Linus Torvalds worries kernel 5.12 might be ‘one of those releases’ that lands a tad late Linus Torvalds has expressed concern that work on 5.12 of the Linux kernel is moving at an uncomfortably slow pace. In his weekly state of the kernel report, Torvalds announced the fifth release candidate for the new version of the kernel and described it as “bigger than average”. “We're not breaking any records, but it …
"can't they start the release now and mark it completed later"
Good idea! Maybe they could call it a "release candidate" instead of a release.
"when all of the development operations have ceased?"
Haven't you heard? All development operations will cease RealSoonNow[tm], when the kernel is done.
It's never late.It's either ready to release, or it's not. Late, in this context, is an artifact of diseased minds. .... jake
Oh? You think so?
Late, in this context, is surely much more an indication of a healthy concern for the power of its rc2 [release candidate remote control] impacts on allied systems, jake, while minor and miner driver glitches are fully sorted with battle hardening.
An abundance of caution in such an operational environment is default de rigeur common sense not least because of the catastrophic havoc, madness and and mayhem many of its impacted systems and CHAOS can choose to release to mentor[s] and monitor[s].
I suppose Linus is the prime, and probably only qualified candidate to adjudicate and pass comment on that tricky sticky matter which has elicited this novel brace of divergent opinions.
* That's a pretty bold statement to be responsible for whenever able to supply straws by the baleful. You don't get many of those out of nowhere for nothing nowadays. Others today would tell you that you don't get any, for they all each cost an absolute fortune which one cannot afford not to buy because of the price that is exacted whenever not protected from the aforementioned harms ..... catastrophic havoc, madness and mayhem.
** Clouds Hosting Advanced Operating Systems
No, I don't think so, amfM. I know so.
One can not put the date of art completion on a calendar, regardless of what Management's much loved Gantt charts might suggest. This is doubly true of FOSS software, where the Person In Charge cares more about functionality than punctuality. Would you bet the farm on something marked "fine vintage wine" before it had stopped fermenting?
Sorry, not a specialist, could somebody explain why it would be of any importance if the next version lands today or in a week?
It's not like they have deadlines, missing contractual dates and losing money. I would think their only concern is to improve the kernel, so why would a slight delay be of any importance?
(Genuine question)
"(Genuine question)"
I have the same question: I noticed that:
"few users rush to put the latest version of a kernel into production. Nor is version 5.12 designated as a release with long-term support"
actually ... who uses non-LTS kernels at all ? Yes, there was a time where I would compile the latest-and-greatest, but that time seems to have passed, I use the pre-compiled kernel of my distribution (5.8 currently). I suspect that the only people who use these non-LTS kernels are kernel-people themselves.
There are a lot of kernel fanbois who eagerly compile the latest, greatest kernel for their system, just to be able to boast that they are running it. Who they think they are impressing by demonstrating their complete lack of anything important to do with their lives is anyone's guess.
> who uses non-LTS kernels at all ?
Well I do, sometimes, if the current kernel has a bug affecting my (rather special) work flow.
But that doesn't mean I get me the latest-and-greatest one, I just try out a couple newer ones (chosen by checking their "known bugs" list), till I find one which fixes my problem. My computers are just tools, not a purpose...
I use non-LTS kernels all the time. And when I use LTS kernels it isn't the reason why - I just pick the newest production kernel every couple of months, compile and go. Never had a problem. But I wait until the kernel has had at least 5 or 6 updates. Right now I'm running 5.10-11, but before that I was running 5.9.8.
Why?
Serious question. If a given kernel works for your hardware, and gets regular security updates, there is no real reason to install a newer kernel. I'm still using 4.4.x in a lot of places, including my daily driver, and MeDearOldMum, GreatAunt and Wife's computers. It does what it needs to do, supports all our hardware, and blissfully stays out of the way while we use our computers. Can't ask for much more than that in an OS.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2023
