How long can the C languages maintain their primacy in the kernel?
https://arstechnica.com/gadgets/2024/09/rust-in-linux-lead-retires-rather-than-deal-with-more-nontechnical-nonsense/
2024_09_04_15_19_38_.jpg
@romin @j
In this particular case there are no good guys, this is a reference case of shit prevailing over the piss. Linux kernel is an absolute clusterfuck — all things considered, it's kinda surprising it works at all, that is precisely why it's insanely hard to maintain in this state.
The cons of Linux kernel are… legacy code, the pros… also legacy code. There are zero good things about it.
@romin @j
No, I'm not going to reimplement that pile of garbage, but even if I intended to — it won't take THAT long because you don't have to implement all the quirks that it had accumulated historically, you don't have to support all the hardware and all the exotic architectures which were supported historically (and all this still contains bugs).
@romin @j
Linux detects liquid cooling pumps in my PowerMac G5 which never had them — because some variables don't get initialised properly, maybe it worked when someone from IBM submitted that code over a decade ago, but now it doesn't — and reporting it would likely only result in support for ppc64 being removed entirely.
@romin @j
I probably won't be using the re-implementation of Linux in Rust even if they did it — I would most likely only support x86_64 and AArch64 — architectures I'm the least interested in.
I'm just going to switch to FreeBSD — they don't have Perl in the base system, can you imagine how cool is that? 🤩
@romin @j
Besides, OS kernel in Rust already exists and works (kinda): https://www.phoronix.com/news/Redox-OS-Faster-VMs
That's what happens when you want to write code and make things — no need to reimplement Linux in its entirety and "face the toxicity head on".
@romin @j
Unlike Asahi Linux — a project that is completely pointless IMO as it boils down to making modern Linux system work on hardware that isn't open quite on purpose, this actually looks interesting. If they make this work even on a limited set of hardware, it would be great! But again, it only currently works on x86(_64) and ARM 🤷
@JoseMariaHDZ @j @romin
This was a great topic to explore in a longer post… and I actually did, but I wanted to proofread it properly and then I forgot to get back to you with a link to it 😅
gemini://m0xEE.Net/gemlog/posts/2024-09-05-asahi-linux-is-pointless.gmi
It's also available via HTTPS, https://m0xEE.Net/gemlog/posts/2024-09-05-asahi-linux-is-pointless.gmi , but I encourage you to use Gemini.
@JoseMariaHDZ
Oh, and sorry for quoting your post and putting it where you no longer have control over it. I can update it if you wish or remove it entirely, but I'd like to keep it for the context, otherwise the designer being brought up makes no sense 😆
@JoseMariaHDZ
Ha-ha-ha, yeah, when I was re-reading it, I got this impression too — but no, I didn't ignore it, I probably just didn't get the accents right, truth is — these people will never get a reliable system and the "harm" isn't people doing something they like instead of what I want them to do — like I said myself, I'm not in power to decide that.
@JoseMariaHDZ
We have very reliable system on Intel hardware, not only thanks to community effort, but because Intel contributed too — for Apple there is simply no incentive to do that.
@JoseMariaHDZ
What I think is bad about it is summarised in TLDR: they make it seem to other people that one day they would get a reliably working Linux system on this hardware, some might consider buying this hardware based on this promise. And I explain, from experience, why this won't happen this time — one essential component simply isn't there.