IshKebab

Isn't that basically how supercavitating torpedoes work?

Do you even need to fine tune? What happens if you just pick the next token but exclude options that contain e? Or would that send you down cul-de-sacs? Maybe it would work with diffusion based models...

For simple-to-medium boards you can assemble it yourself... No help if you're making a motherboard or whatever though.

> Buggy initialization shell scripts are not unusual

You can't really separate the buggy SysVinit shell scripts from SysVinit itself. That's classic "guns don't kill people, people kill people".

I prefer to not get shot.

> It's a rock compared to systemd.

Complete opposite of my experience. I've never had a single problem with SystemD - it's one of the most surprisingly reliable parts of Linux IMO.

In contrast SysVInit worked ok on servers and sort of on desktops, but on laptops forget it.

It wasn't meant to be bait. I'd really like it if Linux was as rock solid as Mac or Windows. But it clearly isn't and it doesn't help anyone that so many Linux fans pretend it is (look at the replies!).

That said, I do have hopes that AI will basically solve this. It's extremely good at diagnosing low level bugs, reverse engineering drivers and so on. A silver lining of AI taking our jobs at least!

We're at the dot com boom stage of AI. You could say it didn't deliver because boo.com didn't make a profit or because the average person still used paper maps, but that was just because of momentum.

It's the same with AI. The technology is here now. My friends in financial advice and market research have not lost their jobs immediately, because of momentum.

> at the level of any other Linux machine

So, 97% stable then? (Downvote me but it's true.)

> the tendency even now of LLMs to hallucinate

That seems to have been mostly solved through scale. I was asking Opus 5.5 about some obscure stuff and it kept hedging "this is from memory since I don't have internet access" (it's sandboxed), "I'm reasonably sure about this but not about the numbers" etc.

Even though it was obscure enough that Google couldn't easily find the paper that Opus remembered (I couldn't find it at all with the traditional search results; the AI summary found it on my third try), it actually didn't hallucinate it.

In fact I can't remember Opus 5.5 ever hallucinating. They've clearly done some work on that.

Obviously this local LLM is going to hallucinate all over you though.

Not purely, but "charity" and "only doing it to lock people in" are not the only options.

I don't think that's true. Many big companies contribute to FOSS in ways that aren't just to attempt to lock people in. Often it's because they use the software and don't want to maintain patches forever - everyone wins if things actually get fixed. Or even because they just want to be known as a good place to work.

> I actually think more of this software ought to be in assembler - whereas perhaps you'd expect I believe it should be Rust.

Yes obviously. Or potentially Zig once it is more stable.

Doing it in assembly in 2026 is absurd.

Not what I said at all.

Sure, but Qt (at least QtWidgets) is still really good, even if it is aging quite badly at this point.

The signal/slot system actually doesn't need MOC anymore, and the custom container classes are pretty obvious.

That said I think some Rust GUI frameworks are finally starting to mature. I quite like gpui-kit (formerly gpui-component). Maybe not as easy to use as Qt but it doesn't matter so much now with AI.

In places where you can use C you can just ask easily use C++. It has a number of nice ergonomic improvements that make it a better option than C, though you would typically opt out of exceptions, RTTI, and maybe the standard library.

Do we really need to post projects like this? It's obviously of no practical or even entertainment use. In the past it would have been a somewhat notable achievement due to the effort required. Today it's obvious that you can easily do this so what's the point? Showing off how many tokens you can waste?

It's like... if someone typewrites out every number from 1 to 1 million, wow totally useless but impressive effort! If someone prints that with a laser printer... Why? Who cares?

You want to use Qt. Or you're writing for a platform where you have no choice (most microcontrollers for example).

From a language perspective the only one I can think of is that C++ coroutines are much more powerful than what Rust gives you at the moment, even though they are very footgun-dense (quelle surprise).

But yeah, overall there is very little reason to pick C++, especially now it's feasible to translate existing C++ to other languages.

Wasn't alive then, but there's very obviously a difference between hype of Prolog, and hype of what is verging on AGI. Come on now.

The UK has dozens of these ancient laws that were never removed from the books. It doesn't mean they are enforced.

Maybe for "soft" subjects like History and Philosophy, where almost nobody who does those courses actually becomes a historian or a philosopher... But there are plenty of subjects where you actually do learn things that are useful beyond university. Engineering, law, medicine, chemistry, etc.

Are they just going to copy LFS? That's a real shame. I had some hope they would do something that isn't as shit. Oh well.

I guess they'll just copy submodules too. Kind of feels like there's little point using it if it's just going to copy Git's mistakes.

> Would we say the same before AI? Bad docs so bad project?

To a large degree, yes. But I would say it's more like "lack of care, bad project".

I don't think you can conclude that nobody cares about it just because they don't constantly mention it.

In any case, I don't think the lack of caring about CO2 emissions is unique to AI. How many people care about Trump's war on green energy? Not many it seems.

It did though. I've vibe coded three programs now. Two are replacements for existing programs that didn't work as well as I wanted and were missing features. The vibe coded ones work much better.

The third is definitely not as good as the commercial version, but it also doesn't cost $100k/seat/year.

Unfortunately work paid for the tokens so I can't share them, but I've seen it with my own eyes. It really works now. The code is not slop (and I am a perfectionist; it's not just that I have low standards).

It doesn't work with pre-Astra models for sure though.

I disagree. I was totally on the "vibe coding doesn't work and it's stupid" side a few months ago, when that was definitely true... So technically I was right, but I would say the "coding is solved!" people were more usefully right even if they were 6 months early. That's really not very early and it gave them a head start in preparing for today's reality.

Six months before the start of WW2 would your rather be "we're not at war, don't panic" or "peace is over, stockpile food"?

I keep hearing this. People being premature before does not mean they are premature now.

How long have people been warning about global warming? Do you still think it isn't real?

PLEASE rewrite this without AI-isms. AI can write perfectly good code now, but its prose - even the latest models - is just awful to read. It hurts my brain to read this.

I imagine that would go down well and result in zero forks...

Yeah because vibe coding has only really worked for 3 months. Give it a chance!

It'll probably be a couple of years before most companies even realise it's an option.

Oh don't worry, shitty WebOS is enough to put me off that (I unfortunately own one). However in this case we were talking specifically about that recent study of the effects of opt-outs, and for LG TVs the opt-out did result in the no longer detecting any tracking.