Performance optimizations WITHOUT Reagent – part 3 (and probably final)

This is probably the last post about these improvements. At the last post, I showed how I threw every performance optimization I tried away (combine-leafs, other stuff, etc) just to throw it all away because React didn’t like what I did.

So, after a lot of time, I decided to do the horrible, insane thing that I did a lot more than I wanted – rewrite the rendering process of Chlorine to not use Reagent. The reason was actually quite simple: I wanted things to be simpler to implement, and honestly, Reagent was getting in my way. Like, A LOT.

Before I continue explaining the performance improvements that I am doing, I have to say something – in the first version of the rendering code (the one that I was trying to optimize at my last posts) I had a huge beginner’s luck – the actual rendering time is amazing for most cases, and while it is slow for some very specific situations, it’s actually not that bad; but there were also some weird trade-offs that I had to do and I cannot use them anymore.

The reason being, Chlorine started to become very strong in the customization code and “interactive renderers”. If you’re not aware, you can configure Chlorine to add custom commands and other amazing stuff, just by opening up a config file, writing some ClojureScript code, and you’re good to go. Unfortunately, also at the first versions of Chlorine, sometimes you would evaluate a code – let’s say a list of a thousand elements – and you get back only 10 or 20 elements. Worse yet, you would not get back a “List”, but an object called an “Incomplete List”. Which was not ideal, and basically made the whole “customize your editor” difficult to say the least.
(more…)

Performance optimizations in Reagent – part 2

Have I mentioned that it’s a nightmare to install React Devtools in Electron? Turns out, it’s worth it. But it’s still a pain. If you want to try, you’ll need to first install the extension in a chrome-based browser (like Brave, the one I use) and then install it by code. At Chlorine, what I did was to copy the whole extension folder to a devtools directory and then changed the script to load electron to the script below (see the BrowserWindow.addDevToolsExtension):

const path = require('path')
const {app, BrowserWindow} = require('electron')

app.on('ready', () => {
  let browser = new BrowserWindow({
    width: 900, height: 600,
    webPreferences: {nodeIntegration: true}})
  browser.loadURL(path.join('file://', __dirname, '/index.html'))

  BrowserWindow.addDevToolsExtension('./devtools')
})

Originally, the React Devtools was installed at ~/.config/BraveSoftware/Brave-Browser/Default/Extensions/fmkadmapgofadopljbjfkapdkoienihi/4.21.0_0 (tilde is my home folder). If you’re trying to install on Atom editor, you’ll have to open developer tools and issue the following script:

r = require('remote')
r.BrowserWindow.addDevToolsExtension('./devtools') // or the right directory

This command will fail, but somehow when you reload Atom, the React Devtools will be there. There’s only one feature that I use – highlight which elements were updated. And that’s where our performance tuning starts again.
(more…)

The magical wonders of programming with pure data

I’ll start with a disclaimer – if you’re developing in ClojureScript because of performance, you’re doing the wrong thing. Although a Clojure code can be really close to Java performance, ClojureScript simply can’t match JS performance – specially because in Javascript, the fundamental data structures like arrays, sets, and objects are implemented in a faster language, and as of today, no “user space” data structure can get even close the “native” ones. And, let’s not forget that JS is slower than Java.

So, if you want to beat JS performance in ClojureScript, you have to make use of what ClojureScript offers you – the REPL, the fast experimentation, and the wonderful experience of programming with data. So, this is a tale of a huge performance improvement that I’m currently working in Chlorine.

Rendering EDN

In Chlorine and Clover, when you evaluate something, the result will be rendered in a tree-like way at the console. This works quite well for Clojure, because UNREPL makes structures that are too big small, and you can query for more data. So far, so good.

In ClojureScript, Clojerl, Babashka, etc, things are not so good. Big data structures will be rendered fully on the view. They can lock or crash your editor, they can occupy all your memory, etc. The reason for that is that tree structures are hard – when you’re rendering the “parent”, you don’t know what the children will be. Currently, in Chlorine, rendering (range 80000) locks my editor for 4 seconds until it calculates everything, layouts all data, etc… and I wanted to change this.

Reagent vs Helix vs React

I was always intrigued on how much performance hit I get when I use Reagent’s Hiccup data structure instead of React objects – after all, there must be some performance problems, right? After all, that’s the promise of Helix – to not pay this performance hit because you’re closer to what React wants.

So, first things first: profiling Javascript inside Electron IS A NIGHTMARE. I decided to install React Devtools extension on Electron (inside the Atom editor) but that got me into so much trouble, false-positives, wrong profiles that I ended up deciding against it and simply used the “performance” tab to do my profiling. There, I could see batching.cljs hogging my performance, so the next move was easy – move away from Reagent and use Helix.

Except… that Helix uses some macros and I ended up doing the code in React. My testcase was simple: render a vector of 80000 elements and see how it would perform, obviously without all the bells and whistles that Chlorine offers today (otherwise this experiment would be waaay longer). And that’s where things get surprising: with Reagent, I was hitting 1800ms of scripting, and about 120ms of rendering. With React… 1650ms of scripting, and about 200ms of rendering. I decided to do more benchmarks and probably because of OS caching, warm-up, or whatever, the results got even closer, with Reagent sometimes performing better than React – but still too slow.
(more…)

Reagent Mastermind

One of these days, a friend of mine posted about his experience writing the “Mastermind” game in React (in portuguese only). The game is quite simple:

  1. You have six possible colors
  2. A combination of 4 colors is chosen randomly (they can be repeated – for example, blue,blue,blue,blue is a valid combination) – you have to guess that number
  3. You have up to 10 guesses of 4 colors. For each color on the right position, you win a “black” point. For each color in the wrong position, you win a “white” point
  4. If you can’t guess on the 10th try, you loose.

So, first, we’ll create a shadow-cljs app – create a package.json file, fill it with {"name": "Mastermind"}, then run npm install shadow-cljs. Traditional stuff.

Then, we’ll create the shadow-cljs.edn file. It’ll only contain a single target (:browser), opening up a dev-http server so we can serve our code, and we’ll add reagent library dependency. I also added the material-ui dependency, but you don’t really need it for the code. Now, running npx shadow-cljs watch browser will start a webserver at port 3000, and we can start to develop things.
(more…)