Yesterday, there was a big discussion on Twitter on how hard it is to start hacking on a js project. One comment by Dan Abramov struck me in particular: "Right: don’t use tools, face a problem, choose a tool or roll your own. Wrong: learn tools that don’t solve your problems, hate the tools."

This is spot on. All the solutions presented in this thread do not solve the problems I have when I'm hacking on a new project.

My dream PHP Setup

Back when I was 14, I had the best setup I've ever used in my life. Here it is:

  • Launch WinSCP and connect to
  • Create a file test.php, write <?php echo 'Hello World'; locally with EditPlus 2, press cmd+s to save, wait 0.5 second for the upload.
  • Open my browser and go to Click on test.php.
  • ?????
  • Profit


I want to get the same attributes but with JavaScript and React. Here are the constraints:

  • No setup: I'm happy to have to setup something initially (dedicated server, apache, php...) but nothing should be required to create a new project. No npm install, react-native init, creating a github project, yo webapp...
  • One file: I want to write a single js file to start with. No package.json, no .babelrc, no Procfile...
  • Sharable: I want to be able to share it with a url without any extra step. No git push heroku master or git push gh-pages.
  • Keeps working: Once online, it should stay there and keep working 6 months later. No npm start to run it, no special port that's going to conflict with my 10 other prototypes...
  • Not generic: I don't care about it being generic, I will use whatever transforms you decided. Happy to write js without semi-colons and using SASS for my CSS if you checked all the boxes above.
  • Not prod-ready: This setup doesn't have to be prod-ready, support unit testing or anything that involves it being a real product. This is meant for hacking on stuff. When the project becomes good, I'll spend the time to add all the proper boilerplate.

So, that's the challenge. Can you make it happen?

One common pattern when implementing user interface optimizations is to compute some value for a node where the computation involves looking at neighbor nodes and want to keep this value updated when the tree is mutated.

On this article, I'm going to explain the pattern we implement to solve this use case on various places in React Native.

Example: background color propagation

On React Native, we implement an optimization where the background color is propagated from the parent instead of being transparent. This provides a guarantee to the GPU that it won't need to paint pixels that are underneath. You can read this release notes for a more complete explanation.

In any case, the algorithm is pretty simple, if the background color is not set on an element, we take the one from the nearest parent that has one set. Here's an example:

Now, let say that the red node background color is being unset, the example would look like:

In order to implement this behavior, we first need to traverse up the hierarchy and find the background color and then down to propagate it, but stop at nodes that have set background colors. This means that we have to implement two different algorithms: one for the initial rendering and one for the update.

The complexity of this example as explained is small enough that it is easy to maintain the same invariants. But in practice, this algorithm is a bit more complex: we don't forward the color for transparent nodes nor for the image component in certain cases... We also experimented with more conditions that we didn't end up using.

Dirty-up and execute top-down

What would be great is to just write the top-down recursive algorithm you learn in school to apply colors however you want and whenever there's a mutation just re-run it on the entire tree. The problem with that approach is that you're going to spend a lot of CPU time running this algorithm on the entire tree when you only need to update a few nodes.

Instead, you can dirty the node you are mutating and all the nodes up to the root.

Then, run your algorithm top-down starting at the root.

You need to implement a way to figure out if a non-dirty node needs to be recomputed. The strategy we use is to cache all the arguments of getColor, for example (parentColor, nodeColor, isNodeText, ...) along with the result. If we're being called with the same arguments and the node is not dirty, then we don't need to go further and can just bail. The pseudo code looks like this:

function getColor(isDirty, prevArgs, prevValue, nextArgs) {
  if (isDirty || prevArgs !== nextArgs) {
    return realGetColorRecursive(...nextArgs);
  return prevValue;  


The benefits of this technique is that you can write your business logic once in a traditional top-down recursive way. You don't need to care about the much harder update case.

The downside is that you are doing more work: instead of touching only the elements that have changed, you need to traverse everything up from the root.

Another downside is that it requires more infrastructure as you need to add a caching layer and refactor your code to add a pass that starts at the top.

We're making use of this technique for a few aspects of React Native:

  • Background color is propagation as explained in this article.
  • Computing the layout (top, left, width, height) values based on the css attributes (eg marginTop, flex: 1...).
  • Flag if a text node is the root of the tree of text nodes to be able to flatten it into a single native view.
  • Flag if a node only affect layout and doesn't paint anything so that it can safely be removed from the UI tree.

I talk about various strategies to improve the performance of your React app: shouldComponentUpdate, Static Container, Element Caching, Raw DOM Mutations and finally, Data Binding, which is powering the Animated API.

I go over their respective trade-offs as sadly there is no perfect solution. Hopefully you can use the ones that make sense in your context.

I can't believe that I had the opportunity to do the opening session of React Europe, in front of 700 people! This is total madness how big this React thing has become and I'm so lucky to be able to play a small part of it.

In this talk, I tried to give an overview of the front-end ecosystem that formed around React and try to tell where I want us, as a community, to go towards!

How it all started

I was talking to frantic about some trick JavaScript question where we needed to use bind and apply in non obvious way and that reminded him of a similar problem where you needed to implement an add function that both works with arbitrary number of elements and can either be used as a number or another function.

test(add(1, 2), 3);
test(add(3)(4)(), 7);
test(add(3)(4)(5), 12);

It sounded fun so I asked him to provide me with some tests so that I could try to implement it. He did it and posted on the internal JavaScript group.

Then, Tadeu Zagallo, another co-worker working on React Native, saw it and while we were sharing our solutions mentioned that just closed down. I got emotional as I loved solving those kind of problems and for fun decided to shrink my implementation to see in how many characters it could fit. And this is how my Friday and Saturday vanished :p

In order to spread the fun, I posted it on Twitter and a lot of people helped out, including RReverser and sainaen who improved the existing solution!

Getting Started

In order to get started you need a valid implementation. Here is the code I wrote in ES6:

var add = (...args) => {
  var sum = args.reduce((a, b) => a + b);
  var fn = (...args) => add(sum, ...args);
  fn.valueOf = () => sum;
  return fn;

and "minified" it to be

91:var add=(...a)=>{var s=a.reduce((a,b)=>a+b),f=(...a)=>add(s,...a);f.valueOf=()=>s;return f}

But, unfortunately (or fortunately), ES6 has a lot of shorthand helpers that let you express this in a very concise way without much room for crazy hacks. So we decided to go the ES5 way instead.

The first version we started from was a direct translation of the ES6 version:

189:function add(){
  var s=[],function(a,b){return a+b});
  var f=function(){return add.apply(0,[s].concat([]};
  f.valueOf=function(){return s};
  return f;

There are three parts in this function: the sum, the returned function and valueOf. We'll see all the experiments we've done with the three.


Sadly, arguments is not a real array object so we need to do [] instead of the normal version with less characters arguments.reduce(.

57:var s=[],function(a,b){return a+b})

One of the realization of this game is that creating a function is very expensive in characters. You've got to pay for function(){return} just to play. Let's see with a normal loop how we can write it shorter.

56:for(var s=0,i=0;i<arguments.length;++i){s+=arguments[i]}

We already won one character, but we can do better. arguments is very long, we can make a variable out of it.

52:for(var s=0,i=0,a=arguments;i<a.length;++i){s+=a[i]}

Turns out that the braces are not needed if there's only one statement in the body. 2 free characters.

50:for(var s=0,i=0,a=arguments;i<a.length;++i)s+=a[i]

At this point we're a bit stuck with the normal for loop. Thankfully, JavaScript has another way of looping through an array: for(var key in obj). It is technically supposed to be for objects but also works for arrays :p

40:var i,s=0,a=arguments;for(i in a)s+=a[i]

This is the shortest we could find using a normal loop. But JavaScript has many builtin functions that can be exploited in a few characters. In this case eval and join. Yes, eval! We care about code size, not best practices πŸ™‚

39:var s=eval([],'+'))

The caveat with using this version is that join doesn't respect valueOf but instead calls toString. So we need to lose one character there.

Curried function

The objective for this one is to create a function that calls add with an initialized value of the current sum. The naive version is the following:

74:var f=function(){return add.apply(0,[s].concat([]}

Again, we don't want an entire function for doing that, the trick we're going to use time and time again is bind to generate a function in much less characters.

33:var s=this.s||0,f=add.bind({s:s})

Turns out that this only work in non strict mode. But if you enable strict mode it doesn't work anymore because undefined.s will throw an exception. Also, if you have a global variable s it will break. We found a way to make it a bit smaller using arrays instead of objects.

32:var s=this[0]||0,f=add.bind([s])

What you would really like to do is to just bind(s), turns out that you can do that if you are in strict mode. Because if you call add(), then this will be undefined and ||0 will work fine.

27:var s=this||0,f=add.bind(s)

What if we just did this|0, it will convert the element into a signed 32bit integer and has the nice property that window|0 === 0 and undefined|0 === 0 so it works in both strict and non strict mode! The caveat is that the function no longer works with floating point numbers as 3.6|0 === 3.

26:var s=this|0,f=add.bind(s)

But then we realized that we didn't need to pass that initial value as this, we could just pass it as an additional argument. Not only making it shorter but another few characters, but also working with strict/non-strict and floats πŸ™‚

23:var s=0,f=add.bind(0,s)

Constant function

The last piece of the puzzle is to write a function that always return a number. This is something simple yet there are many ways to write it. Let's start with the naive version:

20:function(){return s}

The first attempt was to look at array functions for one that returns the first element and pass an array with just one. Turns out that join fits the bill. The only trick is that it converts the number into a string, so we need to cast it back to a number when we do the sum s+=+a[i] losing a character.


Then we figured out that we could use Number as a function to generate a number


But, we're going to be evil again and use eval to shorten it even more!


Putting it all together

We start with the following

108:function add(){var f,i,s=0,a=arguments;for(i in a)s+=a[i];
f=add.bind(0,s);f.valueOf=eval.bind(0,s);return f}

We have four variables f, i, s and a that we have to declare, but we don't really need all of them at the same time. We use isa together and then fs together. So we don't need to allocate a variable for f, we can reuse a or i.

106:function add(){var i,s=0,a=arguments;for(i in a)s+=a[i];
a=add.bind(0,s);a.valueOf=eval.bind(0,s);return a}

var i,s,a is a lot of characters, would be nice to be able to move it inside of the function declaration function add(i,s,a). Unfortunately, this only work in strict mode because in non-strict, if you override the variables it's also going to override the variables in arguments in the same position.

105:function add(i,s,a){s=0,a=arguments;for(i in a)s+=a[i];
a=add.bind(0,s);a.valueOf=eval.bind(0,s);return a}

Thankfully, when we switch to the eval version of the sum, JavaScript reads from arguments before setting the variable so it works even in non-strict mode πŸ™‚

105:function add(s,a){s=eval([],'+'));
a=add.bind(0,s);a.toString=eval.bind(0,s);return a}

Now we're getting in the final phase where we're trying to inline as many expressions as possible. When you have s=...;bind(0,s), you can turn it into bind(0,s=...) which saves two characters: a ; and a repetition of s.

103:function add(s,a){a=add.bind(0,s=eval([],'+')));
a.toString=eval.bind(0,s);return a}

In the same vein, we can transform a=...;a.toString=...;return a into return(a=...).toString=...,a. It transforms 3a and 2; into 2a and () which is one less character!

102:function add(s,a){return(a=add.bind(0,s=eval([],'+'))))

Here, we use eval (4 characters) twice: once to compute the sum s=eval(...) and once to have a function that returns a constant eval.bind(0,s). It turns out that we can build a function that returns the sum by doing s=eval.bind(0,...) and in order to get the sum we use s(). Doing this we save another two characters to reach 100!

100:function add(s,a){s=eval.bind(0,[],'+'));

It turns out that you don't even need to eval in the sum itself, you can pass the string '1+2+3' as first argument to the add function and when it is finally time to toString, then you can evaluate this entire expression.

96:function add(s,a){return(a=add.bind(0,s=[],'+')))


  • This was super fun πŸ™‚
  • I would have never imagined that there were so many avenues of optimization for this seemingly simple problem.
  • New JavaScript standards like strict mode and ES6 make writing more concise code easier.
  • In order to spawn collaboration, having a runnable specification is the critical.
  • No one person came up with all the optimizations, it was truly a collaborative effort.