Hacker News: rcxdude

New comment by rcxdude in "The adder at the heart of Intel's 8087 floating-point chip"

rcxdude — Sat, 13 Jun 2026 21:45:01 +0000

There's a surprising amount of optimization possible in them. You can improve the latency of them substantially at the cost of a lot more transistors.

New comment by rcxdude in "GameBoy Workboy"

rcxdude — Sat, 13 Jun 2026 21:40:10 +0000

The style is also very LLM-y, though not a complete slam dunk.

New comment by rcxdude in "There Is Life Before Main in Rust"

rcxdude — Sat, 13 Jun 2026 21:36:31 +0000

But it can cause problems when there's multiple crates trying to do it. If the call is explicit, the application developer can sequence the calls appropriately (or at least deterministically), as opposed to having the order determined by details of the implementation (something that was learned from the C++ 'static initialization order fiasco')

New comment by rcxdude in "Every Frame Perfect"

rcxdude — Sat, 13 Jun 2026 19:50:00 +0000

If the animations are effectively 'cancellable', i.e. they don't block input or delay the change in state, this can be reasonable. You can put in a sequence of actions into a UI at a much faster pace than 100ms, if you have the muscle memory for it.

New comment by rcxdude in "MTG Bench: Testing how well LLMs can play Magic"

rcxdude — Sat, 13 Jun 2026 10:42:48 +0000

only if there's a player choice in the loop. If there's a mandatory infinite loop the game ends in a draw.

New comment by rcxdude in "Statement on US government directive to suspend access to Fable 5 and Mythos 5"

rcxdude — Sat, 13 Jun 2026 10:05:57 +0000

if it's open source there will be many potential providers, though.

New comment by rcxdude in "Tailwind and slop apps"

rcxdude — Sat, 13 Jun 2026 09:46:27 +0000

I would personally like a little more consistency and a little less creativity in my UIs, TBH. Most UIs should be boring and effectively invisible, because if I'm paying attention to the UI, it's a distraction from the task or content that I actually care about. The web and electron have hastened a trend which I intensely dislike of every single application looking and acting completely differently for what is fundamentally the same thing. It's not that this variation doesn't matter: it's actively bad. (Not that UI design doesn't matter. Good UI design is important. But part of what makes UI design good is consistency and that's what web designers seem to actively dislike)

(more precisely, I long for the days of a standard and customizable UI toolkit. I should be the one who can adjust what the UI looks like, not hundreds of different designers with different concepts of how to 'stand out')

New comment by rcxdude in "Reading for pleasure is sharply down among schoolkids, report shows"

rcxdude — Fri, 12 Jun 2026 17:20:07 +0000

For me it made it a little dangerous to read: I hyperfocus on the book and can't get anything else done until I've finished it.

New comment by rcxdude in "AUR packages compromised with Infostealer and Rootkit"

rcxdude — Fri, 12 Jun 2026 13:47:28 +0000

It's only relatively recently that this has shifted from the norm. Debian operated this way for a long time and it was only in 2019 that they forbade it entirely.

New comment by rcxdude in "Grit: Rewriting Git in Rust with agents"

rcxdude — Thu, 11 Jun 2026 19:33:38 +0000

You misunderstand my point: the LLM is not a losslessly compressed version of the text: you need to supply additional information from the original in order to 'extract' it from the LLM (and from that point of view, the extra information would be the compressed form).

New comment by rcxdude in "More AI-generated code doesn't make your team faster. It might slow you"

rcxdude — Thu, 11 Jun 2026 14:55:46 +0000

I think they were being sarcastic.

New comment by rcxdude in "Grit: Rewriting Git in Rust with agents"

rcxdude — Wed, 10 Jun 2026 14:49:22 +0000

That paper is basically using the LLM as a compression algorithm: it's prompting with some section of the book and it's reprompting if it doesn't give the right output. Notably this only works if you already have a copy of the book in question!

New comment by rcxdude in "Notepad++ Zero-Click RCE via Path Traversal (CVE-2026-52884)"

rcxdude — Wed, 10 Jun 2026 11:34:48 +0000

This is a little bit 'if you can execute code as a user you can execute code as a user'. All of the exploit pathways involve capabilities that would give you any number of paths to code execution. The check should probably be fixed but I question whether it's really doing much in the first place.

New comment by rcxdude in "Mercedes‑Benz starts large‑scale production of electric axial flux motor"

rcxdude — Wed, 10 Jun 2026 11:21:13 +0000

Efficiency and cost savings at scale usually involve an increase in complexity: in mass manufacturing, complexity is generally a fixed cost and so can be amortized over larger volumes.

New comment by rcxdude in "Grit: Rewriting Git in Rust with agents"

rcxdude — Wed, 10 Jun 2026 09:04:08 +0000

The people doing the intellectual work are usually not the primary beneficiaries of IP laws. In fact it often constrains them unnecessarily.

New comment by rcxdude in "Grit: Rewriting Git in Rust with agents"

rcxdude — Wed, 10 Jun 2026 08:58:23 +0000

>Serious such rewrites don't start with the code of the closed game!

No, but they often involve reverse engineering the binary pretty heavily.

New comment by rcxdude in "Grit: Rewriting Git in Rust with agents"

rcxdude — Wed, 10 Jun 2026 08:56:43 +0000

black-box/clean-room isn't necessarily required, though. It does make it a lot harder to argue in court, of course.

New comment by rcxdude in "Grit: Rewriting Git in Rust with agents"

rcxdude — Wed, 10 Jun 2026 08:54:21 +0000

> Like, you can almost certainly get these LLMs to output gits full source code with some prompting, so there's not that much difference (as much as we like to pretend that AI generated code has no copyright implications)

Are you sure? LLMs are in some way a compressed version of their input but it's a pretty lossy compression (arguably this makes them more like a compression algorithm than a compressed version of the data). I'm not sure you can prompt a full, accurate, copy of a nontrivial codebase out of them. Even with zero temperature their accuracy is just not that high.

New comment by rcxdude in "Corrupting a ZFS File on Purpose"

rcxdude — Tue, 09 Jun 2026 23:02:22 +0000

It's not the word choice, it's the whole tone and structure of the sentence. It reads like a horror writer building up the tension before a big reveal but it just keeps drawing it out over a whole article and for something that isn't worth the build-up. It gets quite tiring to read IMO (LLM writing in general tends to have a grandiosity to it which really grates with something which is meant to be more informative, in my experience. They will explain a section of tax law like it's the second coming of Christ).

New comment by rcxdude in "Microsoft's open source tools were hacked to steal passwords of AI developers"

rcxdude — Tue, 09 Jun 2026 16:46:05 +0000

Embedded software already has a pretty strong culture of rarely using libraries and vendoring them if they do (for better and for worse). This kind of worm just doesn't really make sense in that kind of environment anyway.