it's invoked, rename the file, you're done. I don't use git (or C, C++) so IDK other systems.

2
0
Silver badge
Boffin

1. Mr. van Zanten was under a "lot" of pressure and in a hurry (I know, both are bad)

2. He overestimated the reading skills of the Pan Am crew, that is, to recognize the letter "C" and the number "3". Had they exited at C3, as instructed by ATC, all would have been safe.

3. The tower was not communicating using official terminology, leading to KLM "believing" they had take off clearance, although they never specifically asked for it

4. Their fate was sealed by radio interference

Now, he should never have taken off with the thick fog in the first place, agreed.

You cannot blame van Zanten for everything, here ... he has his part of the blame, agreed, certainly not the whole blame.

Linus, on the other hand, has responsibility for a kernel, NOT PEOPLE, he knows damn well what he wants and what he does not want. He has strong words for those that oppose his views, unless you can convince him, but then you need a STRONG case!

The idiots over at systemd and Kees, here, are dead wrong. I have the impression Kees did not even read Linus' email he is replying to ... he even mentions WARN() iso WARN_ONCE(), who wants to fill the log ?

I do not agree with everything Linus thinks, like this:

https://www.youtube.com/watch?v=bw58LZTuZjA

But, hey, guess what, nobody is perfect, we are all entitled to our opinions, we should always question our opinions in the light of new data ...

1
0
Anonymous Coward

Linus Torvalds is obviously wrong each time he despises security... and no less than at kernel level!

0
2

"Can you point me to an instance of Linus making a mistake?"

Yes, and I can point to a kernel dev calling him on it and him backing down.

0
0
Silver badge

Errr, well point to it then i.e. link please.

0
0
Anonymous Coward

Big egos make bad code.

No amount of defending bad code you wrote makes it good code.

Good code is not elegant, uncommented, utilising recondite features of a compiler.

Good code is regular, lowest common denominator stuff written in a style that everybody understands including all compilers, that has stood up to being trashed by users' idiocy until its pretty much bullet proof.

Good code is not the creation of a genius, but of sober humble writers working long hours to create and even longer hours to debug.

If Linus has to use a few F-bombs to get egos out of Linux, so much the better. Its a pity that they didn't work with Lennaert...

85
9
Silver badge

Re: Big egos make bad code.

Exactly. There have been a tonne of instances where the developers have written shit code and pushed it through to the build, only for Linus to find them and kick off about it. They're simple issues, and they're issues that a developer really should be on top of. Blaming the checker at being at fault is like me going to the shop and saying the lottery machine is wrong when it says my "winning" lottery ticket doesn't have a single number on it.

32
2

Re: Big egos make bad code.

A tonne of instances? Don't you mean 1000kg ?

2
4
Trollface

Re: Big egos make bad code.

Surely you mean 1.1231 great white sharks? Or 0.0222 Australian trams?

18
0
Silver badge

Re: Big egos make bad code.

I could've meant 1,000Kg.

I could've meant 2,204.6lbs.

I could've meant 1.10 short tons

I could've meant 0.984 long tons.

I could have also meant 1,016Kg if speaking about ton.

I could've also meant 907Kg if speaking about ton to our North American friends.

But, hopefully, you'll still be able to get the idea that I meant he got a metric shit ton of instances.

13
1
Bronze badge

Re: Big egos make bad code.

What's that in imperial arse-loads?

10
1
Silver badge

Re: Big egos make bad code.

"What's that in imperial arse-loads?"

42.

19
1

Re: Big egos make bad code.

I remember, back in the '70s, my boss telling me the following anecdote:

The Chairman of the Gas Board (as it then was in the UK) issued the following edict "Sack all clever programmers"

:-D

1
0
Sil

Re: Big egos make bad code.

No one is blaming the checker for finding real faults in others' code.

The question is, can you check code and hold people responsible for their bad code, while being decent and not unnecessarily virulent, foul mouthed or offensive about it.

The question is, how many good devs did Linux loose, or will never have, based on Linus' insistence on behaving like a non-educated three years old.

1
2
Silver badge

Re: Big egos make bad code.

@Sil: The world already has far too many delicate little flowers in it whose parents convinced them over the years that "mummy's little soldier is special" and that they can do no wrong. I suffer this crap daily having to deal with clueless tits that just will not accept that they don't have a fucking scooby what they are doing. They are productivity detractors.

The man is trying to get shit done and doesn't suffer fools gladly. I have no problem with this. I feel the world would be a better place if there was less tolerance of whiny little twats not more.

1
0
Silver badge

Re: Counter argument

Where is respectfulkernel.org or politelinux.net? Show me a kernel fork making constructive progress without rants and the counter argument is made.

28
6
Silver badge

Re: Counter argument

"Show me a kernel fork making constructive progress"

Microsoft appear to be trying it. https://www.theregister.co.uk/2017/07/31/windows_subsystem_for_linux_to_debut_in_windows_10_fall_creators_update/

Whether it's making constructive progress is a different matter.

"without rants"

And as it's behind closed doors we aren't going to know.

5
4
Gold badge

Re: Counter argument

I don't think this is a kernel (fork or otherwise).

It almost certainly is without rants, however, since MS have to pay lip service to the norms of US employment. On the other hand, they probably sack the lowest performing 5% of their team every so often, which I'd argue was an even worse way to treat your devs.

10
1

Re: Counter argument

Ken,

Do you think that ranting at a coder about his or her bad code will make him a better coder?

If that works maybe you should tell teachers about your new teaching method.

2
4
Silver badge

Re: Counter argument

Given the behaviour, or rather lack thereof, currently observed in classrooms and the quality of the "educated" product that eventuates I'd say teachers are the last people to talk to about what works.

Those who can...do, those who can't...teach.

0
1
Silver badge

As I've said before ...

Wouldn't YOU be fucking pissed off if a so-called "professional" tried to pass off junk as working code, and expected you to put YOUR name on it? Frankly, given some of the egregious errors I've seen on the KML, and the lengths that the owners of those errors go to justify their junk, I'd say Linus shows extreme restraint!

54
8
Anonymous Coward

Re: As I've said before ...

Hmmm, well so far as I can tell all the chap has done is point out that, at the very most, there is some illogicality to the way faults are reported. He's not even written any code, this is old established code that could do with some improvement. Perhaps his only mistake was to assume that asking Torvalds for guidance would be worthwhile...

I'm wondering just how long Linux has got left as a cohesive, foundational project. RedHat / Pottering are trying to steal it. Google must be wondering why the hell they're bothering. Intel have had staff walk away from the project.

And then there's the more existential stuff; Rust as a language is really coming along, and the Redox OS written in Rust has made some stunning progress in a very short time. The whole idea of a kernel written in C is rapidly becoming anachronistic, not that many people realise it. Linux risks becoming a technological back water, though with a lot of momentum.

Other fundamental stuff needs changes. Everyone acknowledges that the network stack performance is below par, and it needs to be moved out of the kernel. That's a massive architectural change that's not happening because probably no one wants to raise the topic with Torvalds.

13
29

Re: As I've said before ...

I have no problem with the kernel being written in C. Good, and bad, code can be written in any language. I do take exception to Linus' approach to handling anything he deems to be critical of his work. That smacks of a God complex and that rarely works out well for anybody. Over the years I have heard many bullies justify their actions by saying it gets things to. The question is, does their approach cause things to get done in an expedient (fast delivery) or appropriate (robust) manner and at what cost?

13
17
Silver badge
Coat

Re: As I've said before ...

Redox...

https://github.com/redox-os/redox/releases

https://www.reddit.com/r/Redox/

Looks most interesting. Getting close to self-building, and they have ported gcc so there will be some applications to run on their GUI when it is a bit more stable. My favourite quote: "I am currently struggling with getting autotools based builds to work on Redox (warning: trying to read and understand a ./configure script may cause mental illness)."

Wondering if the focus is a GUI/desktop or server workloads, obviously shiny motivates the younger contributors but a minimal system that can sit somewhere on a network and run an application or two written in the system language could provide a niche that encourages adoption and a show-case implementation.

The more the merrier, but I think Linux has a significant incumbent advantage over the BSDs, Illumos based OSes and new ones like this.

Coat: 9front for the win!

6
0
Silver badge
Devil

Re: As I've said before ...

"I'm wondering just how long Linux has got left as a cohesive, foundational project"

a long time, I'd say.

"RedHat / Pottering are trying to steal it."

let them try. systemd kernel? HA HA HA HA HA! that's a laugh!

"Intel have had staff walk away from the project."

rage-quitters should STAY 'quit'. it's likely they were unproductive anyway. My $.10.

16
10

Re: As I've said before ...

The amazing progress of Redox OS includes re-writing the kernel from scratch, because existing virtual memory sub-system caused multiple memory corruptions and crashes in user land (https://www.redox-os.org/news/this-summer-in-redox-15/) that developers couldn't fix. So much for "safe" languages.

16
3
Silver badge

Re: As I've said before ...

I don't think Linus would object to the network code moving out of the kernel...

As long as it works, and has equal performance, and doesn't change the userspace interface.

He might even give a sigh of relief at not having to put up with it anymore.

16
0
Silver badge
Pirate

Re: As I've said before ...

"systemd kernel? HA HA HA HA HA! that's a laugh!"

Never underestimate the power of the Dark Side...

15
1
Anonymous Coward

@bombastic bob - Re: As I've said before ...

I wouldn't dismiss the RedHat trying to steal Linux. It is always good when you can lock-in customers and get rid of competition.

5
1

Re: @bombastic bob - As I've said before ...

Sounds more like a Oracle play, they pretty much stole ksplice from the industry at large.

4
1
Silver badge

Re: As I've said before ...

"As long as it works, and has equal performance, and doesn't change the userspace interface."

Linus isn't the only one who'd heave a sigh of relief. Some of those decisions made a long time ago when we were younger, more foolish and had less grey hair seem less smart with 20+ years of hindsight. kernel NFS in particular must die.

2
0
Silver badge
Coat

Re: As I've said before ...

I've come across "professional" programmers that makes amateurs look good, in fact, very good.

3
0

Re: As I've said before ...

"Some people have 20 years of experience. Others have one year of experience 20 times."

8
0
Silver badge

Re: As I've said before ...

@frobnicate,

I think you're missing the point. Crashes are good, they're Rust's way of pointing out bugs in your code. C is nasty because unless you do something spectacular to cause a segment fault or similar, your code will run quite happily causing mayhem that you might learn abroad during development, or you might not.

Rust itself is still evolving, but is a very good systems language. It's pretty hard to code for, because it doesn't let you get away with mistakes. That's its strength. You write junk code, you're going to be told all about it straight away.

Unlike C, where bugs lie dormant for decades undiscovered.

If ISO standardises Rust, it will become the language of choice for low level stuff like kernels.

I'm a long time C programmer, I love it to bits, but Rust is the writing on the wall. The speed with which Redox has gone from nothing to a running desktop is hugely impressive. The fact that they could bash out a whole new kernel very quickly, and apparently it's pretty bomb proof already, shows that it's a language where you can concentrate on ideas instead of worrying about memory all the time.

Large C projects like Linux will be seen as just too demanding of resources. There's a lot of people spending a lot of time chasing down problems in Linux that simply don't exist if Rust was used instead.

0
3
Silver badge

Re: @bombastic bob - As I've said before ...

I wouldn't dismiss the RedHat trying to steal Linux. It is always good when you can lock-in customers and get rid of competition.

Quite. If enough kernel devs end up being RedHat employees, they own Linux. Linus is good at driving people away from the project, leaving it vulnerable to a group with a plan and the money and the motivation to dominate the project.

If that happened, and they then forked Linux and went their own way, everyone else has to follow. "You wanna desktop, use our fork. Use Linus's original if you want but there's no desktop for you". It's worked with systemd, it would work with the kernel too. There's simply not enough people who care enough about the specifics of the lower layer stuff to resist it. Most people just want a working system and don't give a damn if the code itself was personally blessed by Linus.

The longer Linus keeps ranting at hardworking kernel devs, the more likely we'll end with Pottering being in charge of the only fork of Linux one can actually use.

2
3
Silver badge

Re: @bombastic bob - As I've said before ...

Ps that's an outcome that I'm not so keen on, to put it mildly...

0
0
Silver badge

Re: As I've said before ...

I do take exception to Linus' approach to handling anything he deems to be critical of his work. That smacks of a God complex and that rarely works out well for anybody.

I look forward to a comparison between the Linux kernel overseen by a man with a God complex and a rival one that is "designed by committee". That never works out well.

1
0

Re: As I've said before ...

@bazza,

Rust safety features (as well as similar features of other languages) are mostly irrelevant in the context of kernel development. They rely on the underlying model of fully separated protected address spaces. And kernel's task is to *implement* this model, so it cannot rely on this model. The bugs in the Redox where not some kind of wild pointer dereferences or use-after-free-s (these are eliminated by the language), they were perfectly legitimate memory accesses, but they still caused crashes, because the memory locations in question happen to contain page tables or DMA setup structures, etc. That is, those were not typical "segfault" kind or errors, they were logical errors, from which Rust protects no better than C. (And yes, such errors can stay dormant for years.)

I spent (<counts with fingers...>) about 10 years doing kernel programming for multiple operating systems and I can say that chasing pointer errors definitely wasn't anywhere close to the top problems.

0
0
Silver badge
Unhappy

I think I'd be less than delighted

A supposed professional doesn't accept that his own code generates a false positive and dumps everything. Yet the lead developer who's reputation is then being questioned is expected to act 'reasonable'.

Too many snowflakes around these days.

38
13

Page:

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Forums

Biting the hand that feeds IT © 1998–2017