State of JS 2023 Reactions

Welcome to Syntax. Today, we've got our state of j s 2023 reaction episode. We do 1 of these every single year, and I realize it's quite a ways into 2024 already. But, Scott of JS, which is a survey that goes out to sort of the JavaScript community and ask questions about JavaScript features, frameworks, libraries, resources, all that type of stuff, has just rolled out the latest results. And I often I really like these surveys because it it sort of gives you a peer into our industry of what are people using, what Yarn people like, what what's sort of on the uptick, what's on the downtick, what are warp what are people actually using versus 2 knuckleheads talking about on a podcast every single week? So it's an exciting 1, and then we're gonna dive on on into it. My name is Wes. I'm a developer from Canada. With me JS always is Scott. How are you doing today, Scott? Hey. I'm doing good. Back in back in action over here.

I had a, out I had a ordeal getting getting home from Italy, man. I I always like I had to sleep on the floor in an airport for was in an airport for I was in JFK for, like, 24 hours with my kids.

With your kids?

With kids.

They're 57, so they're Scott, like, little little, but, like, they're little. And they slept on the floor in the airport with me just because Delta didn't wanna cancel the flight because they didn't wanna give people, like, hotel reimbursement. So they were just like, oh, we'll tell you in 20 minutes. We'll tell you in 20 minutes. And it's, like, 2 AM, whatever.

And they're like, alright. It's postponed until tomorrow. Tomorrow tomorrow at 8:15 AM, we get to the gate. Oh, the the flight doesn't have a crew or a pilot. It's postponed again. You're just like, oh, there there was an actual literal fight.

The police had to be called.

It was yeah. For you? There was there was 2 airplanes that were, like, leaving at the same time, us and another 1, and we both got on the planes.

No. No. No. And we both had to deplane after being on the plane waiting for, like, an hour. So so we had to deplane, and then the other the other flight was a bunch of people that were going to go on a cruise, and so they were gonna miss their cruise. And so the the people across the way from us that that they they were just, like, basically losing their minds, and it got heated. I don't I don't necessarily know what happened, but it was a kerfuffle.

Next thing you know, there's police protecting the, the Delta employees who were awful, by the way.

I I never fly Delta, and, man, I'm a I'm a Delta hater now. It it sucks. So yeah. I felt kinda bad because I had previously said 1 of my Twitter tips was to mute the name of every airline. Yeah. Yeah. It's And then I went into the chat and saw that you had this massive ordeal, and I was like, oh, poor guy.

It's Especially with kids. Like, I I do not wish that on anyone. There's nothing more frustrating than being delayed, let alone with kids. So Oh, yeah. Brutal. Yeah. It was brutal. But I'm back.

I'm checking back. Feeling good. ESLint ready. And you know what else is ready? Our code base because it uses Sentry to track all of our errors and exceptions. That's right. We make sure that we can solve anything as it comes up. But not only that, we can fix our our, performance problems before they even become problems because we get a user misery score to say, like, well, this page is being hit by a lot of people, and it's very slow. Therefore, this page has a lot of user misery.

Let's tackle that. Let's find let's dig in. Let's look for our slow queries. Maybe something's missing in index. Maybe something isn't being cached that could be cached. And Century makes it easy to solve and find all that stuff. So if you want all of this and more, head on over to century.i0forward/syntax.

Sign up and get 2 months for free.

Awesome. Alright. Let's get into oh, let's let's real quick plug. We just got new T shirts for syntax. Scott and I are wearing 1 of them. I'm wearing the sick pick T shirt. Scott is wearing the blazing fast.

Oh, he's just spilled his water everywhere.

Blazing fast, worldwide Wes, information superhighway. Wes got a whole bunch of other ones as well. We got a Prettier collab 1 and a Drizzle collab 1. They're pretty cool. Look at this.

Yeah.

We got hats coming too. I just ordered the hats. It's gonna be very similar to this 1, except what's that called? The 5 panel Wes it goes, like, right here? Yeah. I'm pretty sure it's called the 5 panel. Mhmm. So it's it'll be a little less clumpy right here, which is this is a this is a tester. So got some cool stuff coming down the pipe. Go to syntax.fm and click on swag in the top.

You'll you'll get to it. Yeah. Alright.

Let's get into the 2023 state of JS survey. Let's kick it off with the front end framework stuff because I always find this the the most interesting.

So front end framework ratios over time, I think this is always interesting to see which front end frameworks are people using.

By far, the highest 1 is React, which has gone up from 81 to 84.

These things Yarn, we should say, they're obviously going to be biased based on the type of audience that answers them as as well as, like, we tell people to to go like, a couple years ago, we told people to go and, take the survey. So, obviously, there's going to be people that listen to this podcast that are, answering this, so it's gonna be pretty heavy in React's felt world, but that's nothing is ever perfect. No data is ever perfect. But React has gone up 3%, and Vue has gone up as well, which is pretty interesting, to see an uptick.

What's interesting here is that almost all of them have gone up. So Angular went down, and has gone down year over year. Since 2018, it was at 56, and now it's at 45%.

And lit is kind of, like, staying stagnant, but the rest of them have all gone up, which is all pretty interesting to me. I don't know if it's just more people trying different things, and then that's it because there's a lot of options. People are actually trying things Node, but, like, it's cool. And and HTMX showing up here at a 5%.

Man, it JS really interesting to see how much some of the like, viewing up, like, 4%, Svelte went up 4%, Quicken up 3%.

So interesting that they most of them seem to go up.

Yeah. I wonder if that's because people are trying other things. You know? Like, probable I I would I would expect React to go straight and the rest to go up because people are saying, ah, you know what? Maybe I should try something else out, but they almost all have gone up so much that, like, they're not it's not that data's not coming from anywhere else. Right? Maybe those are people that are are coming from Drupal or WordPress or something like that because it goes all the way back to 2016 where at the time, only 52% use React.

Angular was 19 and Vue was 10. Right? It's very small percentage of people who are even using a framework at that time, and now it's seems like almost everybody is. Yeah. And and here's a little interesting tidbit because you can view by usage awareness, interest, retention, and positivity.

And I wonder if positivity is maybe a potential precursor for usage in the future. Because if you do this on. Positivity for React, it has gone down from 2020, 81% to 71%. So positivity for React has gone down by a whole 10%.

Meanwhile, positivity for Vue JS basically kinda stayed the same the past couple of years, but gone down overall. Positivity for Svelte has gone up. Yeah. I was gonna say that. What's the only 1 that has gone up every single year? Yeah. Svelte. Svelte. It's pretty wild. And h t max.

And lit. Lit has gone up too, although usage. Solid's gone up. You know, it seems like Solid is becoming 1 of those ones that's getting more and more awareness, in general. I think people react folks who come to Solid are are gonna find that it's a nice experience compared to maybe what they're dealing with React. But I always find the positivity 1 to be 1 of the most interesting aspects of this just to gauge general sentiment around around these things. Because if you look at positivity around Angular, usage goes down. But positivity JS is kind of remained flat for quite a while and even has, like, a small uptick. So I I wonder, like, how this eventually translates to usage.

Let's talk front end framework pain points because these are the things that I think people have issues with their front end frameworks. And it it's funny because, you know, how many of these are specifically to React or any of these things? And it what I like about this is that they've broken them down. Let's say you have very React specific issues.

React specific issues account for 17 percent of the respondents.

So those would be excessive dominance. That's a wild pain point to say.

React is too dominant. And, I mean, I agree. I I don't love working React myself, and it's everywhere. So you can't avoid it. Right? So you fall into this trap of, well, I gotta use it even if I don't like it. Slow progress, community concerns, those are all a lack of signals, which is an interesting 1. You can just use signals. I mean, you don't have to use, you know, reacts baked in signals to use signals there.

Excessive complexity, choice overload, these are all things that we've been hearing all the time.

This is you know what's really cool about this is a lot of these questions this year have been free form, and I'm not sure we should maybe have Sasha on again who runs this, and see if he's using, like, AI or something like that to sort of collect this data 0, yeah. But instead of being like, what's your favorite framework? 1st 1st checkbox, is it React? You know? Now it's just free form input, and you can click on the issues and actually read the respondents that people have have said, and it's it's really cool. Excessive dominance. Let's see. Dominance of React. Predominance of React. React is everywhere, but is used in a standardized way.

React touches everything. Everything's related to React.

React sucks.

Yeah.

It Yeah. But, honestly, this is such good data because, like, people that are taking this thing care about web development. Right? And this is not just, like, hot takes. This is, like, actually people

that are are saying this type of thing. Yep. Yeah. It is a common sentiment, and I think you get that from that 10% decrease in people liking using React. But, also, like, we talk all the time about choice overload. To me, that the numb that being the number 2 problem is kind of a React problem in itself because the React Sanity, more so than Vue, more so than Svelte or Angular, makes less choices for you. And therefore, that right there increases the choice overload. It also increases the excessive complexity because now you're wiring all these things together that might not be connected, where with Vue, you have a very clear options of what you should be using for state management for for basically everything. Right? And same with with Svelte or or Angular. They take care of a little bit more for you. And because of that, I would say that you have less choice overload and less excessive complexity there. And maybe that that choice overload and excessive complexity are are symptoms of the React ecosystem itself, but, you know, who knows? JavaScript Vercel has got, like, a 1000000000 different framework options.

I think so. If you if you click on over to state management and you just kinda peruse the actual answers, a lot of people are saying obviously, state management. A lot of people are saying dealing with forms

Wes state. And

now now we have that, right, in in React. So I'm curious what this will look like once people start to give, form actions a shot over the next year or so.

Yeah. Yeah. Interesting stuff. And it does I mean, all of these track with things that we've been hearing for so long, in terms of, you know, people's fatigue about JavaScript frameworks and how they feel about them. So, you know, it's obvious that this stuff is a real problem, but it's interesting to see it being laid out like this.

Let's talk about meta frameworks unless you wanna unless you have more there. That's exactly what I was gonna go to. So Let's do it. Major jump in Next. Js

from 48% to 56% of respondents, having usage.

Let's see what Wes let's see what the interest. Yeah? Or the positivity for Next. Js goes from 75% to 64%, an 11% drop in positivity while they jump in in increase in usage and a decrease

in retention as well. That's interesting. Wow. So Next. Js usage has gone up, but interest, retention, and positivity all have gone down. Yeah. So it's very similar to the React answers where everybody's using it, but a lot of the complaints Yarn, I I don't want to use it because it's it just has such dominance.

Yeah. It feels like people are being, you know, led to use it and then maybe not having a great time. You know? Maybe that's it. Yeah. What what's got the most positivity,

uptick here? SvelteKit.

Man, SvelteKit is killing it. SvelteKit is killing it. Astro also, though. Astro goes from, oh, since 2021, it goes from 20% to 54%. And I gotta say that tracks with my experience with Astro. I I tend to really like it. I pick up SvelteKit just because it's like it's the Svelte framework, and that's what I use. But Yep. If I'm if I'm writing a React application today or Vue app and maybe not a Vue. Maybe Vue. I don't know. If I'm writing React application today, I'm picking Astra as my my framework for sure.

You know what's interesting is that the Gatsby usage has sort of remained standard, and it still has gone down over the years. Gatsby JS wild. We've talked about this a couple times, but just the the come up and the slide down in the Gatsby world has been been wild.

Jamstack, man.

The jam Jamstack. Yeah.

You know what? What's funny about how that these these comments about Jamstack relate is that, like, I've been doing more what you could call Jamstack now than I did even back when Gatsby was at its peak. I'm I'm like it could be the hipster in me, but I've moved fully into, like, alright.

SSR my my about pages and the SEO stuff, but CSR, everything else, keep my data clients like, I'm going so hard in that direction. It's it's very funny now.

It's just like whatever everybody's talking about, I can't be doing it for some reason.

Yeah. What so what in 2 years from now, what are you what are you gonna be, hacking SSR. Who knows? Yeah. You're gonna be Yeah. Talking about PHP was right. Yeah. I know. No kidding. Yeah. I'm kinda bummed to look at the other meta frameworks here to see some some really good ones that really aren't getting enough love here. Like, QUIC, only 34 people have have written it in. Like, I had to expect Quick to have a a bigger usage than that. Right? And Redwood, only Deno.

Yeah. Adonis, 12. You know? Like Meteor's higher than Adonis. Yeah.

Those things need a bit more love. So we should we maybe we'll describe what they are. So QUIC is a framework for building websites where it does resumability, from the server to the ESLint, and we had on. You can go listen to that episode. It's syntax.fmforward/574, to sort of understand what quick is all about. Quick is really, really neat, and you can bring all of your React components to it. And same with, like, Redwood. We had Tom Preston Warp on from Redwood as well.

Yeah. And it's just a like, a lot of times we say, like, hey. Why is there no Rails for JavaScript in Redwood JS that?

Yeah. It it is funny how these things it it it is like it's weird because, like, yeah, people are saying there's there's too many choices that need to be made.

There's, too much fatigue in that direction. But then the options that reduce that fatigue are, like, not present in the usage category. It's like, hey. These things are here.

Yeah. If you were to give them a look and give them a try, especially with things like Redwood or Adonis or some of these other ones. Even even Meteor does a lot for you. And and I haven't kept up with Meteor that much, but it's it's still chugging along for sure. Yeah. Do you think that's because people just want to be told what to use? Like, the reason why Next. Js

usage went up but the interest and positivity went down is because, like, well, like, I wanna use the thing that everybody's using. It's the most documentation, examples, the most integrations, right, where, like, a lot of people would argue that Redwood, Adonis, a lot of these quick, solid a lot of these things are better. They're they're way better. But, like, if that's why people are in, like, Rails and Laravel world are so happy because they're just told what to use, and everybody uses the same thing.

Yeah. I know. It's so funny because that that conversation has been happening so frequently lately where JavaScript developers are trying Laravel, and they're like, wow. This is actually great. And I think the reason why it's actually great is is not be I mean, Laravel's great, but not because Laravel is anything special. It just makes all the choices for you for you. Right? Like, it does so much for you that you you, you know, don't have to do all the all the pain pnpm stuff that was listed in the pain points. And if we go to the the meta frameworks pain points, the most respondents, number 1, is is Next. Js issues is number 1, for pain points here, including the app router being a big thing, the caching Next JS 13. I I have heard a ton of things about the app router being a pain. I don't know much about the App Router myself. I haven't I don't know what's different about it. But Yeah. It's

applications on, the Pages router.

And now they're like, oh, AppRouter is the way to go. And, obviously, they're still supporting the the old way, and you can mix and match them. But it's a big, it's a big thing to switch over, and it's, like, another mental overhead shift of changing from 1 way to another.

Yeah. And so excessive complexity, number 2. Number 3 is deployment. This to me, honestly, feels like a Next. Js issue as well because I don't find many of these frameworks to be difficult to deploy.

And not that Next. Js is difficult to deploy, but if you want, like, the the features, you're gonna use it on Vercel. Right? And if you try to deploy it elsewhere, yeah, then it's difficult to deploy. But, that's interesting. So many so many respondents said deployment given that anything that adopts the adapter pattern is super easy to deploy because you just install a given adapter and and let it rip. But, again,

sayer side rendering or just spit out a Node app. Yeah. I love that they publish the results, like, what people wrote in here.

Because you can say, like, if we looked at this deployment, we can guess what people are thinking, but you can literally just read the results.

And probably every third 1 does mention Next. Js and Vercel. Not because it's bad, maybe because it's too good. Right? They're so tightly coupled

together, and it Deployment of Next.

Js. Many of Next features are locked behind vendor. Yeah.

Interesting.

And lock in is number 13 on the pain point.

Vercel is 9 on the pain point. And what's funny is that, like, Vercel makes everything really easy. The problem with that is that it makes everything really easy in a way that could potentially cost you a lot of money later on, or you're locked in now to using it and and being on that as part of your platform. Right? You're you're you're locked in to that being part of your platform.

And not that that's a bad thing if everything is good, but it it is something to be aware of. Mhmm. You you don't have that flexibility and freedom that you get with just deploying a node app anywhere because we that that used to be a really nice thing about Next. Js. You could just deploy node app anywhere, and that was just the way to do it or do it on Vercel. It's just a straight up node app. I'm sure you can still do it as a straight up node app. But yeah. Yeah. Interesting.

I I think the thing is people want.

You you can still deploy Next. Js as a node application, and it runs really well. Right? Yeah. But everybody wants they want the CDN. They want the caching. They want the instant deployment, the the Git previews without having to spin up a server.

So we wanna I'll tell you where I've been hosting everything now is basically just Cloudflare. I've been hosting everything on Cloudflare. I I find it to be dead simple. So good. Works every time.

I built a a whole Next. Js site, and I used all of the caching features. I used React Vercel components. I used, like, a good chunk of the Next. Js features, and I used it's called next on pages, which JS, like, Cloudflare's adapter, I Wes, you could call it, and it worked really well. There was still some, like it's not Vercel level. It's not as as like, oh, wow. This just works, you know, and all this nice integration, but it worked pretty good. Enough where I was like, yeah, I like this a lot. And I I use the d 3 SQLite database inside of it. So it's I think it's if I were to say it's probably worth figuring out how to get it to to run on Cloudflare as well because it's once it is working, it's really nice.

Yeah. I just got a bit into Cloudflare workers, and it's so funny. We'll we'll talk about this in another episode because I built a little font serving service to host our licensed fonts in a way that we didn't have to put them in the repo.

And I did a, like, a a bit of a testing here. It's like, where do we host these things? Do we throw them in a bucket? Do I throw it in a a function a pages function? Do I throw them in a worker? Like, what what's the fastest way to do all these things? So we'll talk all about that. But hosting, man, it it's a wild world right now, and and I I tend to just either throw it on Coolify or throw it in Cloudflare and and call it a day.

Let's talk about build tools.

Still coming in at the highest usage is webpack at 90%.

That to me is wild. I would have thought that everybody is is moved over to

to Vite at this point. But It's it's Next. Js is really what it is. And if you go to the positivity, Webpack had a 89% positivity in 27, which is wild to me in the 1st place. But now it's, like, down in the 40%. So it's nearly dropped 40 percentage points in in terms of positivity. I have not touched Webpack almost my entire career.

It is like a such AAA Node thing for me. I'll I'll just say it. Yes.

V, up to 73% usage, and then the positivity

to 80%. Yeah. Yeah. It's great. I have a a prediction, Wes. Yeah. My prediction is in the 2024 respondents, we will see a roll down on this list, because I think people are gonna actually start to get to use it in some regard at some point this year. Because I even saw Evan Evan just, Evan, you posted today that he had his 1st view app using roll down actually working.

I I don't think I've heard of this. What is roll down? Obviously me tell you. Roll up alternative?

Roll down. So here's here's the gist on roll down. Basically, Vite, the thing everybody knows and loves, Vite uses 2 different build tools behind the scenes. It uses roll up for production builds because roll up is very good at all sorts of things in production builds, and it uses ES build for all kinds of things as well. It's very fast. Right? So it uses a combination of these 2.

And they said, like, alright. Well, what would a next generation version of roll up look like that had the speed of ES build, but it was built from scratch in Rust.

So roll down is an attempt to, essentially, have a next generation version of roll up built in Rust, that's super fast and will eventually become the main build compiler for Vite. It will replace ES build and and roll up within Vite. Interesting.

The road map's super interesting if you wanna give the road map a, a check because the it's it's moving quite along. And, I I think it's 1 of those projects that we're we're gonna hear a lot about over the next, I don't know, next year or so. That's my big big bold prediction.

Not all that interesting,

pain points here. Configuration, obviously, being the the biggest 1 of Yeah. With Webpack. All of them. Config's always gonna be a a thing.

Yeah. Performance, yeah, also, Webpack Mini because let me tell you. There is no performance issues with Vite. Excessive complex complexity, no issues with Vite in there for me.

Webpack issues number 4. Yeah. If you're looking at this list for me, I'm I'm trying not to be too much of a hater here, but, like, the top 4 are all Webpack issues if you're asking me. Wes and and CJS, definitely a big pain point for a lot of people, and definitely something that I I don't know how it gets any better, to be honest. So, you know, I I I've moved fully over to ESM a long time ago, but you still have a 1 off situation here and there. And it it still can be a pain ESLint, definitely.

Oh, yeah. It's, I I think a lot of people also don't realize that they are running common JS at the end of the day because a lot of these build tools will still put out CommonJS.

And the biggest pain point is for library authors who still need to publish these types of things, so that everyone can use it. It's such a such a pain.

Let's move into JavaScript runtimes.

This is kinda interesting. Where does the JavaScript actually run? 94% of respondents browser JavaScript. Weird. I don't know if that's that that doesn't feel

browser JavaScript. Weird. I don't know if that's that that doesn't feel like Why would that good information. Yeah. I know. That's that's crazy. But Maybe people just aren't thinking about, oh, you're doing it. Yeah. Like, the browser is a runtime. Yeah. 22%

on BUN Wow. Versus Deno on 15.

I think, like I don't know anybody that's running BUN in production. Have you do you know anyone? That's a great question. I don't several small apps running in in Deno, which is at 15%, but BUN certainly has lots and lots of hype, in the last year, which is probably

has to show for that. Yeah. Service workers always I like about BUN, but I ultimately can't find myself really committing to the things that make BUN. Like, the things that make BUN unique are all excellent, but I find myself having a really hard time saying, like, going all in on that, and I'm, you know Yeah.

Foregoing anything of the you know, being able to move back to Node or or even to Deno.

17, 000 people said Node. Js.

48 wrote in Cloudflare workers.

So that's a write in. It's obviously lower. But do you do you think people who write Cloudflare Workers think that they're writing Cloudflare code? Because in my head, I just think that I'm running, I don't know, JavaScript code. JavaScript. Yeah. I I wouldn't call it Node.

I would assume a lot of people who might even answer this might not even have, like, a firm understanding of their runtimes and and what their Yeah. Oh, we're using the same same dang cup. We got our big cups here. Man, we are just, like, just got it all right now. We showed up to this recording wearing the same T shirt, and I had to change. So

I had, like, a modal dialogue open, so it, like, was covering Scott's face. And I I just saw the same T shirt, and we were, like, pointed the same way. And I thought something happened with my video where it duplicated, but it was just Scott.

Great.

Edge and serverless run times, 33% on Lambda, 24% on Vercel Edge. Oh, here we go. 16% on Cloudflare Workers.

So I guess people think, yeah, maybe that's more of a serverless runtime than a JavaScript runtime. I don't know if that's but, oh, I'm actually surprised by this JS that Cloudflare Workers is almost the same as Netlify at 11%, Azure Functions, 9%, DigitalOcean Functions. That's something I have not tried out yet JS DigitalOcean's implementation of functions just yet.

Yeah. I you know, DigitalOcean is a platform I think about for VPS. They're, like, whole app platform thing I have not really dove into at all. Yeah. I I tried it once when I was trying to deploy

my server, and I realized, like, oh, I gotta pay twice because I had a back end and a front end. And I was like, a VPS is kinda better, because you can run as many apps as you want on that.

Yeah. It obviously doesn't come with all the niceties, like the the auto build and cut Vercel and staging URLs, all that nice stuff.

Yep. Let's talk about back end frameworks, which is interesting because Express is still king. 73% of the response, which is wild to me. I get there's a lot of people who have built on Express. It's it's I'm still on Express, and I'll tell you what. Wow.

I'll tell you why I'm I'm not off of Express. ESLint Tell me. The the server is often everything is built around that, and it's I think it's very hard to to move that type of thing away, especially because, for me, at least, a lot of my middleware, a lot of my auth, and all of that, it's all based on connect, which is the request, the response.

You have that next middleware.

And and moving off of that is kind of tricky, especially because a lot of the other features, halo JS is probably the closest thing to it.

They don't implement that as well. However, now that we have the, async local storage API, I've been moving all of my middleware of sticking data on the request.

I've been moving it to a single storage, which allows you to share data between function calls.

So if you have a middleware function that runs and then later on, you have another middleware function that runs, You can put everything in a store and just access it without having to pass it in, and that has been really, really nice. So and, so that's my 1 reason. The other reason is people are not mad at express. You know? People don't have express pain. I think that the all the pain I'm starting to feel is that it's it's not a standardized web request. It's not a re it Yeah. It's a it's an express Wes. And now I wanna move Compare it. Everything.

It's not, though. Like, I my website JS fast as hell. You know? Like, yeah, you can look at the the graphs and see that there's 20, 000, 000 requests. It's you can do a 1000 times more in other ones, but I don't think people that have apps in express see that or or actually run into it. Right? Like, my website is our websites are fast as hell. You know? And I've got caching in place and and things like that. The only reason why I wanna move off of it, Express now is because Scott of the new stuff using a fetch Wes, send a fetch request from the client to the server. Oh, now I gotta like, if it was just a web request, I would have the request from the client would be the same as the Vercel. And now I gotta convert it to

the express version, you know, and and that's a bit of a pain. Yeah. It's it's interesting to see Nest at number 2 here. I always figured Nest was a little bit more niche, but it it does a little bit more for you. It it's definitely 1 of those ones that the Angular folks are all definitely on, so that tracks because Angular has a lot of usage.

Fastify at number 3, nice to see. I've always loved working in Fastify.

Strapi, Koa, Happy, these are none of the ones I've I've I feel like these gotta be, like, a little bit legacy because you don't hear that much about them these days, but they have been existing for a long time. Meteor.

Yeah. ESLint interesting to to still see things like we talked about, maybe perhaps like Redwood or or sales or some of these being so much lower or even, you know, the much hyped up Eliza JS is super low on this, which it goes to show you that, like,

Twitter hype ain't everything. No. I would expect Hano and Eliza to be much higher. And I'm also surprised to see I think Strappy, Directus, Redwood are the are the only 3 keystones on here. Mhmm. The only 3, like, CMSs.

You know? The rest of them are just, like, server frameworks. Oh, sales is on there too. Yep. Yep. Sales. Cool.

Alright. Let's what do we got next? Non JavaScript languages. Okay. Don't did you don't look at this. I wanna I want you to guess. K. What are the top 5 non JavaScript languages used by people in the survey?

The top 5, do I have to guess them in order?

No. Just if you well, try, but I won't I won't stop you. 5

non JavaScript languages.

So, alright. Python, a lot of AI.

Number 1, 44%.

Be I'm surprised. That's that's very high.

Yeah. I mean, there's a lot of stuff written in Python. Yeah.

You know, you'd probably have to say PHP, obviously. Yeah. Number 2, 31%.

Feel like Ruby has to be here still?

Man. Ruby number 10.

You gotta be kidding me. Really? Number 10 on 9%. But I I think that that shows more of Ruby being its own ecosystem Yeah. I agree. Versus Instead of JavaScript. People not actually using Ruby. Because I think Ruby is Ruby is below Kotlin and Bash.

Yeah. What? You know? K. There's no chance. That's not that's not usage. That's just usage in the JavaScript world. Yeah. That's Yeah. Yeah. That makes sense. Okay. So I got 3 of them or I got 2 of them.

I have no idea. C warp. We just talked about C sharp. C warp, number 4. Actually, I had

I sat next I I had dinner with Bruno from Century, and he does all the c sharp stuff.

And he's like, c sharp is way cooler than you think. Yeah. And I said, but do I have to get a Windows laptop to use it? And he goes, no.

You know? Like like, that's what people think of of c sharp. Like, you gotta, like, use this light theme Windows XP, Dell Latitude to to use it. Obviously, it's not. So I think we should do, like, a show on that or have somebody on. I think it'd be kinda cool. I know you you and CJ were talking about it as well.

Yeah. Yeah. CJ was schooling me a bit. Alright.

Let let's, I feel like I should know these last ones. But yeah. Yeah. Number 3,

Java.

Java. Java.

Yeah. Yeah. Yeah. Yeah. That makes sense. Number 5, bash?

That doesn't make sense there. Bash is higher than Go and Rust.

Yeah. Go Node and Rust are still a little niche, but III am surprised to see that at least Go isn't higher than Bash. Although, I I do write a fair amount of Bash. I guess if if you think of it JS, like, these are languages of people who write JavaScript

Yep. Probably who are very heavy JavaScript.

What are the other languages that they will reach for? And Yeah. Yeah. Go, Bash, Rust, those are things that you reach for when you need tooling in that area.

Yeah. Interesting.

Hosting service number 1 hosting service, 47% is AWS.

For sale, very close behind at 45%.

GitHub Pages, number 3.

Okay.

I've never hosted anything on GitHub pages, so I don't understand this 1. But yeah. I do. But it has to be unless you're using their, like, weird Ruby

static site builder, you have to have, like, a build step that outputs So HTML. HTML. Yeah. Node use it for all my conference talks. Like, the some of the live. Yeah. Those run on GitHub pages, but it's just Okay. It's HTML. Right? Nullify, number 4. Heroku still Wes? Still Heroku is still number 20 number 5 at 25%.

Man, it just goes to show you, if you could get enough market share, you you'll be you'll be cruising forever, because I can't imagine deploying on Heroku today. I see. That's why all these

companies give you cheap and free hosting Yes. For for a couple of years. And then, like, once you're there same thing with, like, GoDaddy.

You know? I have a Bluehost thing that I pay, like, $120 a year for, and I don't even have any of my own websites on it. I've got my wife's site on it, and my dad has, like, 8 different sites on it. And I still pay for it every single year. You know? Like, that's why Bluehost had such a high I remember they're giving people a $100 if you sign somebody up for Bluehost. They were giving me a $100 if I signed people up for Bluehost. Yes. Yeah. I yeah. It's wild. And the reason is is because once you got stuff on there, man, it's just not worth moving it.

Yeah. For real. Hey. You know, it is it's wild to see that Heroku's ahead of Cloudflare and DigitalOcean.

Yeah. You Node, it's not it's it's not surprising to see it ahead of, like, some of the smaller smaller guys like, you know, Fly dot io or or Render or or Railway or 1 of those. But Hetzner is is starting to get up here. Node Hetzner the big challenge with Hetzner is even getting an account on there. I I Wes like, they their security is so intense. I had to, like, send a DM and, give them the blood of my firstborn child to get an account on there. But I I do have a server on Hetzner now, and it's it's really nice. So I'm a big fan of Hetzner because I have an Yarn an ARM based server on there, and it it's super slick for really cheap. Yeah. That's great. What else is on here? Azure Google Cloud 1918. I think

those are are a little bit more, like, corporate, I think. You know? I think a a lot more people host on those than we think, but, generally, that's, like, the infrastructure team running those, and they're not taking JavaScript surveys.

Yeah.

Usage. TypeScript, JavaScript balance, 32%, 100% TypeScript.

80% TypeScript. Skewed way more to TypeScript than I was expecting. Yeah. Yeah. I remember last Yarn, we were surprised that pnpm every single year, it it goes. And, like, if you were to look at, like, 8 and 9, the top 2 ones, they're so much bigger than the rest of the respondents here.

So it's it's almost at a point where I would say, like, most people are using TypeScript.

Which is it's wild to think. You know? I mean, given that when we started this show, it wasn't a thing. And then, like, a few years into the show, we were like, which one's gonna Wes else is it gonna be Flow TypeScript or Reason? And at that point, it wasn't even clear. And now it's like TypeScript is so ubiquitous. I I don't even I don't start a project without TypeScript these days. Yeah. And I haven't for a while. And and, honestly, I I don't necessarily get the grumbling about it. To me, it's like TypeScript is a fancy linter that helps me out when I need it.

That's pretty nice.

9% of respondents don't use TypeScript.

Like like, a 100% JavaScript.

So yeah, that's that's good. I was very, very impressed.

JavaScript usage, 98% doing front end dev. That doesn't track with the JavaScript runtime ones we had earlier.

Yeah. That's right. Yeah. 67% back end, 27% mobile, 20% desktop app.

That's really cool. We're having the Tubi folks on, which they build a app for all the TVs.

Like, if you have a Samsung or LG or whatever, if you want an app to run on all these TVs, they're all built in JavaScript.

And I am very excited about that. Yeah. III met 1 of the Tory guys at, JS Nation. We should have him on the show. Oh, awesome. Yes. Yeah. Yeah. Yeah. Cool. We're talking quite a bit about the challenges and some of the cool stuff they're doing over there with with desktop apps. So, you know, III like building desktop apps using JavaScript because you can lock in the specific features. You know? If you're you're shipping on a web view, you know exactly which features are there. Browser compatibility goes out the the window. You don't have to worry about it as much.

Did you look at the missing features in in JavaScript?

What would you say the number 1 missing feature in JavaScript? I I didn't look at it yet. Okay. Number 1 missing feature, I'm gonna say types.

Yes. I was gonna say this about the conversation we just had about TypeScript being Okay. So you're making gonna say it, but I was like, I don't know. But, yeah, built in TypeScript

50 7% of respondents said types. Number 2, standard library.

3, better date management. Hey. That's coming. That one's gonna be solved, folks.

100% of you. You can use the polyfill today. It's great. I feel like the standard lib and the date management

have gotten so much better in the last 3 or 4 years.

So I'd be curious to see this versus previous years or or we'll take a look at it next year to see because, man, standard lib the Lodash usage, we didn't go over this 1, but Lodash usage is still date functions and Lodash is still huge, huge, huge.

And not that you shouldn't use those. They're they're great libraries, and we use them on as well on on things, but I find myself reaching for them less and less, not because I don't like them, because I I don't need them.

Yeah. Yeah. I don't need them is really what it comes down to. What what's wild to to me about this is that for all the hype around signals, it's only 26 people responded saying signals.

Only 26 individuals responded saying signals when when there is, like, serious talk about adding it to the to the language. So that's crazy to me. I want I want signals. Pattern matching is really high up in here too. Pattern matching is great. We've seen pattern matching being really nice and rust. A lot of people have noticed that. So I I think pattern matching would be cool. Observables, pipe operators are in here.

Immutable data structures,

26% on them. The pipe operator, I want all the time. I want it all the time. Yeah. Sometimes you're doing, like, a map filter reduce, and you have this beautiful chain.

And then you wanna, like, stick that in a variable, or I do this a lot in the in the JavaScript console, in the browser Wes I wanna copy the output. And there's a function in the the dev tools called copy, and you can just wrap it in a copy, and it will throw it in your clipboard.

But in bash, you can simply pipe to p b copy, and it will throw it in your clipboard. And I often think, I don't wanna wrap this whole thing in a copy. I want a pipe to Scott.

And, I would I would love love to have

the pipe operator. I I wonder where that's at. Bro, I'm just looking at this, and the proposal pipeline operator feels like pretty stagnant.

You know, you you I don't have no idea what the conversations around this Yarn, but, like, the official changes thread Wes last updated, August 2022.

Yeah. What's up with this? I think we should get somebody who knows exactly what's going on. You can you can often go through the t c 39

meeting notes to see. Yeah. So it's at stage 2, which is just in draft mode.

So maybe sometimes these go away for a while because they say they they go through the stages where they say, okay. You know what? This is a good idea to add to the language.

Now let's go off and draft the pnpm. And that's that's what takes the longest JS what does this look like? What happens in this case? What are all the test cases? So maybe it's in a spot right now where it is being drafted up. Yeah. It just feels like

most of the comments or anything on there are, like, from 2021, 2022.

You Node, I know a lot of this stuff happens offline. So or off of these GitHub issues. So it'd be really interesting to see what is actually going on. If if you're someone out there and you know the status of of what's going on, leave a leave a comment or hit us up on on Twitter or whatever. I'm I'm very interested to know what the status of this

is. Industry sector. I find this interesting. Like like, what industries are people working in? Right? 46% in programming and technical tools.

Does that mean that half of the people that take this thing are just building more JavaScript for the JavaScript people? Oh, yeah. That's kind of that's kind of interesting. Programming and technical tools. And then 33% consulting and Vercel. That makes sense. Right? You you you build apps for clients, 27% ecommerce and retail. That's something I think we don't talk about enough. It's, like, how big Scott is in this industry. Right? It's giant.

Sure. 17% finance, 16%, education, 16%, marketing sales and analytics tools. I would have thought that would be a bit higher. I think I think people who work on apps put themselves in that first 1. You know? Like, if you work for a company that builds an app for insurance, you know, quotes, you know, where would you put Vercel in there? Insurance. Probably technical tools. Oh, is there insurance on here? No. I don't Node. There is.

49%.

Okay. But insurance is big. Remember? We met, Christy from yeah.

Christy from where does she work again?

The the,

Imu 1. The Limu Imu 1. But, yeah, had they had something like 3, 000 developers that were deploying serverless functions every single day.

Damn.

Big world, insurance. A big world. I know. Yeah.

For real. Health care, media, government, social media, cybersecurity, 4%.

So most people working on I think it's safe to say that most people are either building apps for people or working on applications for customers or building an Scott website.

Yeah. Yeah. For sure. Cool. Well, that's really interesting. I love this survey every single year. Again, it gives us a good, indicator. You Node, we can all get ourselves into little bubbles, and it's easy to to, like, be in your your Twitter bubble Wes everybody JS suggesting that, this thing is the 2nd coming or something, and it's incredible and whatever. Yeah. You look on here, and it has 10 respondents saying they use it. So it's important definitely to to to get your your mind in a recalibrated way with these kind of surveys. And, also, be cognizant of the fact that, hey. The survey is a result of the people who responded to the survey, not a result of every single person in our industry overall.

Can we do the finish up with the awards real quick? Or we yes. Do that. Let's do that.

Wes both get to guess. The most adopted technology, ordered to the technology with the largest year over year progression.

Vite.

Oh, yeah. I think it was Veet last year too, wasn't it? Scott be Veet. Yeah. Most adapted technology. I'm gonna say bun.

Veet, you got it.

Highest retention awarded to the technology with the highest percentage of returning users.

Veet.

Just riding riding Vite on this thing.

I'm going to say Next. Wes.

Vite.

It's doing me good. 2 for 2. But runners up. So 98%

of developers are using Vite again.

Number 2, 96% is Vite Wes or Vite test.

How do you say that? Vite test. Yeah. And number 3, Playwright. 95% are using it again. That must feel good to have people love your your tech so much. Highest interest. Awards says technology developers are most interested in learning once they are aware of it.

That's a tough 1 for me.

This is a tough 1, because, you know, so many of these little things, they deal in small numbers even if it is. And it's, like, often small little, like, feature.

Yeah. I'm going to say bun.

You're gonna say bun. I'm gonna say, solid.

V test.

Oh, V test.

Really?

V test. Most write ins awarded to the item with the most write in answers. Okay. So this 1 can't be beat.

It's gonna You know? It's gonna be beat. The most write in answers, oof. Oof. Oof.

I don't know. III couldn't tell you Deno.

I'm gonna say quick.

Bun. Ah, I think that's a bun for all of them.

So what time you did say bun? Oh, good. That's great. Runner-up and Amber, 81.

Wow. Amber still not. Expected.

Yeah. Well, I think that's that's a that's a almost a bummer that it's not even an option, to choose, anymore.

Yeah. Most commented library are a a worksheet library which received the most comments. Next. Js for sure. React or Next. Js. Yeah.

React. Yeah. Yelled it. Yes. And most loved library. Words to technology with the highest proportion of positive opinions. Vite is Veet.

It's all Veet.

Yes.

Wow.

Wes in React, number 23.

Okay. Cool. Okay. And, oh, that last 1, was Next. Js was number 2 and storybook number 3. Sick. PIP people loving it. Awesome. Well, that was super fun. I really enjoyed that. Let's move into some sick kicks and shit. Wes do,

number 1 on the podcast once again, guys.

Syntax.

Woo hoo. Yeah. So shout out to everybody who wrote ESLint in here or at this.

Yeah. Thank you so much for that. That's it's it's incredible every single year to see us be at the top here. So it's shocking. So thank you so much.

I appreciate everybody who who does that.

Yeah. There's so much more stuff on here. A lot that we didn't even go into, so check it Wes. 2023.stateofjs.com.

They run state of CSS, state of HTML. The HTML 1 this year was super interesting because you think, like, HTML, like, what? You got a new tag? You know? But HTML has so many new interesting, features coming up, and, big fan. Big fan. Alright. What do you got for a sick pick today?

Sick pick. Hey. I I feel like I have a lot of sick picks because there's some of them that I wanted to share or wait until you were back because I thought I felt like you would specifically appreciate them. 1 of the cool things that I got was a digital and I don't know Node digital JS the right word. A a rechargeable lighter.

So, you know, you have those, like, gas lighters you click and you light a fireplace or something like that. I got 1 that is a USB C chargeable lighter, and it just has a little electricity burst between like, a little taser between 2 posts.

And they're, like, $15 or something. They're super cheap, charges with the USB c.

You can light your fire pit or your, fireplace. If you got 1, you can light birthday candles, anything.

And you never have to worry about it running out of gas or collecting those around or any of that stuff. So there's a ton of them on Amazon or whatever. I'll link to the 1 that I got. But, man, I've been using this thing all the time. It's 1 of those, like, little cheap cheapo gadgets you can get that you could just have around the house and never have to worry about having an actual, like, flame based lighter around there again.

Oh, that's awesome. I many, many years ago, I was just looking back at 20 episode 2/14, I picked a torch lighter, which is I think it's meant for lighting cigars.

It doesn't run it's not electrical, but it runs on butane, which you gotta refill it every couple months. But my wife loves it because she lights a lot of candles, And it's just like it feels really nice. It's like aluminum, and you don't have to, like, fuss around with the click click click click click. Like, it always works. It has a really nice strong to to light it up and even lighting the barbecue. It's been such a nice thing. We've had it for probably 4 or 5 years, and every time we use it, we're like, this is so good. You might like 1 of these things too. I I when I said it was $10, I I was wrong. It's actually 5.99,

and it's a little, like it's long, so you don't have to be near it when you're lighting it. And Oh, okay.

Oh, I see that. It kinda rules. Yeah. So you maybe maybe a nice little candle lighter for you, and then you have to deal with flames when you're starting your flame.

Yeah.

So, well, you got flames anyways. Wes. But I don't. I Yeah. I think this would be really nice at the cottage because it, like, will always work. Right? Especially when it's windy. You know? Like, you don't have to worry about the wind blowing it out. That's always super annoying with those crappy little barbecue lighters.

I'm going to sick pick a toy that I stole from my kids, and that is you might have seen me playing with it. It's called a monkey noodle.

And if you fidget at your desk, like I do, it's nice to have a quiet toy rather than something. Sometimes I've I'm playing with stuff here at my desk, and I'm thinking, I'm sorry, Randy. Like, you probably hear me dropping something on my desk, and it's it's it hurts the audio. Right? And my kids have all these, like, little you know, those, like, Poppets and what what are those called? Like, sensory toys? Push them in. Yeah. Yeah. Yeah. Yeah. Yeah. And they have these, like, monkey noodles, which is just a little silicone booger.

And I love just playing with it and just tying it and moving around. I've been playing with it this whole episode. So if you need some sort of something to do with your hands while you're while you're working, check out grab a monkey noodle.

Yeah. That it's it's not surprising that you would like that. And in fact, somebody left a comment on 1 of our videos when I was with CJ here in the past couple episodes, and they're like, since the move to video, it's become increasingly obvious that Scott has ADHD.

So it's like

we moved around the whole episode. Yeah. Oh, yeah. The video. Yeah. Another thing is, I had a chipped tooth for I had a chipped tooth for, like, probably, like, 8 years, and I never really got it, fixed. And then I whenever I see a TikTok, I'm like, oh, I should probably get my my chipped tooth fixed.

And, somebody commented, Wes, did you chip your tooth? And I was like, yeah. Like, 8 years ago, and I'm getting it fixed next week. And, I just got it fixed this morning. So What? Oh, yeah. That's crazy. They're just so wild. Yeah. They can just add a whole bunch of tooth stuff on.

You got a tooth. I I chipped mine on a, PETA chip.

Oh, yeah.

Which which 1? Like a front 1? Yeah. Well, the the front 2 ones. I chipped out another Stacy's giant pita chip that was apparently like a piece of concrete.

Oh, man. And so did you do they have to, like, build it up? And then Yeah. Yeah. It was right on the edge too. Yeah.

I know. It sucks. It's wild.

Cool.

Shameless plugs. Check us out at the Swag Store. Go to century dot shop or go to syntax.fm and click on swag in the top, corner. We've got all kinds of really cool stuff. We've got these new t shirts. There's 4 new t shirts that are gonna be up there. We got a couple skate decks left. We got a couple basketballs.

We got the coos the can koozies, on there. We are all out of Yetis, which we realized we were selling the YETI's for cheaper than you can buy a YETI in the store.

So they're all Node, but we have some can koozies and and whatnot, and we've got some really I almost just said what what we what we ordered, but we have some really cool stuff coming down the pipe as well. So check it on out. Grab some stuff before it's all gone.

Sick.

Alright. Thanks for tuning in. Peace.