Hacker News

Why wait? It’s been possible since forever. I wrote UI in OCaml eight years ago. https://github.com/kccqzy/drawing-guessing/blob/master/clien...

There's also Fable for F# but I believe this commenter is being sarcastic. Javascript is a common backend language.

A really great place to reach for this is Elixir/Phoenix now that the set theoretic type system is in place with LiveView. I have been doing some pretty large/complex codebases gradually working towards fully integrating working this way for the past 5 years and it has been great.

Similar attempts include Scalajs.

The general challenge becomes integrating the fractional front-end code written in your backend-language that compiles to JS with the rest of the JS ecosystem.

JaneStreet have a love of writing their own stuff from scratch so it doesn't apply to them but it might to you.

Hence most people end up with frontend-as-backend rather than backend-as-frontend.

> Finally! I was waiting for this to become possible!

Sarcasm?

Do you want to have a talk as to why JaneStreet wasn't built on JavaScript?

A framing from that podcast that helped me better understand the role of Bonsai is that it's really a framework for building incremental distributed state machines.

My understanding is that js_of_ocaml has been around longer and likely has a stronger ecosystem, but Melange was designed more specifically to play nice with the JS ecosystem.

I’m no UI expert - what’s wrong with the margins?

Jane Street made it's way into today's Wall St Journal for its quantitative prowess and hiring of Math PhD's. What was stunning to hear was that their 2nd quarter profits were a record $10.3 billion, nearly double that of Goldman Sachs and Morgan Stanley (much larger firms).

They're a trading firm that hires really clever people to use math to trade. Their strategies are, presumably, very complicated and ever changing, and secrecy about them is the point - you get bigger margins for longer if people don't learn about your 'one weird trick' (that and also that there is presumably another side to your trades losing out on upside, and they might not be best pleased to learn about it.

They’re a market maker

Print money?

Basically most of Jane Street's published code[0] via its dependency on Core.

[0]: https://github.com/janestreet/bonsai/blob/f31661450eb133fe89...

The main thing that Bonsai gets you is a nice programming model for making your UIs interactive. If you're mostly producing static reports (with maybe some basic interactive visualization widgets or something), it will not get you much over generating some variant on HTML + JS.

It makes more sense for applications where you have complex interactive logic in the UI, with a lot of different components that need to control and display some shared state.

There was this submission two months ago: https://news.ycombinator.com/item?id=48365904 . Basically they have a TUI frontend for the same thing. And the blog post says: "Bonsai_term would feel familiar to anyone who’d ever done web development here, and it had the huge advantage of being especially amenable to AI assistance. It was actually somewhat of a mystery to us how good the models were at writing Bonsai_term code, given how relatively obscure it is [...]"

Probably not, because your agent won’t have much info about this tool in its training set.

There's a full terminal implementation of Bonsai as well. I actually use it in my personal workflow these days to manage my contacts database! https://anil.recoil.org/notes/aoah-2025-9

Jsoo supports self-recursive and mutually recursive tail calls, but for general tail calls it would need to trampoline every single call, since the only major browser with proper tail calls is Safari.

Ocsigen and Eliom were started 2005 and are pretty similar, although I can't find any well-known projects using Eliom :/.

Presumably this has learned some lessons from it, so it's probably better.

>"This would of been cool back in 2014."

Pretty much. UI libraries are more or less meaningless now, and I say this as someone who spent the last decade building them. Given a proper data API to work against, you're a single prompt away from a fully tested bespoke lib that's more maintainable and has no dependencies. The weeks of work you saved by reaching for one before is gone, and the tradeoff you had to make of dealing with upstream changes and lack of customization just isn't there anymore.

And there's nothing wrong with that. Many languages attempt to do the same thing. Why bother with 15 different languages if you can leverage one?

See also LiveView and Hologram for Elixir.

Agreed, looks like it's reinventing the wheels. Frameworks does not really matter today, AI agents write the code anyway.

Been here 5 years, never saw it

HN is obsessed with anything done in niche programming languages, is my guess.

Windows and especially Macintosh UIs from the late 90s/early 00s were by far the most functional UIs we've ever had since the dawn of the desktop era - user research was a serious discipline, accessibility was baked in to the frameworks, and the HID guidelines that devs adhered to gave us uniform conventions that we users could rely on.

If the biggest criticism of Bonsai is "it ain't pretty but it works", then long life to Bonsai - give me more, please!