Slashdot Mirror
Linux Kernel 4.14 Will Be An LTS Release (softpedia.com)
prisoninmate writes: Development of the Linux 4.14 kernel series did not even start, as the version that's being developed these days is Linux 4.12, which should be promoted to stable early next month, but Softpedia reports that renowned Linux kernel maintainer Greg Kroah-Hartman announced earlier this morning that the upcoming Linux 4.14 kernel series will be an LTS (Long Term Support) branch. The developer promises to support the Linux 4.14 kernel series for at least two years after its release in November 2017, probably until November 2019.
"Development of the Linux 4.14 kernel series did not even start" I can't even get past the first clause, much less the first sentence.
Bahaha! Did he require them to call him that or are they just kissing ass?
"I believe in Karma. That means I can do bad things to people all day long and I assume they deserve it." : Dogbert
And if you had some actual understanding of the matter at hand, you would not post such nonsense. Here is a hint: A kernel is not an OS and what MS supports today does not have the same kernel that it had 15 years back.
The sheer stupidity of some people...
Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
You seem to be unaware that you can use a newer Kernel with an older installation and usually that just works.
Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
I'll address the "long term support" component of your comment, as your remarks about usability and features are clearly just trolling.
"Long term support" seems to mean different things in different circumstances.
While Win XP certainly had long-term support, at least for security patches, take a look at Windows 10.
Are the frequent mega-updates really long-term support, or effectively a new OS under the old name? After a couple of mega-updates, can we really say that it's the same OS? Especially in light of some massive breakage.
I do agree that 2 years isn't my idea of "long term" --- the Mint/Ubuntu LTS releases are 5 years, but even they have kernel updates from time to time.
That is 2 years for "same minor version". In most cases you can just move to a newer minor kernel without any problem. This is not a Windows-like set-up.
Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
I used to always run the latest Linux kernel, but since I run VMWare, I ran into too many problems with broken modules doing that. While I would love to see them get all of their modules into the mainline kernel (I fail to see why they really need something that couldn't be a generic service), that's really not relevant to this discussion. I've also had some problems with Nvidia modules, though to a much lesser extent.
So now my strategy it to always go with the latest LTS kernel. This has proven to be a successful strategy that keeps me with a relatively recent and stable kernel while also having one that will work consistently with outside modules.
Hopefully by pre-announcing the selection of a LTS kernel, distributions will make a point of selecting it to minimize their work in maintaining a stable and secure kernel. It was harder for them to do that when the LTS decision wasn't made until after the kernel was out.
No, but it means long term for the upstream Linux kernel. But you don't use that, you use the kernel supplied by your distribution where long term is something completely different.
my are you confused, you can pay for support for many other distros including Debian (my former employer used to support that and some other Linux distros). Open source is about choice, including whether you want to pay at all or not. It's also about having the ability to see into the software, even if you're not into coding you can find things from those who are that are useful.
Lowering expectations? You're telling me the Windows ports of some of the software I use runs better? No they don't.
I hope the fuck that it is better than 4.4. Having a LTS kernel come out with updates every few days is a complete joke. You may argue often, release early, but these are kernels used by the distos which do no such thing. Those that want release often, are more than likely to be using bleeding edge kernels, not LTS. Opensuse is at the moment using 4.4.70, yet the latest is 4.4.73. If you want Linux to succeed on the desktop, then bringing out new kernels every few days is not the way to do it.
You seem to be unaware that you can use a newer Kernel with an older installation and usually that just works.
"Usually [...] works" is not good enough for companies that have millions of dollars on the line. The whole idea of long term stable is to minimize risks like "usually".
Red Hat supports their OS and backport fixes to the kernel for 10 years. They just now stopped supporting 2.6.18, and still support 2.6.32. What Red Hat does not have is a light or embedded version, and unfortunately, linux.org does not support kernels for long when even LTS is only supported for two years. The result is that there's an awful lot of devices out there that run old and buggy kernels, and won't get updated because either backporting kernel fixes or bringing everything else forward on old hardware will take a lot of resources.
Why not just call it 4.2?
For simplicity. Personally they should have gone with a year-month versioning system, so you can see how non-bleeding edge you are...
"Usually [...] works" is not good enough for companies that have millions of dollars on the line. The whole idea of long term stable is to minimize risks like "usually". Red Hat supports their OS and backport fixes to the kernel for 10 years.
So if you want service and support you pay for it and you get it, I sorta fail to see the problem here. The linux.org team is there for development, for a relatively short period they'll fix their own releases but it's not meant as end user support. It's the point they expect to hand it over to Red Hat, Canonical, SuSE or what in-house or volunteer maintenance model you pick. It's not going to happen without someone making an effort, if nobody cares enough or is willing to pay enough to have it maintained well then it will be unsupported. Bring that grievance to the makers of the product you bought, not the kernel development team.
Live today, because you never know what tomorrow brings