Bruno Pinto
@bpinto.com
Achei justo. Ele saiu muuuuito a frente por não ter seguido a pista. Se fosse justo passar por ali, então todos poderiam passar por ali e economizar 1 segundo da volta.
Era só ter freado e voltado na mesma posição que estava quando saiu da pista.
Era só ter freado e voltado na mesma posição que estava quando saiu da pista.
October 26, 2025 at 9:23 PM
Achei justo. Ele saiu muuuuito a frente por não ter seguido a pista. Se fosse justo passar por ali, então todos poderiam passar por ali e economizar 1 segundo da volta.
Era só ter freado e voltado na mesma posição que estava quando saiu da pista.
Era só ter freado e voltado na mesma posição que estava quando saiu da pista.
Ah okay, I thought the shim would actually be a simple wrapper to a throw promise, I guess my assumption is incorrect then :)
October 7, 2025 at 3:22 PM
Ah okay, I thought the shim would actually be a simple wrapper to a throw promise, I guess my assumption is incorrect then :)
What is your concern with the shim? I don’t understand the problems it may cause… I guess the type of obscure problems we wouldn’t want to deal with :P
October 7, 2025 at 3:11 PM
What is your concern with the shim? I don’t understand the problems it may cause… I guess the type of obscure problems we wouldn’t want to deal with :P
I’d love if it could show me only the most liked posts from my “following” list.
There are some people I follow that just post way too often.
There are some people I follow that just post way too often.
October 6, 2025 at 9:14 PM
I’d love if it could show me only the most liked posts from my “following” list.
There are some people I follow that just post way too often.
There are some people I follow that just post way too often.
Likely a minority opinion here. But I think you could delay the decision.
IF (big if) it happens that changes are needed to make RQ transition compatible and assuming it would require a major release, then I think it could be useful to delay ESM only so more apps could migrate to the new API/design
IF (big if) it happens that changes are needed to make RQ transition compatible and assuming it would require a major release, then I think it could be useful to delay ESM only so more apps could migrate to the new API/design
October 5, 2025 at 9:50 PM
Likely a minority opinion here. But I think you could delay the decision.
IF (big if) it happens that changes are needed to make RQ transition compatible and assuming it would require a major release, then I think it could be useful to delay ESM only so more apps could migrate to the new API/design
IF (big if) it happens that changes are needed to make RQ transition compatible and assuming it would require a major release, then I think it could be useful to delay ESM only so more apps could migrate to the new API/design
YES! I only saw the announcement a few minutes ago and I'm already in the process of updating our app to 19.2. 🎉
It's awesome that now that you all have built most of the foundation needed, new features are being released frequently. It must be a great feeling for all of you.
Congratulations!
It's awesome that now that you all have built most of the foundation needed, new features are being released frequently. It must be a great feeling for all of you.
Congratulations!
October 3, 2025 at 3:21 PM
YES! I only saw the announcement a few minutes ago and I'm already in the process of updating our app to 19.2. 🎉
It's awesome that now that you all have built most of the foundation needed, new features are being released frequently. It must be a great feeling for all of you.
Congratulations!
It's awesome that now that you all have built most of the foundation needed, new features are being released frequently. It must be a great feeling for all of you.
Congratulations!
And it has been released in canary. We are getting closer! 🎉
Maybe we get a release during the conference?
github.com/facebook/rea...
Maybe we get a release during the conference?
github.com/facebook/rea...
Ensure `useEffectEvent` implementation is available in Canary by eps1lon · Pull Request #34614 · facebook/react
Follow-up to #34610.
I forgot there's a separate feature flag not just the entrypoint like for <Activity>.
github.com
September 27, 2025 at 11:46 PM
And it has been released in canary. We are getting closer! 🎉
Maybe we get a release during the conference?
github.com/facebook/rea...
Maybe we get a release during the conference?
github.com/facebook/rea...
The first time I checked this I thought the middle tab “react compiler” was meant to be the faster version. But it’s actually the last tab, React, that shows the benefit of using transitions.
I wonder if more people are having the same misunderstanding and thus replying with million being faster.
I wonder if more people are having the same misunderstanding and thus replying with million being faster.
September 23, 2025 at 11:11 PM
The first time I checked this I thought the middle tab “react compiler” was meant to be the faster version. But it’s actually the last tab, React, that shows the benefit of using transitions.
I wonder if more people are having the same misunderstanding and thus replying with million being faster.
I wonder if more people are having the same misunderstanding and thus replying with million being faster.
Maybe useStore() is the missing hook that will allow useEffect() to be removed?
September 23, 2025 at 9:33 PM
Maybe useStore() is the missing hook that will allow useEffect() to be removed?
What if you created a draft blog post that you could share with him privately first?
Then, once you understand the concepts better (after getting his feedback), you could publish a blog post that shows both the correct understanding AND your initial misconceptions.
Then, once you understand the concepts better (after getting his feedback), you could publish a blog post that shows both the correct understanding AND your initial misconceptions.
September 20, 2025 at 8:19 AM
What if you created a draft blog post that you could share with him privately first?
Then, once you understand the concepts better (after getting his feedback), you could publish a blog post that shows both the correct understanding AND your initial misconceptions.
Then, once you understand the concepts better (after getting his feedback), you could publish a blog post that shows both the correct understanding AND your initial misconceptions.
I know 🤯
I was thinking that the `isPending` returned by the `useSuspenseHook` would, somehow, be aware of transitions that have started automatically by react-query.
And so `const { isPending } = useSuspenseQuery()` would still be part of the API to handle scenarios like this.
I was thinking that the `isPending` returned by the `useSuspenseHook` would, somehow, be aware of transitions that have started automatically by react-query.
And so `const { isPending } = useSuspenseQuery()` would still be part of the API to handle scenarios like this.
September 19, 2025 at 6:07 PM
I know 🤯
I was thinking that the `isPending` returned by the `useSuspenseHook` would, somehow, be aware of transitions that have started automatically by react-query.
And so `const { isPending } = useSuspenseQuery()` would still be part of the API to handle scenarios like this.
I was thinking that the `isPending` returned by the `useSuspenseHook` would, somehow, be aware of transitions that have started automatically by react-query.
And so `const { isPending } = useSuspenseQuery()` would still be part of the API to handle scenarios like this.
September 19, 2025 at 6:01 PM
I'm not sure if my thinking is right, but I thought about this:
The change you did on the PR was on the `shouldSuspend()` function. That function, when true, will cause a `throw fetchOptimistic()`.
Should we replace that `throw` with a `use` hook?
The change you did on the PR was on the `shouldSuspend()` function. That function, when true, will cause a `throw fetchOptimistic()`.
Should we replace that `throw` with a `use` hook?
September 19, 2025 at 4:08 PM
I'm not sure if my thinking is right, but I thought about this:
The change you did on the PR was on the `shouldSuspend()` function. That function, when true, will cause a `throw fetchOptimistic()`.
Should we replace that `throw` with a `use` hook?
The change you did on the PR was on the `shouldSuspend()` function. That function, when true, will cause a `throw fetchOptimistic()`.
Should we replace that `throw` with a `use` hook?
I'm still reading the codebase to see if I can reason about this, but one thing that has caught my attention is that if you click "refetch" on the react-query dev tools then the suspense fallback isn't displayed.
I don't know yet why that would be the case, do you have any idea?
I don't know yet why that would be the case, do you have any idea?
September 19, 2025 at 3:58 PM
I'm still reading the codebase to see if I can reason about this, but one thing that has caught my attention is that if you click "refetch" on the react-query dev tools then the suspense fallback isn't displayed.
I don't know yet why that would be the case, do you have any idea?
I don't know yet why that would be the case, do you have any idea?
I tried, but not sure this is the correct path since I'm not noticing any difference when running on the react-query PR version.
stackblitz.com/edit/tanstac...
stackblitz.com/edit/tanstac...
[Transition] Query Pagination - StackBlitz
Run official live example code for Query Pagination, created by Tanstack on StackBlitz
stackblitz.com
September 19, 2025 at 3:23 PM
I tried, but not sure this is the correct path since I'm not noticing any difference when running on the react-query PR version.
stackblitz.com/edit/tanstac...
stackblitz.com/edit/tanstac...
That said, I would double check if using Activity for this is what you want.
September 19, 2025 at 8:10 AM
That said, I would double check if using Activity for this is what you want.
I believe you would need to use a hook version and then do something like
<Activity mode={isMobile ? “visible” : “hidden”>
<ComponentA />
</Activity>
<Activity mode={isMobile ? “visible” : “hidden”>
<ComponentA />
</Activity>
September 19, 2025 at 8:06 AM
I believe you would need to use a hook version and then do something like
<Activity mode={isMobile ? “visible” : “hidden”>
<ComponentA />
</Activity>
<Activity mode={isMobile ? “visible” : “hidden”>
<ComponentA />
</Activity>
By the way, implementing it in a different way is fine, doesn't mean any alternative approach is worse. I am only using RQ as a reference here because that's what I'm used to, but someone coming from react-relay or other implementations could have a different idea.
September 18, 2025 at 8:36 AM
By the way, implementing it in a different way is fine, doesn't mean any alternative approach is worse. I am only using RQ as a reference here because that's what I'm used to, but someone coming from react-relay or other implementations could have a different idea.
I can't say anything about threads because that's beyond my knowledge, but regarding cache warmup, invalidation, etc. I think react-query has solved it in an elegant way?
It allows requests to be initiated (or fully cached) at the router level (anywhere) instead of doing at component render time.
It allows requests to be initiated (or fully cached) at the router level (anywhere) instead of doing at component render time.
September 18, 2025 at 8:31 AM
I can't say anything about threads because that's beyond my knowledge, but regarding cache warmup, invalidation, etc. I think react-query has solved it in an elegant way?
It allows requests to be initiated (or fully cached) at the router level (anywhere) instead of doing at component render time.
It allows requests to be initiated (or fully cached) at the router level (anywhere) instead of doing at component render time.
Or the challenges you are trying to solve.
September 17, 2025 at 10:23 PM
Or the challenges you are trying to solve.
Going back to the initial thread topic, Cache API, you mentioned it is hard which makes me think that the ManyKeysMap approach is not what you are envisioning.
Would you be interested in sharing details about the leading plan?
Would you be interested in sharing details about the leading plan?
September 17, 2025 at 10:22 PM
Going back to the initial thread topic, Cache API, you mentioned it is hard which makes me think that the ManyKeysMap approach is not what you are envisioning.
Would you be interested in sharing details about the leading plan?
Would you be interested in sharing details about the leading plan?
This is a small detail but wondering if naming these as “action” is usually okay or if you normally end up with onClickAction and alike names.
September 17, 2025 at 10:10 PM
This is a small detail but wondering if naming these as “action” is usually okay or if you normally end up with onClickAction and alike names.