I would argue that if you rewrote the same application twice, first from X to Y, and later again from Y back to X, the final X rewrite will outperform the original X version.

During these migration, we also revisit the architecture, remove accumulated tech debt, adopt newer best practices, and, especially, put more emphasis on performance. So, it's no surprise that this new version with framework Y performs better.

I'm very skeptical of claims like "We migrated from framework X to Y, and metric A improved by…" In most cases, these migration stories over-credit the framework and undercut everything else that changed during the rewrite.

these translations are so bad, but also looks like a bug (?), since the translation of "la eterna canción" is switched with the translation of the last phrase, "Ni tiene perdón"

useActionState – React The library for web and native user interfaces

useActionState is a new hook for forms, async reducers, and more. It now has great new documentation: react.dev/reference/re...
Thank you to @ricky.fm for writing.
Thank you to @aurorascharff.no, @danabra.mov, @samselikoff.com, @stephanjnoel.bsky.social, @brenelz.com, and @yadiel.dev for reviewing.

woah! I love the animation on the first page

What a night to be Puerto Rican 🇵🇷

I’m biased, ofc, but that was the best halftime show ever 🔥 #SuperBowl

typesafety in Python is a mess, TypeScript is way better

😩

I was rewatching Ryan Carniato's stream you were featured in, and it's really cool how all the pieces connect, especially with the ViewTransition stuff.

thanks! this helps a lot. can't wait for libraries to start baking this stuff in, especially component and form libraries

also, I checked the profiler tab and it re-rendered two times, the first render triggered by the setOptimistic fn has a "immediate" priority, and the setState has a "normal" priority, pretty cool

okay, it's clicking...

continuing with the controlled inputs example, we don't need to wrap the setState in a nested transition, right? Unless I have to await a promise for data, then the setData would be in a nested transition.

I noticed that I got practically the results in the scheduler track

yeah, maybe text inputs are not the best examples, but I get what you're saying, thanks 🙏

we're doing animations with JS libraries because it's way easier, I bet that is one of the possibly many reasons

I have doubts about my mental model because don't we want that state update to have the highest priority? 2/2

I'm not sure if I have my mental model about transitions 100% correct, but let's say that I have controlled text field, does "everything should be in a transitions" also mean that the state setter for that controlled input should be wrapped in a transition? 1/2

thank you! this makes perfect sense

what do you mean by "except useOptimistic"? not sure if I understood that part

There're even languages with 5 or more grammatical gender. Gender in this sense doesn't always mean female/male, it can also mean animal, or plants 🤯
You might find this interesting: wals.info/feature/30A#...

0 spots left 😕

remove synthetic events

I'm not sure explanations for these terms really belong on react.dev. I've actually been thinking about starting a dedicated site that explains all of these concepts in one place, so frameworks and libraries can simply reference it. I even picked up a domain for it 😂

Hand them a messy codebase, a bug, a feature request, and an AI agent.

The real signal is how they explore, iterate, and collaborate with the tools that actually ship product.

hot take🌶️: An engineer who can’t use AI slows the whole team down (specially in early startups), and hiring should reflect that.

Modern dev interviews should test how well someone works with AI, not how fast they grind leetcode.

Angular really nailed releases: a major every 6 months + codemods

i know bro

maybe controversial take: React is cutting edge

MacOS 26 is so fuckin buggy I can't work

I'm getting a PhD in Auth