Funny to think that LLMs could finally achieve linux on the desktop. Writing your own software is finally worthwhile for the average user, so open apis and rich open source training data is a competitive advantage.

for all my philosophical grievances with having LLMs code for you, the main reason it surprises me that people do it so much is that Claude is just kind of stupid

in a win for open protocol applications, WebDAV (underlying contacts and calendar sync) is finally moving beyond being polling-only

AI codegen is walking us backwards into literate programming. Instead of jumping directly to code, I now dump my half-baked ideas into md files in a "specs" folder and let the agent rip. Or write a comment describing the function I want and AI completes it. My code has better docs and tests.

Cursor/Windsurf's "pair programming with the AI" approach is far better. Yes, you cannot generate an entire project in mere minutes this way, but it's still VERY fast, and you have human eyes on the code developing a real understanding for all the moving parts.

You can now apply for tickets to @protocol.berlin Ticket's will be distributed on a "first come, first serve" basis for all eligible applications. And as we always like to say, tickets are free, as in free lemonade! Looking forward to seeing you all in June tickets.protocol.berlin

LLMs That Don't Gaslight You A new language model uses diffusion instead of next-token prediction. That means the text it can back out of a hallucination before it commits. This is a big win for areas like law & contracts, where global consistency is valued timkellogg.me/blog/2025/02...

The more I think about "functional" components, the crazier they seem to me. They try to combine: - One-time initialization - Instance-local state - A repeatedly called render callback - Other lifecycle callbacks into what looks like - but definitely *is not* - a pure function.

All told, it's been a year and a half of hard work to get to today. We've got so much to share about this project. Let us know what you'd like to hear about! :) youtu.be/0mCsluv5FXA

it's already here and it's called lean 4

writing code with LLMs is fun because you get to hear "You're right" and "Good catch" all the time as you tell it the mistakes it's made, instead of the other way round when the compiler tells you what you did wrong

there may be a category of questions, though i am highly doubtful that there is, for which stackoverflow answers are worse than for CSS

Latest video from us. Might be the most effort anyone has ever put into explaining an impossibility proof www.youtube.com/watch?v=_VAA...

not to be a contrarian, but whenever i find myself agreeing with the top comment on hacker news i want to interrogate that

writing proofs in lean feels just like coding except your entire program is a noop and your only job is to make it typecheck

do not use <Suspense> in React 19 if you want your app to be snappy. it will take a minimum of 300ms to resolve. github.com/facebook/rea...

i keep coming back to this by @bobaekang.bsky.social (and the linked post by @tannerlinsley.com) RSCs are the 'easy' way out here for the React team, but hardly the 'correct' one... bobaekang.com/blog/compone...

TIL that React throws away useRef (and all other 'stable' hooks, but this tree is the most egregious) references on suspend, which is yet another instance of this pattern

firefox is kinda busted but every time a website is actually broken in functionality it turns out to be the same in chrome

As of yesterday it has been longer since @solidjs.com re-introduced Signals and Fine-Grained Rendering back into modern web dev than the gap between the initial releases of jQuery and React.

i like this POV on bookmarklets and extensions as "interventions" for user agency in a user-hostile browser world: andregarzia.com/2025/02/reta...

haskell is a low level language send tweet

react is the "worse is better" of frontend frameworks (which is also why it won, and why it won't remain dominant, like others in this vein): a simple model that offloads most of the complexity onto the user

new (nix) blog is up! please stop using `nixos-rebuild switch`. there's a better way! aly.codes/blog/2025-02...

it's wild to me that after all these years responsive design is still a third class citizen in CSS. why are people letting themselves be mollified with flexbox and container queries

Wrote a thing about some things I've learned about software development over the years qntm.org/devphilo

The Patron Saint of Lemon Fairies

someone's working on making docker containers runnable as webassembly

we’re in a funny transitory period for ai codegen where you can often get away with not being able to write code, but you still have to be able to read it

this is great www.cs.cmu.edu/~pavlo/blog/...

What would happen if we had a way of defining signals directly in HTML?

This has been a long time coming! Have you ever wondered what lenses and UI programming have in common? Can we recreate React, Elm, and Redux entirely using pure category theory? Read this series to find out. cybercat.institute/2025/01/21/u...

TIL: Lightweight multitenancy for server-side JS → github.com/pdubroy/til... I did some research into how various cloud providers are doing lighter weight isolation (i.e., not using containers) for server-side JS. Comments/corrections welcome!

Flutter powers more apps than React Native: but more iOS apps are RN than Flutter. Large-scale case studies published are mostly RN. Flutter case studies are usually smaller apps. More details on each technology: newsletter.pragmaticengineer.com/p/cross-plat...

this has been in the work for *months* at @expo.dev 😄 but we're finally launching EAS Hosting! this is the first of many steps for building out Expo web, and also sets the groundwork we need for SSR-support and React server functions & components 🚀 expo.dev/blog/expo-an...

in which RSCs could be good but Vercel is economically disincentivized to cache them properly

“Breaking in” to tech is hard, but staying and growing in tech is harder. We never really talk about it, so I wrote a long blog post sharing why i think it’s hard and how to navigate tech. dev.to/blackgirlbyt...

🌠 A long-requested CSS feature is the ability to transition height between a known value (`height: 20rem`) and a derived one (`height: auto`). A new global CSS property makes this possible now. 😄 html { interpolate-size: allow-keywords; } More info in thread 🧵

So happy to see the results of the work on @servo.org during the last couple of years to bring the project back to life. This effort has been lead by @igalia.com with the help of the wider Servo community. Thanks everyone! For more details check my last blog post: blogs.igalia.com/mrego/servo-...

splitting the difference in that HTML is definitely not a programming language but CSS 100% is

One programming bitter lesson that Factorio really drills into you is that efficiency and scalability and fundamentally at odds. Maximum efficiency exploits the local structure of a context, and that structure changes when you make the context bigger

opening up inference scaling with @nousresearch.com's Forge

I'm involved in tech hiring. I've interviewed thousands of engineers. And I'd like to tell you about a disturbing trend I've seen in tech in the last 10 years 🧵

great thread, did not think about it like this before

i wodner when if ever scale-agnostic programming (run this in a function vs run this on a thousand machines) will become a thing. everything i've seen so far either just presupposes scale, or happens to use a runtime that scales well. in an ideal world the business logic would be fully separated....

oh crap. i missed that @cloudflare.social is planning a container offering... this will be huge, making their platform almost a no-brainer for anything web

i think i get it now. everyone looks at React Server Components and sees post-initial-load server rendering (because their current setup doesn't give them good client-server DX), while the important story is actually that they approximate partial hydration...

found a very cool software that makes it super easy to create RSS feeds for web sources that don't offer them docs.rsshub.app

Here's my end-of-year review of things we learned out about LLMs in 2024 - we learned a LOT of things simonwillison.net/2024/Dec/31/... Table of contents: