After the AI Engineer conference, I'm thinking about some problems from AI models to overcome. I got this list from the top of my head. Anything missing? • continuous learning or better memory? • long-term planning • prompt quality resilience: different prompts shouldn’t alter the outcome cont

tradcoding

Designed by davesnx, implemented by opus oh god

this will be implicit in all we will consume, in fact, "Made entirely by humans only" is the flex

Tricks from React that feel between hacky and genius: - Suspense fallback replace in SSR (the complete boundary script) - Bounded references in server functions - Form action hydration with hidden inputs - well, the entire Form progressive enhancement - useId being deterministic

gif about ochre without much context

every developer I know is at their peak only great things can come out

a dark mode, but it doesn't show white images

🧠 Rethinking Pair Programming in the Age of AI Why "pairing with AI" is harmful to your team and what to do instead learnings.aleixmorgadas.dev/p/rethinking...

Let's have great documentation for your OCaml project. Markdown is not the excuse anymore I wrote a blog post about odoc's Markdown backend sancho.dev/blog/ocaml-...

I'm super close to finishing a Next.js-like experience for OCaml on top of Dream: file-system routing, server components, typed routes, and a CLI that hides most of the dune machinery with LSP, error overlay, hot reload, melange integration...

Some thoughts about AI-assisted coding and how it affects my brain. I'm more productive and more exhausted at the same time. marvinh.dev/blog/ddosing...

It is the era of finishing the side-project of the side-project and go to the deepest rabbit hole possible

I wrote a guide on how to implement your first parser with parseff. Should be handy! davesnx.github.io/parseff/gui... is there anything you would want to learn/know about parseff?

exploring the idea of making parseff faster, autoresearch suggested to remove effects entirely from parseff and it makes fused operations equal, but primitive operations are much faster

yep, that's me

publishing parseff v0.3.0 - Binary parsing (big and little endian) - UTF-8 with uucp and new ocaml stdlib apis - Parseff.locations () that give offset, column and line

Are you going to be in Miami between April 20/25th? Let’s meet at React Miami

AGI is here. Visionaries were right www.reddit.com/r/ChatGPT/c...

writing parseff is mega fun this is a parser for a key/value (with utf8 support)

OxCaml is the extension of the extension of a dialect of ML If we fork OxCaml to add mlx... ex of the ex of the ex of ml

a pull request out of the blue 10 minutes after I publish in ocaml discord... just feels nice

Parseff, a parser combinator library for OCaml 5

If you try opencode, it's really hard to go back to claude Even now that opus/sonnet won't work in oc. I just hope 5.4 codex and other models become better, and oc implements some of the cool features from cc

Parser combinators are great until your code becomes operator soup `>>=` `<|>` `<<=`

BREAKING (nothing is broken): I just deployed a new html_of_jsx documentation site build with mlx, yocaml (@xvw.lol) and ochre (for the syntax highlight, still wip). That's what's possible with doc-markdown and odoc 🎊

Every Haskell project is just an identity choice. Nobody cares about the code. The "written-in-haskell" IS the product. Nobody picks OCaml for a web app because they ran the benchmarks. They pick it because they want to be the type of person who writes OCaml for a web app.

I love how @bobkonf.bsky.social become the 2nd house for ocaml talks

I have a mutual who's a developer who teaches me functional programming for free and always puts a lot of effort into sharing open-source projects, talking about them but I feel a bit bad because nobody uses them. One day I went to their profile to see who they were and it showed me 'edit profile'

who said you can't have nice documentation in OCaml?

Wow Eliom + Wasm Of OCaml is released (thanks to the @ocsigen.org toolchain) ! discuss.ocaml.org/t/eliom-12-c... #OCaml #ocsigen #wasm

i didn't expect opencode would make me install nvim...

this is what claude code was doing before your silly prompt

people, I found the absolute best name I could possibly imagine for a reason/ocaml project... a syntax highlighter called ochre

Time to go for a run

such a great feeling when you find all your stack used in the wild

TUIs are hot

poc: vibecoding from the slopes

me (to my friends who don't understand ai): you should always validate gipiti's answers, security matters alot, blabla also me: don't even read the output by whisperflow, --dangerously-skip-permissions, no plan mode, accept any code change without reading, not even a in git, auto deploy to prod

zo.computer is the real openclaw + Mac Mini I just installed pi, so I have an orchestrator for 5 instances that I can access anytime and ask them shit like schedule a barber, but also, clone repos, push PRs and deploy to fly.io feeling the agi rn

Finally I had time to finish!!! react rules of hooks for reason-react components - feature complete - deployed in our monorepo - linting the effects, the memos, the everything github.com/ml-in-barce...

I just implemented context support for React Server Components in server-reason-react. unlike React.js where createContext/useContext are client-only, our OCaml implementation supports context in server components, including across Suspense boundaries.

pov: I just opened my first non-trivial PR into ocaml/ocaml

1M context window changes the game again, I spend 2 weeks "cleaning" up rules, moving some into skills, some into hooks... and that's kind of useless now? I'm not super sure, been using it for 4 hours, but the model doesn't get dumb anymore I need to test the limits, to know more

Native support for mlx in the ocaml-lap-server (syntax highlight, and all LSP features from ocaml in mlx)

Adobe buys semrush (our biggest competitor)

This is pretty sick React.cache was a missing piece for RSC to work well™ github.com/ml-in-barce...

Modular explicits are coming to OCaml 5.5 🎊🫰🥳🎉🎂 github.com/ocaml/ocaml...

At this point, Github should implement the same UX as cursor to annotate files. The attention to details is so well done with the @ menu to annotate files/past chats/folders... that's the only reason why I'm still using Cursor over claude code

I finally released v1.0.0 of a ppx that enforces the rules of hooks for reason-react - exhaustive deps in useEffect/useMemo/useCallback - order of hooks (hooks at the toplevel/not conditionally) github.com/ml-in-barce...