Can we have a Wikipedia entry for these? per model? I wouldn't mind using this page for that

jazz hand(s) !

I love how they will always have *one metric that is lower than a competitor's model, like these metrics are reflecting usage.

ACP still exists, not sure why no one other than Zed is using it. Its best of both worlds, because you're using their CLI but in another tool

With the coming changes in June, ACP will charge towards the same budget as claude -p and the Claude Code SDK (since it uses the SDK), so ACP no longer solves this. It's (I think) why Zed added "Terminal Threads" [1] to their agent workflow

1: https://zed.dev/blog/terminal-threads

The ACP budget change is so bizarre to me. If i was more adventurous with my subscription i'd be interested to see if you could intercept UI/input from CC TUI and render that in a native GUI without it being a TUI. That would be "interactive Claude Code" but you'd get a programmatic interface.

But that would be banned almost instantly i'm sure lol.

Same dude that made jiff. Love that library, so I'm assuming biff is built on top of jiff.

also made ripgrep

And xsv. Burntsushi projects have certain quiet sensibility that I appreciate.

I only recently realized that xsv is now unmaintained. The author now suggests using qsv or xan.

So this is who I get to "blame" for LLMs constantly saying "I want to use `rg` but I see that it's not installed on this system...".

If you've got enough permissions on the device to install a local llm, go ahead and install rg. After a few decades of installing my pet toys on every *nix box I maintain, I've seriously scaled back. Yet I continue to install rg. I can think of only two other non-default Debian install tools that I use regularly, those are git and ncdu.

ripgrep is a required dependency of codex when installing it via brew on macOS. So you should at least not see it if you fall into that scenario. :-)

> someone

in this context, that someone is an AI bot that spat out words.

Well, you see, the house always wins

Does it? Unusualwhales already provides the same data with a bit of insight. They also trade via shell companies (used to just call them straw buyers).

Id say its just another version

Drive-by PRs have been an issue before, but with AI it's just getting disrespectful

So which language had it right from the start? is there a language that has a very low rewrite status?

I think Elixir is a good candidate here. It's small, coherent, and composes well, and (at least to my understanding) the authors consider the language finished, with no new major features planned.

Elixir is missing static types though, it's hard to go back to work on dynamic languages.

Hopefully, it's coming.

https://elixir-lang.org/blog/2023/06/22/type-system-updates-...

I'd particularly like examples of statically typed languages that "got it right" (since I love me my types)

Ocaml maybe? Multi threading didn't seem necessary and introduced the possibility of data races.

I'm not familiar with D, but Zig and Rust are well-known for continuously evolving.

Zig has the (in)famous "Writergate": https://github.com/ziglang/zig/pull/24329

And besides Rust's high count of RFCs, there are things like async (I'm not complaining about it, but its an obvious large-scale "change"), module system changes, etc.

(To be clear, I like both languages a lot. But I wouldn't call them slow moving or right from the start.)

D literally can't even maintain backwards compatibility between minor version updates not to mention a big part of the D community left when D reinvented itself with D2. Among languages it's probably the one that is constantly in a state of flux.

what's a big project built with D? I feel it gets mentioned a lot on hackernews but I've never run into any project using it in the wild.