wqhff

Schizophrenia

[wqhff @wqhff3]

@sabatage_xxi Они фиксят по 30-70 CVEшек с каждым релизом) https://t.co/zkiTZixe5f

[wqhff @wqhff3]

@L01sValencia @splinedrive > simple
> linux
LOL

[wqhff @wqhff3]

@splinedrive @popovicu94 But Rust gives you the ability to avoid introducing an enormous number of stupid CVEs.

[wqhff @wqhff3]

@nktauserum А чем оно лучше, чем XORнутый wireguard?

[wqhff @wqhff3]

@vetroff https://t.co/gh3aY14me1 вполне рабочий на м1-м2, потихоньку вроде более новые процы прикручивают

[wqhff @wqhff3]

@valigo niri forever

[wqhff @wqhff3]

@Cluster_M Вроде можно ллмку нормальную прикрутить к Хоум ассистанту

[wqhff @wqhff3]

@Zeroawen @PETAUK

[wqhff @wqhff3]

@xeophon @teortaxesTex *146%

[wqhff @wqhff3]

@melqtx What's wrong with systemd?

[wqhff @wqhff3]

@MiraSecretAlt @Jush21e8 Why do you use little snitch instead of LuLu?

[wqhff @wqhff3]

@wateriscoding https://t.co/BJZE7s3oVG

[wqhff @wqhff3]

@effectfully **G**NU's **N**ot **U**nix!

[wqhff @wqhff3]

@kalomaze Okay, that's weird. Why didn't they fix this yet? I don't usually ask models to code, and when I do, brain filters the response, so I didn't really notice. I accept that I was wrong.

[wqhff @wqhff3]

@kalomaze As long as it separates the code from the text, it should be doable.

[wqhff @wqhff3]

@kalomaze I think you can work around this with a few lines of code and prompting. It will be faster and cheaper than training a model, and it will run quickly everywhere. I don't think the cursor uses a model for this.

[wqhff @wqhff3]

@kalomaze Maybe it's free in the sense of liberation, given there's bloody hand imprints on the van

[wqhff @wqhff3]

@tokenbender Isn't this setup essentially pretraining with extra steps?

[wqhff @wqhff3]

@MiraSecretAlt @ApriiSR @4confusedemoji @Calm13794 Why doesn't shitter make replies private when the original post is private? This schizo thread wouldn't exist if it worked that way.