Stephan

Deep Dive: Caching and Revalidating · vercel next.js · Discussion #54075 The Next.js App Router introduced new heuristics around caching and revalidating. To ensure we’re all discussing the intended behavior of how caching is designed to work, I thought it would be help...

There’s a deep dive here github.com/vercel/next..... Not sure how much has changed since it was written though.

I’ve been using revalidatePath in Next14 and it works fine. Docs for Next were a bit confusing as to what “triggering on next navigation” really means.

If you do a router.refresh you lose the single round trip revalidation benefit and it invalidates a lot more.

YouTube video by ThePrimeTime banned from github

youtu.be/7gCCXCSs734?...

@tangled.org new video, help Primeagen figure it out?

How can one get involved?

I’m subscribed and have been following along but it’s locked to invite/past contributors.

Now I want to know your zod opinions

JSON Schema | Zod How to convert Zod schemas to JSON Schema

I was thinking when I got back to at proto stuff I’d use Zod 4 with toJsonSchema(). zod.dev/json-schema. I guess that doesn’t work out?

Wow, that was fast. I’m in now. Thanks to both of you.

Thanks! I opened an issue here tangled.org/@smokesignal.... Not sure if that’s the right place. Let me know if I can help.

Unfortunately getting a “failure?message=openid_connect_sub_mismatch” when trying to sign up or login .

Thanks for the info, I’ll be there

Quick question. I’m kinda new to At Proto so bear with me, but if a PDS has ultimate discretion over what it hosts, how does that interact with the question of privacy, private data (possibly e2ee) and data ownership? Trying to figure out how data privacy is expected to play out eventuality.

I’m trying to think if/how I can help. Making issues/pull requests to the docs doesn’t seem like the best path (maybe it depends?).

Not sure if making a blog would help either.

If there’s some preferred way of helping, please let me know.

Hmm I’ll look more. I see it in the reference section, but feels like the reactwg discussions address the “why” a lot more, esp showing how you’d have to address things without the feature.

Real world example: adding startTransition for slow renders · reactwg react-18 · Discussion #65 In React 18 we announced a new startTransition API and shared a high-level overview of the problem it solves. In this post, we’re going to dive into a real-world example of speeding up a slow updat...

I found these for the first time today

github.com/reactwg/reac...

github.com/reactwg/reac...

My path was I did a google search and a Reddit thread mentioned them. Not sure how I didn’t find it earlier.

Looking forward to it!

I hope there was something useful here, it ended up more disorganized than I thought it would be.

I was thinking maybe a deep dive and understandings bit more about how these work under the hood of Fiber based on the concept of concurrent React might help. Or maybe seeing more production use cases? I’m not really sure.

Maybe it’s that I feel React’s approach to maintaining an api with a minimum surface area has shifted and I’m looking for something to help group/connect all these new features/concepts together in a coherent way to see how they’re part of a bigger vision/model instead of disconnected pieces.

I saw someone (maybe you?) mention you need to understand Suspense to understand transitions. I thought it was the other way around, and thought it would be interesting to learn more about that connection.

Where, when, how and and how often these new hooks are being used and expected to be used in production apps isn’t clear to me. useDeferredValue still seems niche, even more so with the compiler.

Like I know transitions are useful for submitting forms, and routers like NextJS use them internally but I’m having a hard time finding concrete use cases outside of that. You could add an action prop to design system components but should you and why/why not? Transition all the things?

I saw Ryan’s attempt to make a clipboard copy using async transitions for the pending state. Like async transitions as a general purpose way to coordinate and batch things async. It was interesting, but even though you could, I wasn’t sure if you should.

The “action” terminology (vs just client server functions) is still kinda fuzzy. I get it’s a state update or server function called in a transition, but that definition feels kind of like memorizing instead of understanding. I feel like I’m not getting the the “why” behind it.

I’m not sure I can put it in words well myself, but there’s a nagging feeling I have that I’m not understanding transitions on a deeper level.

Really interested in the concurrent stores mentioned in the context selectors RFC and how that might contrast to signals.

Also, a deep dive into transitions/actions, useOptimistic, Suspense, Activity, and Fiber, and how they all fit together. Trying to see the big picture of concurrent React.

Do you have any thoughts on local-first software or sync engines potentially helping here?

Zero Docs The official documentation for Zero by Rocicorp.

If you’re interested in why offline writes will be temporarily disabled in zero sync, you can find more here zero.rocicorp.dev/docs/offline. The referenced “Lies I was told about Collaborative editing” article is very insightful.

This was a great discussion about local-first vs sync engines, the relation between local-first and AI, and also the temporary disabling of offline writes in zero sync (upcoming day 2 talk by Peter will cover Patchwork’s approach I think).

Not saying it’s a valid reason to build your own or that there aren’t other possible motivations.

I feel like there are valid reasons not to use some solutions like Cursor and Windsurf, no? Microsoft employees mentioned this was a factor for them open sourcing Co-Pilot (despite some components being server side still). You’re giving your private code to yet another company that pinky swears.