Seems to be specifically a desktop containers runtime for Mac written in Rust. It uses Nftables for networking on linux, too, unlike docker which is really nice.

Looking at the codebase this is either not majorly AI generated or the spent a lot of time cleaning up the tells I recognize, so I'd say not slop.

almost 200K lines of code materialized in about 2 weeks, and the code is full of em-dashes. if this wasn't vibecoded, they sure tried their hardest to make it look like it was...

> 200K lines of code

And who is going to read all that to make sure it doesn't contain malware?

I'm extremely loath these days to casually clone any repos from untrusted sources.

This problem has preceded llms by at least a decade

But you couldn't have the code generated for you then. That friction is gone now.

If you have enough tokens, a LLM will happily read every line. That’s how we are all getting inundated with bug reports now.

> code created by an AI is not copyrightable and it’s output is in the public domain

This isn't settled yet, and while the license may ultimately be meaningless, it lets people compile and use the code without risking copyright law violation.

I do agree about AI washing though, ethically any code you generate using AI should be licensed as similarly to the source material as possible.

A USCO opinion, and then two more confirmations of that opinion, one before Congress, is not settled? Unless the law changes the only thing licensable is the prompts.

The opinion was only for fully ai-generated content, notably the human-authored components of the comic were still protected. Whats unsettled is when the artifact is built up over multiple rounds of targeted edits to existing output in collaboration with a human, and whether that output is protected.

I'm guessing there's too much money in that being protected for it not to be protected so, no I think providing LICENSE is still worthwhile.

USCO doesn't make the law nor do they decide what the law says, they just register Copyrights. However, registration is not necessary for a work to be protected under Copyright law.

Congress makes the law then the Court interprets it, which hasn't happened yet.

If choosing the position / angle of a camera and its settings is enough to gain Copyright over the images it generates then prompting a model to generate an image is enough too. I don't see how anyone can come to any other conclusion.

The creativity bar for Copyright is not high. An alphabetically ordered list of customers does not meet it. Pointing a camera at someone and pushing the button does.

Congress has made the laws and USCO interprets them and makes rules based on its interpretation. Those are the implementation of those laws.

Directing is not copyrightable. The person who takes the picture owns the copyright (fixation rule) unless some other contract says otherwise. Treating the prompt writer as a director reduces their claim to ownership.

(Edit to to clarify interpretation of parent)

There was a thing, but it didn't happen often, where a GPL advocate would take an MIT or BSD license project, make some additions, license their changes under the GPL, and effectively copyleft their release of the combined work.

I'm not a lawyer, but I don't think the law is simple.

Code exists in the Linux kernel that is licensed under permissive licenses. This works because (some) permissive licenses contain no restrictions incompatible with releasing the code and/or derivative works under the GPL.

This is why the original BSD license (with advertising clause) was incompatible with the GPL, while BSD without the advertising clause is not.

"Public domain" is not a license, and has different meanings in different jurisdictions, and thus it can be useful to release code under a license even if the intent is unrestricted use.

That doesn't justify its deletion, it might justify warning the user, but unconditionally deleting data not owned by the app shared with other apps?

I didn't read it as dismissive. I think the "Why doesn't this work?" at the end is key, implying they know they're probably wrong, but don't know why. It's pretty common to form and present hypothesis like this, hoping that anyone who knows the actual theory can easily provide a counterexample showing why it's wrong. It's not at all intended to be dismissive.

Its not enabled by default, but enabling MGLRU setting /sys/kernel/mm/lru_gen/min_ttl_ms to 70 has entirely resolved the issue for me. It _does_ usually pick firefox or vscode to kill, but my system remains interactive, with no need for earlyoom.

Docs https://docs.kernel.org/admin-guide/mm/multigen_lru.html#thr...

Usually its me who has been committing with a bogus email address, because I don't like having my real user/email set in my $HOME/.gitconfig

Hi there, webstrand@Webstrands-MacBook-Pro.local - we meet again!

ES6 class syntax is still mostly just syntax sugar overtop prototypical inheritance.

JS _does_ still have TCO (called Proper Tail Calls), Safari's JavaScriptCore implements it, and is technically the only conforming interpreter.

Yes but the syntax encourages patterns which would be uncommon in pre-ES6 JS.

I can’t rely on TCO if chromium doesn’t have it.

You're in luck <https://hnrss.org/threads?id=SirHackalot>

I also do this, except 7.0.0.0/8 instead. Its great for not conflicting with hotel wifi dhcp.