The Useless UseCallback
tkdodo.eu
tkdodo.eu
{} !== {}
and that it uses reference equality as a proxy for value equality.
This is why the absolute simplest thing to do is to just memoize all the things (see https://attardi.org/why-we-memo-all-the-things/ for a deeper dive into this approach).
I know that the conventional wisdom (e.g. https://kentcdodds.com/blog/usememo-and-usecallback) is not to do that, but it's worth noting that Kent doesn't actually bring receipts for his claims, and also the discussion is typically limited to toy examples, not huge applications that have the potential for huge cascading renders when anything changes. In my own profiling of real world applications I have found that memoizing everything is not actually a performance regression. And then also consider the React Compiler, which of course does the same, at an even more fine-grained level than any human could be bothered to consistently do by hand.
This creates a massive gulf between beautiful simple frontend toy examples and the reality of applications made with those same frameworks when deployed for multiple years and product cycles. I think it is a big part of what fuels shiny framework adoption. You think: "yes, this here is what will resolve the complexity and pains I see in my work codebase", and this is bolstered by your carefully curated personal side projects.
One example could be a form where the validation rules of one field depend on another, where you have to pick an option from a list that is loaded from a server (what do you do when the current value is not among the returned list?), etc.
A bunch of `useCallbacks` are rendered ineffective because the top of the chain wasn't properly computed. And the fix isn't `useCallback`, its to rearchitect the tree. This feels wrong - either the API, the framework, or the language isn't powerful enough to ensure that someone does the right and performant thing. It seems like the new React Compiler tries to fix this - but I'm surprised React Compiler is a thing that needs to be built.
I lack the expertise to be able to go in and rip it all out, and while I can share articles like this and others such as "You might not need useEffect", there isn't a really obvious way to deal with it all, especially if there are places which genuinely could do with proper memoisation.
Now I actively go out of my way to avoid doing any fullstack because I spend more time trying to fight against the various fragile towers of incantation-like hooks than just implementing the business logic in a way that feels readable and maintainable. I'm not sure there is a way to remove the complexity either, but so much of modern programming with React just feels unintuitive and convoluted, which is exactly the opposite of how it used to feel when it first came out. I'm not sure how to solve it.
I use React (actually Preact), but the old way, not with "use" anything, no hooks. It's pretty simple, but does a good job at what I need, and I'm doing pretty complex web applications with it.
There's no law saying you must use the bleeding edge of everything, the old React way is still plenty good for most things.
The way they rushed out hooks and everyone jumped on board was a red flag for me. I've used it, I've built stuff with hooks, but it's always seemed like fixing problems by creating more problems, and the hype around it was ridiculous.
I'm speaking more generally about the trend of chasing "new, shiny" in front-end. Some people knee-jerk and just go for the "new, shiny" thing as soon as it happens. And other people still use the existing tech, because it's good enough.
Hooks are a good translation layer.
Footguns, learning curve? Sure, but I'll still make that trade.
React in its current incarnation is the best framework to create unmaintainable code.
That’s an antipattern. Like the 200 lines function in imperative or the god class in OOP.
Your components should be small and composable. Before directly using the standard hooks in the components, you should think first if it would not better to write a custom one. Scenarios should be handled by custom hooks and component should just get data from them. If you cannot test scenarios apart from your views and vice-versa, you’re doing it wrong
useState
useState
useState
... // 20 more
useEffect
useEffect
useEffect
... // 20 more
I think a lot of front-end devs don't realize the necessity of custom hooks for composition and modularization. My personal rule is "Never use react hooks (useState, useEffect, etc.) in component code. All react hooks MUST be wrapped by custom hooks." and it makes for MUCH better code:
useSomeData
useOtherData
useMobileView
useScrollToTop
useInteractionLogging
... // 10 more hooks
This is also what makes TanStack so good and popular. It's all just behavior wrapped in hooks.
It's the language. React uses a bunch of magic methods and requires you to follow rules about when to call them, with limited guardrails, because if you try to make developers do this properly (with higher-kinded types and monads) they throw their toys out of the pram. Our culture prefers "there are a bunch of subtle rules you have to follow and if you break them you'll get weird bugs or performance issues" to "there are a bunch of explicit rules you have to follow and if you break them your code won't compile", sadly.
Before that you could write:
class Page extends React.Component {
componentWillReceiveProps(nextProps) {
if (nextProps.url !== this.props.url)
logVisit(nextProps.url, this.state.nItems)
}
}
const Page = ({ url }) => {
const onVisit = useEffectEvent(visitedUrl => {
logVisit(visitedUrl, nItems)
})
useEffect(() => {
onVisit(url)
}, [url])
}If you really wanted to take a jab at React people, show them
const that = useLatest({
// bunch of stuff you want to pass around to effects / callbacks
})It’s not complicated at all.
It’s quite easy. The complaints people have about it are due to friction from using it wrong.
And keeps in mind, react hook until 18 (a long time since hook introduced) didn't offer a official way to connect component to external changing data. The only way you can do is by using useEffect incorrectly.
You mean correctly. That’s was what it was for. Syncing with external systems.
If your large, complex, react app is slow - what is easier? Fixing the data architecture or adding a useMemo? Multiply by that by 1000 and you have a nightmare. The bad thing/escape hatch is vastly simpler than the proper thing, because something about language/framework/api makes doing the proper thing hard or unclear.
The language is sort of settled. The browser runs JS, so for most projects, it's the right choice.
> If your large, complex, react app is slow - what is easier? Fixing the data architecture or adding a useMemo?
It might be easy to add useMemo, but it doesn't work. People using a tool for the wrong thing because they don't understand it isn't a problem with the tool. It's unfortunate that a tool with a steep learning curve is the "default" for many devs now, but again that isn't a problem with the tool. I don't have any guardrails keeping me from using my saw as a hammer, it just doesn't work.
> because something about language/framework/api makes doing the proper thing hard or unclear.
It's unclear because it was built on functional programming concepts and most devs aren't taught that way. If you are familiar with those concepts, the whole thing is quite simple.
If the tool is easily misused, the tool is wrong. i.e. "You're holding it wrong" is a problem with the phone, not the user.
>The language is sort of settled. The browser runs JS, so for most projects, it's the right choice.
TypeScript exists. Something could be made that compiles to JS that makes this easier.
There are more opinionated UI frameworks (angular) and frameworks with less magic baked into them (svelte, solid). React gives you the freedom to architect your app the way you want, for better or for worse. It was made by functional programmers to allow them to use composition and other FP paradigms in UI development. If you don't understand those concepts, or you just don't care for them, you're going to be swimming against the tide. Lots of tooling has been created to help those who do (Redux, for example).
Maybe you're right and "you're holding it wrong" isn't a valid answer, but it's not React's fault it took over the world. It was made to do a certain thing, and it's pretty good for that. That thing requires a way to create side effects (useEffect) and a way to provide referential stability (useMemo, useCallback). The React team has done a lot to document the right way to use these tools.
If it's not someone's thing, that's cool. If they are forced to use it at work, might as well learn a little FP and make life simpler.
And out of them, the react hook is the only one that 'doing it wrong' is easier. A good API interface should have least resistance for the user when doing things correctly. But React Hook almost feels like the complete opposite of this phenomenon. It's complex, comes with tons of foot gun. And yet still get popular?
Wondering if there is something other like this exists in the computer history.
I've got pretty sizeable codebases, less than 5 useEffects ever (nowadays I only have one for a single location where I need to manually poll, the rest is useSyncExternalStore) and I only do useMemo for tables.
Compiler is just an additional performance improvement that's "free" and at least the compiler doesn't make it harder to understand how your code executes because it just works
Unfortunely, it has won the mindshare game, in many SaaS products I tend to work on, they only provide React and by extension Next.js based SDKs, if I want to use vanila.js, Angular, Vue or whatever, it is on me to make it work, if it is possible at all, given the product expectations on its extension points.
Eventually the "everything must be React" wave is going to pass, hopefully.
I think this is wrong? If you memoize a callback with useCallback and that callback uses something from props without putting it in the dependency array, and then the props change & the callback runs, the callback will use the original/stale value from the props and that's almost certainly a bug.
They might be trying to say you just shouldn't use useCallback in that situation, but at best it's very confusingly written there because it sure sounds like it's saying using useCallback but omitting a dependency is acceptable.
IMHO useCallback is still a good idea in those situations presuming you care about the potential needless re-renders (which for a lot of smaller apps probably aren't really a perf issue, so maybe you don't). If component A renders component B and does not memoize its own callback that it passes to B as a prop, that is A's problem, not B's. Memoization is easy to get wrong by forgetting to memoize one spot which cascades downwards like the screenshots example, but that doesn't mean memoization is never helpful or components shouldn't do their best to optimize for performance.
> I think this is wrong? If you memoize a callback with useCallback and that callback uses something from props without putting it in the dependency array, and then the props change & the callback runs, the callback will use the original/stale value from the props and that's almost certainly a bug.
The problem is up a level in the tree, not in the component being defined. See the following code block, where they show how the OhNo component is used - the prop is always different.
Most source of bugs happens when people ignore this simple fact. Input start from the root and get transformed along the way. Each node can bring things from external, but it should be treated careful as the children down won't care. It's all props to them.
Which also means don't create new objects out of the blue to pass down. When you need to do so, memoize. As the only reason you need to do so is because you want:
- to join two or more props or local state variables
- or to transform the value of a prop or a local state variable (in an immutable fashion, as it should be).
- or both.
1. Enforce that you're always memoizing everywhere always
2. Memoize every new function and object (and array) and pray everyone else does too
#1 is pretty miserable, #2 is almost always guaranteed to fail.
The mismatch is that you need stable props from the bottom up, and you enforce stable props from the top down. Any oversight in between (even indirectly, like in a hook or external module) creates a bug that appears silently and regresses performance.
> you need stable props from the bottom up, and you enforce stable props from the top down.
You don't need stable props from the bottom up. Component will always react to props change and will returns a new Tree. The function for that are going to be run multiple times and you're not supposed to control when it does. So the answer is to treat everything that comes from the parent as the gospel truth. And any local transformations should be done outside of function calls.
#1 is not needed. Memoization concerns is always local to the component, as long as you ensure that the props you're sending down is referentially stable, you're golden.
#2 is not needed. Again your only concern is the current component. Whatever everyone does elsewhere is their bad practices. Maybe they will cast your carefully constructed type into any or try/catch your thougthful designed errors into oblivion along the way too. The only thing you can do is correct them in PR reviews or fix the code when you happen upon it.
This is strictly false, because of effects. The bottom of your tree consumes props and triggers events. The author has plenty of examples of this, and even calls out useEventEffect as being helpful.
> The only solution is to have everyone have discipline and respect the convention of React.
The article specifically goes out of its way to show that mistakes still happen and linters aren't capable of solving this alone.
> as long as you ensure that the props you're sending down is referentially stable, you're golden
This is literally the point. You can't. As the component, you're powerless to ensure this. And components that use your component can't even be sure, because referential stability doesn't only require your code to be correct, but every third party library as well.
What I meant is that the API design of B, where the hook only works as intended if the consumer A memoizes the input props, is not ideal because A has no way of knowing that they have to memoize it unless they look at the implementation or it’s clearly documented (which it rarely is). It’s not A’s problem because B could’ve used the latest ref pattern (or, in the future, useEffectEvent) to work without passing that burden onto its consumers.
A second but unrelated grief with hooks is when you get async updates happening in between renderers where you want to update some state and then do some logic. Let's say I'm waiting for 2 requests to finish. They both should update a value in my state. However, to not lose an update if both happens at the same time, I need to do the setState with a callback to see the latest value. But then, what if I want to trigger something when that new calculated state value is z? I can't do that imperatively inside the callback, and outside it I can't see the new value and don't know that I now should do something additional. I have to do it in a useEffect, but that decouples it too much, the logic belongs where I want to put it, in my state hook. Not in some useeffect wherever. The other option is to misuse refs to also maintain a current state there in addition to the useState. So always use the ref, but set the state in the end to trigger a re render.
Hooks are so unelegant for real problems. I hate how they force your hand to write spaghetti.
If y never changes it should be declared as a const function outside of the context of the component, then it can be safely included in the dependency array without triggering unnecessary rerenders.
If y does change, you've just created a bug where your component will keep using an old version of y.
React is a closed ecosystem. Anything that enters either do through the root component or through hooks (in the functional version). If both x and y enters the ecosystem, they must adapt to the rules, meaning you write what happens when they change. Because the only reason for them to enter is that they are not constant outside React's world (which is the ultimate truth). If they are constant, then just use the reference directly.
If you're the one writing that reusable thing, you don't need to do anything special or write the kind of defensive code you're talking about. It's up to the original source of those props to make sure they're not creating new objects every render if they're in a context where that even matters
Components seem extremely similar to classic OO, why can't I use the same syntax.
Arugably this isn't a misuse, but just accepting that Reacts state management isn't powerful enough for what you need. Something akin to useRef + useSyncExternalStore is often what developers want.
For example, there's no guarantee that your useEffect will be invoked with the intermediate state you wanted to observe.
> I hope this example shows why I'm passionately against applying memoizations. In my experience, it breaks way too often. It's not worth it. And it adds so much overhead and complexity to all the code we have to read.
This design means that unlike vanilla JS, web components, Vue, Svelte, Solid, etc. where your state explicitly opts in to state changes, React requires component code to explicitly opt out of state changes (making the model much harder to wrap your head around and harder to do well).
This is the actual root cause of all of the underlying complexity of React. In signals-based frameworks/libraries and in vanilla JS with callbacks, there is rarely a need to explicitly memoize.
For the curious, I have some code samples here that make it more obvious: https://chrlschn.dev/blog/2025/01/the-inverted-reactivity-mo...
Sometimes I think it would be fun to make an immediate mode rendering framework similar to React -- and given that it can take 10x or more refreshes for a page to be done rendering, it might as well be immediate mode.
But yeah, I figured out the need for signals back in 2006 or so when I was working on a knowledge graph editor based on GWT and having to explain why there were "race conditions" depending on whether or not data was cached. When I got into modern JS around 2017 everything seemed worse than the frameworks I'd built for some very complex (think Figma) applications more than a decade ago.
[1] Of which 80% of the examples don't really need macros
To do what? Most of the times, simplicity beats cleverness.
As an example, Lit + lit-html is nowadays what React was before hooks. It is faster, simpler, more fun to code.
I have been writing react professionally in large codebases for 5 years now and I am quite disappointed by the course react followed. It became more complex but not more powerful, it introduced features that feel like bandaid fixes rather than new tools that open up new opportunities, it broke its own rules and patterns (the new “use”, the “useForm”) and it aggressively pushed for server side rendering, alienating hobbyists and small scale users who enjoyed SPAs.
Like another user mentioned, I am irritated by the linter rule for full exhaustive dependencies in effects, because it takes control away and also forces me to memoise everything. Or not, as the article points out. The library is not easy to build with anymore and it’s comfortably sitting at the point where it’s used everywhere and people are getting tired of its quirks. Which, if we are lucky, means it prepares the way for something better.
Once you know about the reactive render cycle, it becomes a second nature to never create any object other than via a memo function with declared dependencies. Anything else—any instance of something like <Component someProp={[1, 2, 3]} />—is, to me, just a bug on the caller side, not something to be accommodated.
Furthermore, ensuring referential equality has benefits beyond avoiding re-renders (which I would acknowledge should generally be harmless, although are in practice a source of hard to spot issues and loops on occasion).
— First, it serves as a clarity aid and documentation for yourself. It helps maintain a mental graph like “caller-provided props A and B are used to compute prop C for a downstream component”. Code organized this way tends to have self-evident refactor routes, and logic naturally localizes into cohesive sub-blocks with clearly defined inputs and outputs.
— Second, and it is obviously an ES runtime thing, but it is simply weird to have to reason about objects being essentially the same but different—and if memoization allows me to not have to do it and, with some discipline, make it safe to assume that (() => true) is the same as (() => true), why pass on that opportunity?
That said, now that I see an alternative take on this, I will make sure to document my approach in the development guidelines section of the README.
personally, I want to have a comprehensive understanding of exactly when and why each of my components re-renders, and to do that I use hooks and design around referential equality. like you said, it's a kind of documentation, more about organization than performance.
not to say that this way is intrinsically better! just that the style is appealing for its own reasons
it sucks both ways - the default is that stuff is different if it looks the same so you can't not do anything, but the only solution is both a lot of work and really easy to mess up completely. The TC39 Composites proposal[1] would be a great step in the right direction, but it will take years in the best scenario, there's really no guarantee that React will take advantage of it, and it still leaves out a lot of cases (Dates, to name one)
I think the other benefit is enough to justify the effort, though. Essentially, you don’t just make up objects, but instead you define each object’s constructor and all of its inputs. This makes it natural to keep logic contained and easy to shuffle things around when refactoring!
CostPostCache = CostOfCache + CacheHitPercentage * CostWhenHittingCache + CacheMissPercentage * CostPreCache
CostOfCache is pretty cheap compared to the CostPreCache IMO... CostWhenHittingCache is very low... and CacheMissPercentage is also probably pretty low in your typical React component that has more than 2 children node involved. Mathematically it feels like a no-brainer!
"Fix your slow renders first" just feels off. Yes you want to fix slow first renders! You also want to avoid wasting render cycles. These are distinct issues from my view. What am I missing? Why shouldn't React "just" do useMemo by default?
That’s what the react compiler does, and it’s a good idea in that case because the compiler knows how to do it correctly, for _everything_. When humans try to do it, they will likely get it wrong (see the real world example in the article, this is the norm imo).
> Adding non-primitive props you get passed into your component to internal dependency arrays is rarely right, because this component has no control over the referential stability of those props.
Expecting referential transparency is a fine constraint. Inversely, the useRef technique might lead to its own weirdness where your UI is using some stale composite object that your event handler doesn't see anymore (and I prefer slow perf to chasing down correctness issues because you have a ref _and_ you have some accidental mutability somewhere in the stack _and_ you have stale UIs)
Sure performance is a concern but is recursive re-rendering really cheaper than memoing all props?
But I wonder should it even work at first place? It feels like you are covering the problem instead of actually fix it. Make broken things broken is a lot healthier for the code base in my opinion.
Immediately find out things to be broken is 100x better than find out it suddenly not works and can't even figure out why is it broken months later in my opinion.
Use external value in Vue is very noticeable in contrary to React. Because it will definitely not work even once. The in component computed or component itself both enforce memorize . Render will be skipped completely it you used external values there so the framework did not notice anything is changed.
Sure reducers could be a bear (and asynchronous actions were hard), but the easily testable and portable logic-less views were really nice to work with.
Having worked in a number of codebases from React’s earliest days until now, I see echos of the bad old mixin days in the usage of hooks.
I do get what you mean about the conceptual separation, although something about the HOC approach also led to a lot of historical user confusion about "where _are_ my props coming from?".
As for writing the rest of your Redux logic, our modern Redux Toolkit package has addressed the historical concerns about boilerplate and other pain points:
- https://redux.js.org/introduction/why-rtk-is-redux-today
- https://redux.js.org/tutorials/essentials/part-2-app-structu...
- No Hooks.
- Don't try to sync any state. E.g. keep a dom element as the source of truth but provide access to it globally, let's call this external state.
- Keep all other state in one (root) immutable global object.
- Make a tree of derived state coming from the root state and other globally available state. (These are like selectors and those computations should memoized similar to re-select)
- Now imagine at any point in time you have another tree; the dom tree. If you try to make the state tree match to dom tree you get prop drilling.
- Instead, flip the dom tree upside down and the leaves get their data out of the memoized global state.
- Parent components never pass props to children, the rendered children are passed as props to the parent.
One note, is that the tree is lazy as well as memoized, there's potentially many computations that don't actually need to be computed at all at any given time.
You need something like Rx to make an observable of the root state, some other tools to know when the external state changes. Some memoization library, and the react is left with just the dom diffing, but at that point you should swap out to a simpler alternative.
As an example: Say the user clicks a button to submit a form, clicking the button updates the local state to include something like `status: 'SUBMIT_REQUESTED'` then you make the request conditionally, and update the state to `status: 'IN_PROGRESS'`.
It might become a mess, but the point is to do no side effects based on any events, all effects happen conditionally based on the state. My hope is the forces you to actually track everything you should be tracking in your state object.
If your parent component doesn't need the optimization, you don't use it. If it does need it, your intention for using useMemo and useCallback us obvious. It doesn't make your code more confusing inherently.
The article paints it as this odd way of optimizing the component tree that creates an invisible link between the parent and child - but it's the way to prevent unnecessary renders, and for that reason I think it's pretty self-documenting. If I'm using useMemo and useCallback, it's because I am optimizing renders.
At worst it's unnecessary - which is the point of the article - but I suppose I don't care as much about having unnecessary calls to useMemo and useCallback and that's the crux of it. Even if it's not impacting my renders now, it could in the future, and I don't think it comes at much cost.
I don't think it's an egregious level of indirection either. You're moving your callbacks to the top of the same function where all of your state and props are already.
Not really. The problem goes beyond re-rendering 15 times. For instance, how do you instrument usage? It can’t be simply debounced.
Similarly, you’ll be making unnecessary requests. e.g., we need to fetch X when this prop changes.
Or, we need to lazy load X once.
And back to re-rendering, there’s plenty of apps rendering @1FPS when dragging elements.
I will still say though, I have not actually had this happen to me yet with all the years of using hooks. Generally when I'm fetching when X prop changes, it's not in response to functions or objects, and I guess if it's ever happened it's been fixed and never broke or hasn't caused problems.
Not to say it isn't an issue - it is - but the number and degree of issues I saw with lifecycle functions was much worse. That was with a less experienced team, so it could just be bias.
Things were simpler when we had lifecycle methods to manage some of this things. But I`m sure that the next version of react will change everything and make us come up with more patters to try to fix the same problem again...
Useless Machine:
https://en.wikipedia.org/wiki/Useless_machine
A useless machine or useless box is a device whose only function is to turn itself off. The best-known useless machines are those inspired by Marvin Minsky's design, in which the device's sole function is to switch itself off by operating its own "off" switch. Such machines were popularized commercially in the 1960s, sold as an amusing engineering hack, or as a joke.
More elaborate devices and some novelty toys, which have an obvious entertainment function, have been based on these simple useless machines.
History
The Italian artist Bruno Munari began building "useless machines" (macchine inutili) in the 1930s. He was a "third generation" Futurist and did not share the first generation's boundless enthusiasm for technology but sought to counter the threats of a world under machine rule by building machines that were artistic and unproductive.
A wooden "useless box"
The version of the useless machine that became famous in information theory (basically a box with a simple switch which, when turned "on", causes a hand or lever to appear from inside the box that switches the machine "off" before disappearing inside the box again) appears to have been invented by MIT professor and artificial intelligence pioneer Marvin Minsky, while he was a graduate student at Bell Labs in 1952. Minsky dubbed his invention the "ultimate machine", but this nomenclature did not catch on. The device has also been called the "Leave Me Alone Box".
Minsky's mentor at Bell Labs, information theory pioneer Claude Shannon (who later became an MIT professor himself), made his own versions of the machine. He kept one on his desk, where science fiction author Arthur C. Clarke saw it. Clarke later wrote, "There is something unspeakably sinister about a machine that does nothing—absolutely nothing—except switch itself off", and he was fascinated by the concept.
Minsky also invented a "gravity machine" that would ring a bell if the gravitational constant were to change, a theoretical possibility that is not expected to occur in the foreseeable future.
When he went to edit his program it had been deleted. So he rewrote it and tried compiling it again and found the compiler had deleted itself too.
I like the idea of a program that takes increasingly dire steps to prevent you from compiling it each time you try.