We’ve been searching for a memory-safe programming language to replace C++ in Ladybird for a while now. We previously explored Swift, but the C++ interop never quite got there, and platform support outside the Apple ecosystem was limited. Rust is a different story. The ecosystem is far more mature for systems programming, and many of our contributors already know the language. Going forward, we are rewriting parts of Ladybird in Rust.
Porting LibJS
I used Claude Code and Codex for the translation. This was human-directed, not autonomous code generation. I decided what to port, in what order, and what the Rust code should look like. It was hundreds of small prompts, steering the agents where things needed to go. After the initial translation, I ran multiple passes of adversarial review, asking different models to analyze the code for mistakes and bad patterns.
“We want rust for safety. That’s why we made the code in the most unreliable way to code known to humanity”.
Not surprising from the devs that think that not being a fascist is political I guess
Edit : ah apparently the guy is not trying to be an enlightened centrist anymore and is a full-on nazi. Heh, who could have guessed? /s
I will copy my comment here:
Just so you know, Drew DeVault is a pedophile apologist. He cannot even remotely be considered a reliable source of information.
However, I completely agree regarding the use of AI, especially in a project such as a browser engine.
EDIT: I accidentally replied to this comment instead of the one with the link to Drew DeVault’s blog. My bad.
That is appalling, and I don’t mean to diminish the importance of this, but if the claims on fascism stand to scrutiny then it’s not relevant, and if the claims don’t stand to scrutiny they can be easily refuted.
Edit: and to be clear, I know DHH, I’ve read his blog and his social media activity, he’s clearly a fascist.
I am completely disappointed in Ladybird, even though I had some hope for the project until now. I guess all that’s left to do is wait for Servo.
Your final hope at this point is announcing you will not use LLM to write any part of your program, including tests. Adopting a language which your developers are less familiar is bad enough, and you’re writing it with LLM? I’d rather you stick with C++.
The requirement from the start was byte-for-byte identical output from both pipelines. The result was about 25,000 lines of Rust, and the entire port took about two weeks. The same work would have taken me multiple months to do by hand. We’ve verified that every AST produced by the Rust parser is identical to the C++ one, and all bytecode generated by the Rust compiler is identical to the C++ compiler’s output. Zero regressions across the board:
Seems like a non-issue.
The Lunduke vid about this is going to be lit