I don't know if this is a joke, but Minority Report was a fictional movie.

>Plasma will crash on me 2-3 times per day just doing regular things

This is not normal. Do a RAM test to see if you have a hardware issue.

It’s normal if you use the Nvidia proprietary driver. Every notification leaks one fd, so if you get a lot of notifications it’ll segfault once or twice per day.

This was apparently fixed in version 590 of the driver which was released only recently: https://forums.developer.nvidia.com/t/fd-leak-with-explicit-...

So that must be why I'm not affected. I have a gtx1080 which can't use that driver. Good to know.

RAM or the GPU is faulty. Definitely not normal.

I've been running plasma for over a year, there was like 1 crash during the 5->6 transition, it's otherwise been perfectly fine.

You know what else has a dozen crash related bugs in the bug tracker per day?

Literally any piece of complex software. See LLVM for example. LLVM is the backbone of most compiler work in the world right now.

https://github.com/llvm/llvm-project/issues

I have the same issue, only on KDE, never anywhere else.

Does KDE use RAM differently than every other software?

Yes, KDE aggressively caches and indexes things by default whenever you have free RAM unless you disable this behavior in multiple places in multiple applications. For example, in Okular you can tune it to choose how much of a pdf you want to keep rendered in memory, if you have a tonne of memory, this makes it the smoothest pdf viewer I have ever used.

It has become reasonable graceful in giving it back when you you need it nowadays.

The Linux kernel does this too, yet it does not crash like KDE. At any given moment, most of your free RAM is used to cache stuff by the kernel, unless you've recently rebooted.

No, what the Linux kernel does instead is randomly kill user processes :)

It's kinda infamous for that, and had held up Linux adoption for a decade or so.

But you sort of missed the point, I think. The comment chain was about speculating why KDE could possibly crash if there was faulty RAM while other software would be fine. And the kernel absolutely crashes when there's faulty ram.

Right, yet it's KDE that is crashing with extreme regularity, and never the kernel.

The comment by Arch-TK is currently the top for me.

Interesting. RajT88 20 minute old post is top on my end.

For me, the top comment is currently cmarschner's day-old comment enumerating a bunch of other examples of botched redaction

>All the numbers point towards us hitting up against planetary limits, at some point something's got to give.

Sounds like space is the next big thing then.

>I’d think people are looking for these services online or in some gig work app.

Then you'd need to prove your identity and pay taxes on what you earn. This is for illegal immigrants working under the table.

It's also only in some areas. None of the big home improvement centers where I live have anyone hanging around looking for work.

I've seen it in Atlanta, GA, Phoenix, AZ, Kansas City, MO, Anchorage, AK, and Chicago, IL.

What did you use to make that chart? It looks really nice. Its the first time I've see these ASCII boxes on HN without gaps in the border.

>Can you define what "reacting" means exactly in a shooter

A human can't really, which is why you need to bring in ML. Feed it enough game states of legit players vs known cheaters, and it will be able to find patterns.

There is no need for ML. Games arent the real world.

A suitable game engine would have knowledge of when a shadow, player, grenade, noise, or other reactable event occurs for a given client.

Especially if games arent processed in real time but processed later based on a likelihood of cheating drawn from other stats.

And what happens to that pattern, when the cheat engine adjusts? What happens to the enraged players that got wrongly banned for cheating?

Yeah, that's why you need a data scientist or two to figure that stuff out. Its a solvable problem, but you're not going to get solutions instantly for free in the reply section of HN.

But in the reply section you can read about that it has been tried in reality, with not so much success as in theory. But if you see a working solution, then you don't need to tell me, but can market it yourself.

If anyone is wrongly banned the system is too sensitive. Let it capture data for a month before banning someone. Ensure the confidence is crazy high.

Fireworks supports this model serverless for $1.20 per million tokens.

https://fireworks.ai/models/fireworks/deepseek-v3p2

That's the final, fine-tuned model. The base model (pretraining only, no instruction SFT, RLHF, RLVR etc) is this one: https://huggingface.co/deepseek-ai/DeepSeek-V3.2-Exp-Base It's apparently not offered at any inference provider, nor are older DeepSeek base models.

I've never had issues with Debian based distros.