Conversation

Thorsten Leemhuis (acct. 1/4)

After six years the 4.14.y series will soon be EOLed with its 336 release: https://lore.kernel.org/all/20240108141854.158274814@linuxfoundation.org/

Side note & general reminder: You should slowly migrate to the 6.6.y series in the next few weeks on any systems used for "general computing", e.g. "'"where you have untrusted users or virtual machines"'". If that's news to you, checkout this blog post from @gregkh: http://kroah.com/log/blog/2018/08/24/what-stable-kernel-should-i-use/

1
1
0

@kernellogger @gregkh farewell to security patches for another generation of Android phones

2
0
1

@cas @kernellogger @gregkh Anyone can continue to maintain a 4.14 branch. If they find appropriate funding…

2
0
1

@ptesarik @kernellogger @gregkh yes, but better is to get newer kernels working on these devices - you can't maintain 4.14 forever

1
0
0

@cas @gregkh

yup. 🥴 Hopefully we some day reach the point where everyone updates Linux kernels like Firefox is updated:

You either update in small steps every few weeks or make one bigger step once a year.

2
0
0

@kernellogger @cas @gregkh Um. I don't know where to start…

All I wanted to write probably boils down to: I am not very hopeful that hardware manufacturers abandon their practice of shipping their own downstream kernel forks.

2
0
1

@cas @kernellogger @gregkh That's true. Still, open source offers an alternative.

I mean, you can't keep patching e.g. Windows 7, even if you found the time and money to do it.

0
0
0

@kernellogger @gregkh i hope so

and maybe we can hold someone accountable for the millions of perfectly capable devices out there which have been destined to landfill because nobody will fund the work to get them updated again (and because patching firmware is literally impossible without OEM signing keys), but i digress...

1
0
1

@cas @kernellogger @gregkh You don't. If hardware manufacturers switched to an upstream-first policy (or at least upstream after release), kernel version updates would no longer be prohibitively expensive, and nobody would have to support an old kernel for 5+ years.

1
0
0

@ptesarik @kernellogger @gregkh EU mandates and closer alignment between IOT and mobile are bringing us ever closer to this reality

ctrl+f "2031"

https://9to5google.com/2023/08/30/fairphone-5-updates-specs-release-date/

0
0
0

@ptesarik @cas @gregkh

People said similar things about Firefox when they switched to the current model ages ago. We'll see. It's quite possible that you are right. But maybe those new rules and laws from the EU might bring some vendors around.

0
0
0

@ptesarik @kernellogger @gregkh this is great for new devices sure (although without ensuring that firmware can be patched by the community it's still questionable)

but we can't just ignore all the existing hardware

1
0
0

@cas @kernellogger @gregkh Seems I should take some time to crack RSA, so we can sign our own firmware files for those devices. 😉

1
0
0

@ptesarik @kernellogger @gregkh i think there's some legal precedent with right to repair that could help us here. but I'm not a lawyer. hopefully this is something we're eventually able to resolve.

apple at least as far as i can tell fully preempted this, and they actually give you much more control over the hardware (ironically) compared to Qualcomm

0
0
0
@ptesarik @cas @kernellogger @gregkh CIP project maintains 4.4 and 4.19. Feel free to use our kernels, and feel free to join the project.
0
0
2