Hacker News: Retr0id

New comment by Retr0id in "Qwen3.8-27B on the iPhone 18 Neural Engine at 15–70 tok/s (1.5 bpw)"

Retr0id — Sat, 17 Oct 2026 19:04:28 +0000

OP, asking for upvotes is against the HN guidelines: https://x.com/anemll/status/2111531629529456771

New comment by Retr0id in "Tesla Roadster"

Retr0id — Fri, 16 Oct 2026 20:36:44 +0000

The gas tanks take hours to refill, so even with one car that could do it all they'd need 3 of them.

New comment by Retr0id in "Llama.cpp on a jailbroken PS5: run any GGUF model natively on the console"

Retr0id — Fri, 16 Oct 2026 20:22:21 +0000

I'm sure Sony has an opinion of the jailbreaks that enable this, but they really don't give a shit about anyone running llama.cpp. Why would they? They're just glad you're not pirating games.

New comment by Retr0id in "Claude reveals ECDSA curve origins"

Retr0id — Fri, 16 Oct 2026 19:56:20 +0000

I have seen behaviour a little like what you describe, in other tasks. When I was cracking the keys for Sony's camera firmware updates, Claude steered away from an obfuscated library, seemingly because it thought it might be related to licensing checks. It was indeed related to licensing checks, but also firmware update decryption. I had to explicitly tell Claude to write a deobfuscator for it, which it did, and eventually I had the keys.

I later got Claude to crack their licensing checks too, by convincing it to write a "license checker" program, and then a second "test case generator" script for the license checker... (Opus 4.8 was smart enough to see what I was doing, but Opus 4.6 complied)

New comment by Retr0id in "Claude reveals ECDSA curve origins"

Retr0id — Fri, 16 Oct 2026 19:40:34 +0000

I saw no evidence of sandbagging. And since it was an open problem, there'd be no reasonable way to "steer away" from the correct solution.

New comment by Retr0id in "Claude reveals ECDSA curve origins"

Retr0id — Fri, 16 Oct 2026 19:27:20 +0000

I worked on this specific task myself, using Claude, and I was not blocked (although I did not succeed!)

New comment by Retr0id in "Steam Client is now faster than Epic's new vibe coded launcher"

Retr0id — Thu, 15 Oct 2026 18:30:32 +0000

I don't know why anyone's assuming Valve isn't "vibe coding" too.

New comment by Retr0id in "Wayland desktop for Linux with skeuomorphic design"

Retr0id — Thu, 15 Oct 2026 18:14:15 +0000

It's chock full of Claudisms!

> The room is ray-traced, and one desk lamp lights all of it.

> The first login draws the room once: a few seconds on a GPU, longer without one. After that it starts at once.

> Nothing here is a symbol for a thing on a desk.

> each is drawn to look like the real thing set down there

I could go on...

New comment by Retr0id in "Wayland desktop for Linux with skeuomorphic design"

Retr0id — Thu, 15 Oct 2026 15:04:22 +0000

fwiw that is the standard term for that kind of fold (to the extent that it has its own wikipedia article), but I agree that the claude prose is gross.

https://en.wikipedia.org/wiki/Dog_ears

New comment by Retr0id in "Wayland desktop for Linux with skeuomorphic design"

Retr0id — Thu, 15 Oct 2026 15:02:34 +0000

> the kind iOS 6 had

> The room is ray-traced

imho, part of the charm of iOS6 was that everything was baked into native-resolution textures.

New comment by Retr0id in "Wanted to download an ISO, ended up writing a kernel exploit"

Retr0id — Thu, 15 Oct 2026 11:56:16 +0000

There's a 4th tier for the people who enrolled before they introduced the 3 tiers thing. It does almost nothing these days but I think it reduces the guardrails on Opus 4.8, and maybe 5 also, but definitely not 5.5.

New comment by Retr0id in "AI was great until it was not"

Retr0id — Thu, 15 Oct 2026 09:55:27 +0000

There's a wide gap between "introvert" and "no interaction ever".

New comment by Retr0id in "Gaming on Linux without cheating"

Retr0id — Wed, 14 Oct 2026 18:49:01 +0000

Good LPEs escalate you to kernel (which is true for most exploitable kernel bugs).

New comment by Retr0id in "Gaming on Linux without cheating"

Retr0id — Wed, 14 Oct 2026 18:44:38 +0000

Very exciting, I've been looking for a fun target to demo my attestation bypass techniques against.

New comment by Retr0id in "AOSP Source Releases Are So Late Developers Disassemble Libraries to Report Bugs"

Retr0id — Wed, 14 Oct 2026 12:36:00 +0000

Yeah, you could get 95% of the way there with binary diffing

New comment by Retr0id in "AOSP Source Releases Are So Late Developers Disassemble Libraries to Report Bugs"

Retr0id — Wed, 14 Oct 2026 09:16:05 +0000

It'd be mildly amusing if this led to someone doing an AI-powered decompilation of AOSP.

New comment by Retr0id in "Show HN: Spotatui, a Spotify TUI/GUI without web API rate limits"

Retr0id — Tue, 13 Oct 2026 22:52:41 +0000

I'm a bit confused, are you saying you worked around spotify's ratelimits by using non-spotify APIs to achieve similar things? Why not just RE the official client and use the same APIs it uses? (You can get lossless that way, too!)

New comment by Retr0id in "Sqlthon – A lightweight Pythonic layer for SQLite"

Retr0id — Tue, 13 Oct 2026 13:33:01 +0000

Top quality SQLi prevention: https://github.com/SAUMS-1391/sqlthon/blob/3bae9e24fb6a6846d...

New comment by Retr0id in "A Modern Web App in C"

Retr0id — Tue, 13 Oct 2026 13:03:32 +0000

The inefficient part is execing a fresh process on every incoming request. The article alludes to FastCGI etc. as alternatives but they're still workarounds for a fundamentally inefficient architecture.

New comment by Retr0id in "A Modern Web App in C"

Retr0id — Tue, 13 Oct 2026 12:14:59 +0000

I don't think "computers are faster now" (and the unspoken "LLMs write my code for me now") is a great reason to be using an inefficient design.