I get that Linus is a superhero, but it’s still so weird to me that this vital piece of the world’s infrastructure relies on one man.
I think its better to think of it like a president or prime minister. He might set the plan and direction and making the big decisions, but there are thousands of others supporting and making the plan actually happen.
In the past he has delegated the release to others as well.
So if the worst would happen, the linux project would continue operating fairly seamlessly.
Is benevolent dictator still the official title?
Technically yes.
That’s because it doesn’t : ) He is the top level engineer/manager for releases and technical consultation but there are many more engineers “under” him leading and moving the pieces into place.
Linux is developed by a ton of people. As soon as Linus is out of the picture (say, because he retires), someone else will take his place.
Apple didn’t disappear just because Steve Jobs was gone.
It’s also mind blowing to consider that as many other projects, both Linux and Python started as a hobyist project never meant to do more than cater to some personal needs.
This taught me how important is allocating time for your team for their personal projects, as the next school romance anime tagging system could be the cornerstone of every AI in the future.
Except 99.999% of personal projects won’t be that popular and allocating time for personal projects is a waste in that regard. Basically you’d be playing lottery and not get anything out of it.
There’s plenty of reasons to encourage personal projects, but this isn’t one of them.
Lol, downvote this guy for pointing out that it might not make sense for your company to pay for your personal projects
Maybe because it’s a wise investment to encourage knowledge workers gain additional experience working on things they enjoy even if you might not be able to pick up one of those things and directly make another revenue stream out of it.
I didn’t, but I get why. It’s a specious argument — it doesn’t matter if 99% of them are useless. It matters if the 1% that become ubiquitous for whatever reason provide utility that makes the useless ones worth it.
Yeah you can run a company that never provides any time or resources to tinker, but only if you’re okay with innovation never happening again.
That’s pretty much all of open source to be fair. It’s a real problem.
Relevant xkcd https://xkcd.com/2347/
It is a small bunch of people though.
OSS is heavily undermaintained, always has been. But the world hasn’t exploded from it yet (somehow).
If you think OSS is undermaintained, you really ought to look at the way 90% of commercial software is developed.
It’s at least equally bad if not worse, with the added bonus that no one else can step in even if they really wanted to.
Wouldn’t surprise me to see unmaintained software anywhere.
The kernel will figure something out. There are already lots of companies investing their own development resources into it. Would just need a new leader to emerge. Perhaps it’d be a rotating group of people who are responsible for managing a single release.
Tons of smaller but important projects don’t have this luxury, though.
The kernel is totally safe. I don’t see anything happening to it. Even if something were to happen to Linus (oh hell no, please live forever).
But that’s not true for the projects that don’t do headlines, everyone uses, and nobody knows. When you install software and it has like 200 MB dependencies, half of those are probably unmaintained.
Also, the term maintained is not clear. Is a project with.a single contributor and some commits this year maintained? How about tons of contributors in the past but only a release 2 years ago? And you have to differenciate the usages too, curl is dead if it does not get updated, some config parser, ls, or cat is maybe as stable as they can be.
This is the best summary I could come up with:
“So this last week has been pretty calm, and I have absolutely no excuses to delay the v6.6 release any more, so here it is,” Torvalds wrote early on Monday morning, as version 6.6 debuted as planned.
Among the highlights of the release are the KSMBD in-kernel server for the SMB networking protocol, which adds additional features for sharing files and improving inter-process communication in Linux, hopefully speeding I/O.
Speaking of AMD, early tests by the Linux-lovers at Phoronix found substantial performance gains for its manycore “Bergamo” CPUs thanks to the inclusion of the Earliest Eligible Virtual Deadline First (EEVDF) scheduler.
The kernel also added support for AMD’s Dynamic Boost Control tech that allows users to tune Ryzen CPUs for optimal performance.
A change to this cut of the kernel rebrands it as just “SELinux” – a reaction to the Agency’s role in ops that have harmed privacy, per Edward Snowden.
US-based contributors will also have a Thanksgiving-sized hole kicked in their schedules, making it possible work on this release will be slow and Torvalds could push it into early 2024.
The original article contains 506 words, the summary contains 181 words. Saved 64%. I’m a bot and I’m open source!
Two included contextual references got lost - “speaking of AMD” and “this cut of the kernel”.
Crappy title.
It’s paraphrasing Torvalds himself though. It’s a cheeky title.
“… and I have absolutely no excuses to delay the v6.6 release any more, so here it is,”
Actually, I think I have misread it. My bad. I’m the one running out of caffeine, it seems.
I could have sworn he has used this joke before? Like in the past year?
I don’t actually follow the release schedule of the Linux kernel, but yeah, I was a little confused when I saw the mid-story link where he said roughly the same thing about 6.5 over the summer. That said, if we are going to call out middle aged men for repeating jokes, I’m in trouble.
