Linus Torvalds Back in Charge of Linux | Hacker News Reader

Linus Torvalds Back in Charge of Linux

zdnet.com

I don't know what I thought would happen exactly but when he said he was taking time away I certainly imagined it would be a lot longer than a month.

To me on the other hand, it was obvious that he would be back as soon as 4.19 was released and the 4.20 merge window opened (the kernel release cycle has two phases, the merge window where new features and the bulk of the changes come in, and a phase where only bug fixes are supposed to come in). He took advantage of the time when the merge window was closed, since it's a less hectic period, and handed over maintenance to someone who already does something similar (Greg KH maintains the "stable" trees, which also accept only bug fixes).

That's impressive foresight.

Why did Linus make such fanfare about his departure if it was always intended to be so short?

> Why did Linus make such fanfare about his departure

I'm not sure he did. AFAICT he wrote an email on the Linux mailing explaining that he was taking some time off and why. He also did an interview or two, presumably because he was asked.

He probably realised that a portion of the internet would go bananas but that's not entirely his doing.

> That's impressive foresight.

Not really, it was on the announcement email from Linus (https://lwn.net/Articles/764901/):

"To tie this all back to the actual 4.19-rc4 release (no, really, this _is_ related!) I actually think that 4.19 is looking fairly good, things have gotten to the "calm" period of the release cycle, and I've talked to Greg to ask him if he'd mind finishing up 4.19 for me, so that I can take a break, and try to at least fix my own behavior."

If Linus asked Greg to "finish 4.19 for him", it's predictable that he'd be back for 4.20.

Because this was the first time in the history of linux this happened (a non-LTS release that wasn't by Linus himself) and Linus also wanted to take this as an opportunity to apologize for his behavior. Linux is a huge and very important project, so I'm not surprised this deal was a big deal and had a lot of coverage in HN or reddit.

Yeah I actually made the same assumption.

But, I suppose for someone who's always quite involved with everything and doesn't take long breaks- one month can seem quite long.

It was a political move from day one. As you might notice Kernel is not the only environment that adds and modifies their CoC currently. There is a group with political goals driving that effort. He wanted to see where the wind is blowing before making a move himself, as well as be flexible enough to also move into the corresponding direction.

That said, I also wonder if he moves too quickly, or if he found an agreement with other FOSS leaders.

Can you identify this "group with political goals driving that effort" or is it speculation bordering on conspiracy theory?

Can you check the other comments in the thread and any other post about CoCs in FOSS communities over the recent 4-16 weeks?

Check out this discussion for instance: https://news.ycombinator.com/item?id=18281676

Also note that my previous comment is at -4 already of the time of this writing, which must be done by someone who doesn't like that opinion. You wouldn't go ask 5-7 of your friends to give downvotes to that comment just because you disagree with it, right?

I am now waiting for a post from Linus entitled: "Behaviour therapists are just bullshitters" ;)

Welcome Linus 1.0.1-000 (first minor revision).

Linus debugged his own communication routines after users complained.

People can change their habits only to a certain extent...

Except for Unix kernels and version control systems, I would say. #hailLinus

sudo change behaviour -f linus20.json

JSON is likely the least efficient container for a behavior type. ;-)

sudo change behaviour didn't work.

Maybe you used the wrong option?

Not in sudoers file or group more likely.

> Linus debugged his own communication routines after users complained.

And found nothing wrong with them.

Personally, I'm perfectly fine with Linus as he was.

If Linus is really the messiah (ans he probably is), he will soon illuminate us again. About human interactions, this time.

Funny that their code of conduct contains all the standard equal rights no matter what XYZ, while this was never Linus problem. He never lashed out on people because of gender, race or sexual preference. It should simply have said "Don't be an elitistic assh*le aginst people just doing their best."

Obviously the COC isn't written just to control Linus' behavior.

The timing is impeccable, would you not agree?

The timing is easily explained from the perspective of the New Yorker's article publication and a want to do something visible to the contrary as evidence of a willingness to improve as a community.

Hence why the CoC is a whole lot broader in scope than anything Linus has ever done. It's more of a "while we're here on this subject and have everyone's attention, let's clean house a bit."

No. It's designed to be used to wedge, curtail and essentially interfere with the projects to which it is attached.

It's scope will expand to fill the space available to it. Simple as that. Quotas, funding limitations and project oversight requirements on projects will arrive in increasing numbers as time goes on.

Definitely seen the divides it's created in the communities. The CoC discussion is likely a positive thing as it is bringing up issues that need light in them. I'd like to see some evidence of the funding limitations, quotas etc.

That said, a lot of the CoC discussion rapidly goes into political discussions about meritocracy / Marxism. As I'm reading "Gulag Archipelago" at the moment, I'm starting to see some parallels in the modern world we as a species should definitely have already learnt from. The number of deaths from Marxism and Nazism is horrific.

I really wonder what's going to happen to Linux once Linus is gone. I think we've seen now that just a week after he took a break the community started falling apart rapidly.

Kernel fragments into a dizzying array of forks, TrueLinux, RealKernel, Linux Core, LinuxPlus and DogeKernel fight for supremacy on obscure making lists and github issue trackers.

The public remains largely unaware.

Huge PR battles over which logo to use erupt, somebody sneaks backdoors into the majority of the kernels. Somebody else backdoors the backdoors. Pretty soon over half the code base consists of backdoors and embedded nVidia graphics drivers.

Buzzfeed publish a list of 15 linux kernels you won't believe exist. A week has now passed since Linus left the project, and we have hit Peak Kernel.

Linux fades from the cusp of becoming a mainstream desktop OS and eventually settles into a niche server role until the 2038 bug wipes out humanity due to some buggy memory management running on an ICBM launch computer.

Chrome OS sees a slight uptick in users.

I really enjoyed reading this comment. I'm still smiling. Thanks. Now I can start working happy.

carlmr, you issued a nice writing prompt.

I've been keeping a plain text file of quotes for 15 years. I hadn't put anything in it since April this year.

Thank you.

This is perfect.

Did it? People actually involved in the Linux community seemed fine. It was just internet people shouting "the sky is falling!"

What I read about was Kernel contributors having some friction with the CoC. That got blown up by the press a bit, but it wasn't only the press.

That was just because the CoC was updated at the same time that Linus left. There would have been CoC controversy if Linus had never left.

You might have hit the nail on the head here.

The belgian king once resigned for a single day so the parliament could pass a law on abortion.

Could Linus have pulled of a similiar stunt?

https://www.nytimes.com/1990/04/05/world/belgian-king-unable...

I find it unlikely. It came up on the lkml that Linus requested changes to the CoC, and Greg was just the person who ended up managing the changes.

If you mean that Linus didn't want the flak for the changes, then that seems like weak leadership on his part, and since iiuc Greg is likely to take over kernel management when Linus ever retires, passing the controversial work to his eventual replacement is bad tactics as well, since it'll undermine Greg's authority in some people's eyes if he takes over.

If anything, drama related the CoC was what caused Linus' hiatus (in a "straw the broke the camel's back" sense), not the other way round.

He left after detonating a bomb. It would have played the same even when he stayed. He probably just left to avoid the shitstorm and his first violation of the CoC.

Though, it's true that with "eternal" project-leads like Linus, or recently Guido van Rossum at Python, we never think about what will happen when they are gone. But I think with big projects like Linux they will be fine. There are enough people to replace them, even if things might change, the project moves on.

We should think more about important projects from single developers. What about vim? Or all the important plugins/addons? There are so many small important gears around which depends on so few support.

"Some people on the internet with no involvement whatsoever in kernel development shouting about political correctness gone mad" !== "Falling apart rapidly"

Git seems to be doing fine after Linus left it...

True, but Linus is not the leader of Git since 13ish years. He gave the reins to Junio Hamano before it became such a large and influencial project, and Linux is way larger and way more influencial now. There's way more Linus "baked in" the kernel than in Git, I would say

It would probably be for the best if Linus stayed away for a while and saw what happens. That way he and the community could make a better plan for when he retires for real, and start steering in that direction now.

[citation needed]

Hopefully natural selection will bring us a profile similar to Linus that will take over the job.

...or various specialized kernels... Who know...

Very first thing? Kernel ABI becomes unstable, thus securing Linux's position as the OS most incompatible with itself.

Once there are enough moderators, community guidelines and steering committees in place it will just write itself.

Ballmer yelling at people "cancer is cured!"

>Jon Corbet, kernel developer and editor of LWN, thinks the Linux community will become less of what was once a "Wild West world with dueling developers and more of a professional developer community."

Well, that's what all this has always been about right? Power-washing the personality out of open source by subverting the popular movement of the day. A group wants something but there are very few actionable steps to achieve it, and the ones that do exist are long term investments. They're turned into unwitting campaigners for a different goal when someone provides them an easy answer that's really a thin veneer over a different purpose.

https://news.ycombinator.com/item?id=18278072

Also:

https://news.ycombinator.com/item?id=18274192

https://news.ycombinator.com/item?id=18273213

Thus restoring the natural order.

That was easy.

Reply on news.ycombinator.com