2024 Predictions

You're listening to Syntax, the podcast with the tastiest web development treats out there. Strap yourself in and get ready. Ready. Here is Scott Tolinski and Wes Bos.

Welcome to Syntax. Today, we have our 2024 predictions for web development, JavaScript, rid Dripped, server side, frameworks, tooling, CSS, the ecosystem, AI, all in in general, we're going to be talking about things rid we think will happen this year. We do this every single year, and we revisit it in December to see how right or how wrong we were. And rid. Quite honestly, we we nailed it last time, so I'm going a little bit more wild with my predictions this year just to Let's see if I can manifest some of these

things. How are you doing, Scott? Oh, I'm doing good, man. Feeling, you know, ESLint the Christmas spirit, I know you're getting this one in January 3rd, but we're recording this the week before Christmas. So, Hey. I'm I'm I'm feeling it, man. We got the the elves on the shelf. We're just hopping around the house. Our kids are going absolutely feral, And, we are doing our best to hold on before the holidays. So, yeah, that's how I'm doing.

But, hey, if you're getting wild with your Node, like we're getting wild with these predictions, You're gonna wanna have a tool to help you solve your bugs.

Not only solve your bugs, but solve your performance issues, Solve, your your questions, your mysteries that have you have going on in your code base. Hey. Why does this thing function this way, or why is 1 user in particular telling me this happens when they go to the website.

He did all that kind of visibility and more with a tool like Sentry, which provides things like session replay.

It provides error and exception handling and tracking services. It provides performance metrics, query metrics. You can see rid how fast or how buggy your thing is at any time. So if that sounds awesome to you, it is awesome. Check it out at century.io.

Alright. So we're gonna Start with JavaScript and all of our predictions. So you wanna kick it off there with the first prediction?

Yeah.

Yeah. I think types in JavaScript will have some movement.

You Node, we we said this in our last predictions, and sure enough, there was, You know, movement in some regard. It's in stage I believe it's in stage 1 now.

I would like to see more movement, more Public movement. More movement where people are talking about it. Honestly, I don't know what the effort is to take to gonna get this to stage 2, so I don't wanna just say, oh, it's gonna be stage too. But I would love for this to keep marching forward because I think this is a worthwhile API that I personally, you know, am in favor of. Yeah. I don't think that

they are in any rush to get this thing out the door because This is forever. And we saw this with JavaScript modules. We saw this with promises Wes it took years years of deliberation and going back and forth on the API, because, like, rid. We might be writing JavaScript in 10, 20 years. We might not be, you Node. But if we're gonna go as far to add types to the language, rid. And the the proposal for everyone listening, the proposal JS called types as comments, which Yep. Essentially means that it's not A type as a comment. It doesn't like look like JSDoc. It looks like TypeScript.

But the browser simply will We'll render out the JavaScript or the when the browser parses the JavaScript, it just takes the types and throws them out the door. So So the browser JS not actually going to type check your JavaScript, but it allows us to author and use tools that can parse that specific data. So, Yeah, I bet we'll see some major movement this year.

There really wasn't a whole lot talked about it last Yarn, and rid I bet we'll we'll see stage 1 at least 2 meetings where they talk about it as well. Yeah. Wait. I think it already is in stage 1, Wes. Is it? Yeah. Because stage what is stage 1? Stage 1 is where it's proposed? Or No. I think I think of when it's been We always see. We always talk about it. Here we go. Stage 1, which is make the case for the addition, rid. Describe

the shape of the solution and identify potential challenges. So And I think they changed the name from types JS comments Even though we just did that whole message, because it's now listed as type annotations, which I don't know when that change was made, but I don't remember seeing it rid being listed as type annotations before, but I can do Well, we have

Wes can go back and look at the source.

March 31, 2022 was renamed. So that was almost 2 years ago. It was renamed TypeScript comments The TypeScript annotations? Crazy. This proposal is over 3 years old. So, yeah, it's gonna gonna be a long time. Rid. Time flies when you're having fun. No kidding. Next prediction I have here is temporal API will ship in Node browser. So temporal is the new Data API that will likely replace a lot of your use of the Data API in JavaScript.

And it has a whole bunch of really nice utility functions for working with dates, the ability to add days, add months.

Ready. So it's like it's kind of like the data API we've always wanted a little bit of like moment JS added into it.

And we're starting to see like the temporal API has been around for a while. They were waiting on the new read. ISO specification for how to basically time stamps that include what calendar you're in and what time zone you're in. So That's been approved. So we're starting to see a drop in Safari. Safari has just pushed out a couple. Every single new version of Safari technical preview. They've added some. So, like, that's it. Not in Chrome, not on Firefox, but I bet this year we're going to see one of these browsers say, you know what? We're done with the temporal API, or at least rid I don't wanna say 80%. I'm gonna say the whole thing is gonna be shipped in at least 1 browser. Yeah. And I think at this stage, if you're looking at this, it could be Safari given that they have a big head start at this point. Yeah. Next one I have here is perf tooling gets easier to understand.

I think that rid.

I have this in conjunction with a little bit of perf pushback because rid. Sometimes people just go gunning for the lighthouse score, and that's it.

And Sometimes life is a little bit more complicated than that, so I think people will start to push back on just this Node metric.

Rid. And I think the reason they'll be able to do that is because they'll be able to have a little bit more tooling that is easier to understand, Flame graphs and whatnot Wes it's like, all right, regular people can can just look at this, or you can run some sort of tool through your output, and it'll tell you Specifically, what is causing the issue? Yeah. Get

some AI tools in here. Hey. What do I do to make this fast? Yes. Exactly. Next 1, you Node, next next 1 is the this is not related to JavaScript entirely, but CSS will continue to add features that will make JavaScript less necessary, which it has been and continually has been. It has.

Yes. It has been. It has a guy doing a lot of work there. This year.

Yeah. We has had a lot of those.

You know, I it's it's been endless here. And what's rid to me about this debate is that, like, JavaScript haters will always use this as being like, yes.

The browser can do so much to you oh, you know, if you only you knew how to use the browser, you wouldn't have to use JavaScript.

And meanwhile, it's like rid. Three APIs that had landed that single day, and they're you know? The CSS has gotten a lot better. We'll just say that.

And because it's Continuing to get better, we get to use less JavaScript, and I think that's a good thing for everybody, even people who like JavaScript. Because then we can use JavaScript to do cooler things, like, You Node, these neat midi APIs Yarn, you know, artificial intelligence stuff. Who knows? Let's talk about frameworks.

I believe 2024 is the year of the server. So 2023, we started to see a lot of these rid Framework say, maybe we should start to render on the server. Now we're starting to see, obviously, the big one, React Vercel components, Saying, by default, everything is rendered on the server, and then you have the ability to opt in to client side with the use client tag. It's it's gonna be a bit of a rough move, I think, for a lot of applications, but I rid I think it's gonna be a big year for people moving to the server. It's gonna mean snappier applications. It's gonna mean a little bit less 100 layers to get your data to the the actual client. Yeah. You know what? I think an a big thing is,

you know, we'll see some continued trends Like, AngularJS has just been kind of quietly shipping and maybe not so quietly at You mean Angular? Rid

Don't you're gonna are you doing that intentionally to get the pitchfork? I don't know. Confused.

I we talked to Mark Jackson.

I got the the whole breakdown, but, you know, it's just Wes you got 2 things in your brain, right, that are so closely connected, I I don't think I'm ever gonna get that right. I think I'm gonna doubt myself every single time I say the correct one. So this is my specificity, JS Angular rid. J s.

I think I think, you know, Vue is also going to continue to ship, but, you know, one of the things that I think is that It might become more standard and standard for them to have a compiled time mode like a Svelte or a Solid Or a million JS because JS we've seen, the perv seems to be there. I I just can't imagine given how Vue Hence, to take the best from all worlds that this won't become the the norm for running Vue at some point.

On the same regards, re Svelte version 5 and Svelte version 5 is going to ship, obviously, but it's it's going to be very, very, very fast.

It's going to be convincingly fast. Salt Kit version 2, I think that's already out, so, rid You know what?

I I upgraded our it just came out, like, a couple days ago, and I I tried to upgrade our site rid to it, and it's we're pretty close. There's one thing with a plug in that we're just waiting for a new version to ship, but the, like, the NPX rid. Whatever upgrade that SvelteKit put out, it's amazing. It's just like

it updated our code in, like, 18 different places, and, Yeah. All just worked. Let me just say one quick thing about the Svelte community that I sincerely appreciate beyond other communities.

When they issue breaking changes, almost every time they issue a code mod, and I've had those code mods do great work for me. So People who are out there shipping code mod utilities, here's another Node. Sentry, our employer.

They they have a, a wizard to install Sentry on a SvelteKit site, and it just modded my code for me and put the code with my, all of the relevant things that I wanted Directly for me in the right files, and I didn't have to do anything. Hey, man. If you have a a CLI wizard or a code mod, That's a a pretty sweet value add to me. Mhmm.

Astro, what's our prediction there? I think It's funny. We've been talking about Astro many years, and we've seen many iterations of what Astro JS, but it seems like recently It's been popping a Scott lately. Now they're like a full stack framework. You Node, a lot of people are like, this makes a lot of sense to me. So I think, Astro's gonna have a good year. What about you? Yeah. It's tough to say because I think Astro is compelling,

and I think rid. People who use Astro like it. For instance, I was on that 4 devs one app idea show on Jason Langsdorf's YouTube channel, And most people used Astro even if they didn't use React.

And so I think devs are using it, and they're liking it.

But it's all about if that usage will continue to grow into something meaningful because there's always going to be Frameworks that do something compelling and interesting that just can't get the market share. And Astro's fighting a big battle, but the Coolest thing about Astro is that, again, you can bring your own front end framework.

And there's times where I felt like maybe Astro would have been the right thing to build on because, You know, there's so much flexibility there. That said, you know, I think it's a it's a great platform. I would like to think it would pop, But I'm not sure it does just because

I don't know. Yeah. Is it enough to draw people away? It's Scott it's true. It's gotta be enough. Rid. Like people, you hear that a lot. You hear a lot of developers complaining about React. But what what are all the big companies using? You know, it's It's React. And is it that much of a pain point that people are going to want to switch? Like, is it gonna? Rid. I felt like that was the same thing with, like, 11ty and so Sanity. A lot of these frameworks were like It's popping in the developer community. And of course, there's always going to be a bunch of like use cases where somebody used it for a big Scott site or something like that. But JS it going to push past that into a, like, holy smokes, someone's using Astro for X, Y, and Z, you know? Yeah. Yeah, totally.

We talked about the React Vercel components.

Yeah.

I think React server components have been around for what is it? Probably like 2 Yarn. They've been talking about it. Right. Go back to One of our shows, I think it was we did it right at the end of the year, but now finally seeing it ship in rid A few different frameworks.

So obviously Next. Js is is pushing it out now.

I have a prediction about React, and it comes from me playing with React Scott, state management and then playing with Coming from the Svelte site that we built, so enough experience in both. And the one thing that I really still miss from Our Svelte gets site when going into something like Next. Js is the the import export State management, being able to have simpler state management tools for storing all of your data. And part of The React Vercel components is this form actions API that allow you to know when a form is submitting, rid. What function it needs to call on the server Node, any errors that happen. And I Wes, like, kind of, like, looking at that being, like, wow. Like, that's the first Time that React has really gone past it being a library to render stuff and handle clicks into, You know, something's happening on the server. How do you communicate with the actual server? So, like, I I almost wonder, and I'm predicting that rid. React is going to maybe not roll out because it takes a lot longer than that, but we're gonna have something different than use Scott and use context

or being able to manage our our our data. Maybe it's maybe it's signals. Yeah. Well, you know what? I've I'll I'll chime in on on this and maybe augment it or add on how I could see this happening.

I think if React rid. Themselves were to announce this, you they would announce it, and then maybe, like, 4 years down the line, it would be really because that seems to be the flow these days. But I would say If anybody's likely to do this, maybe that seems like a Next JS territory or a Vercel launches A next Scott. You know? Something that that I wish they would.

They they obviously see how signals are A really good pattern.

They're clearly a good pattern. Everybody's using it, and even signal or observable based state In React world, itself is very popular, but it's not the Deno facto. And I think that is something that Svelte really has going forward. It has total control over those things. And once you all see, the new runes rid APIs, Wes, because I know you you saw the initial runs on in Svelte. The initial, announcement of runs, which is basically their Their new state management based solution.

Mhmm.

Once you see the updates to that, I think you're gonna be saying, Dang.

This is what React needs because the way that they're implementing signals inside of Svelte right now for the next version of it. It's it's I think just time over time again Wes I'm working in state and anything, I think rid Really love to have those ruins right now or any sort of signal based architecture that's a little bit nicer than what I'm working with. So Yeah. Yeah. It would It would be interesting to me, and and honestly, this is one of my biggest annoyances with rack React overall. You never know what you're getting into with when you open up a React project in terms of state. Is somebody using a third party solution like Zoos stand? Are they using something like, context and possibly using context poorly, or maybe they're using context extremely.

You know, there there's so many different ways to tackle state in React.

You know, you never know what you're gonna get. Rid Yeah. I totally agree. And and here's another point of why I think it's gonna happen JS Vercel is sponsoring this project called Waku, w a k u. And this is a new React framework based on React Vercel components. So from the ground up with it. And, the author of Wacko is also the author of Zusan rid In Jotai? How do you say that? Jotai. Yeah. Jotai.

And both of those are React state management libraries. Maybe Vercel JS saying, you know what, let's test it out on this wacko project.

Rid.

And if that's really the way that we need to go JS Scott to say that this wacko is going to be implementing state management, of it. But you got to think, like, if the author of Zustan and Jotai is like, now it's time for a react framework, You gotta bet that those things are gonna be, no, work really well together.

Is there 1 1, person or group of developers who I would trust to make a new React framework. It's these folks.

Yeah, so cool. I'd never heard of Wahoo until this moment. Now I'm, like,

Very excited about this. Yeah. It's, I I don't know if I have the energy for another React framework, but I am rid I am excited that somebody is is looking at it also. So JS well, Remix.

I predict Remix is going to move away, and maybe I shouldn't say predict. I feel like this is going to happen, and I'm pretty sure by reading all the rid. GitHub threads and everything is that Remix will move away from page based loaders, which is kind of what self SvelteKit does Two component Node. Because I was talking about this on Twitter where, like, I love the page based loader, But I hate that it's attached to the page.

Yes. And then I want it to be attached to the component.

And Remix is moving more towards the Next. Js way of doing it. So I predict that there's kind of 3 major frameworks now sort of working on this React Vercel component, so we'll see what comes out of that. I would love

rid. If SvelteKit implemented component based loading functions, I have no idea if that's ever even a possibility, But that's something I would really like. I don't I don't even know if they're talking about that, but yeah. So let's talk about server side JS. I think something that we we started hearing about even about halfway through the Yarn, and the first time I heard about it, Wes dropped it to me, Hano JS.

I think this is gonna be one of those projects that becomes more ubiquitous, more used. I think, you know, Hano JS has a lot of rid Positive features about it, and I I believe that as a compatible, fast web framework that works anywhere, rid. Kinda feels like this is this is Scott be the one you pick. Right? Yeah. I

so for anyone who hasn't heard us talk about it a 1000000 times, Deno is the new express, meaning that we now have browser standards for request, response and fetch.

And all of our servers need to be built on top of these standards. So remixes, Astro is, Next. Js is.

Every single time that you write, make it. Sorry, SvelteKit is Every single time that you have a server side request or response route, they are they all add a little bit of extra stuff on top, which is fine. Deno JS does as Wes. Some nice stuff for sending JSON and and for handling errors so you don't have to do the whole headers. That's kind of annoying.

But like this is it. This is the standard for doing it. And you can take a function that is supposed to run-in SvelteKit on the server and move it over to NextJS with little to no movement.

And rid.

Deno.

You're gonna be able to play it Node there's there's a Node. Js adapter, which I think is hilarious. You can you can adapt throw it to sorry. I should say, I think it's hilarious because It's not Node. Js by default. It's anything by default. And if you want to run a Node. Js server, you have to Npm install the Node. Js server for it as well. So a big fan of it. I reach for Deno every single time that I need something, but I also find myself saying, Do I need HANA, or do I just whip up a Next. Js, Remix, or SvelteKit site? Just the these meta frameworks are almost better because Now I have

file based routing, and now I can I can do my client side as well? Yeah. Yeah. And and and then at that point, it's like, what are they using? And that's what you're using. Right? Exactly.

My next prediction, Node will introduce TypeScript support field loaders. So I think rid. We've talked about how this is technically possible, but I bet, especially all the heat that, Deno and Bun have been given Node lately.

I bet like an easy win is for Node to just ship built in TypeScript support via a Node, and it's fine because they don't have to make any. Rid. You can use your own loader if you really want to, and you can opt out of it if you really want to, or you have to opt in to TypeScript if you really want to. But, rid Like, I was thinking about that the other day. It's like like, Npm is not part of Node. You know? NPM is is a totally separate project, And it's owned by GitHub, but they ship NPM with Node when you install it. Right. Yeah. Wes can't you make some like, that's an opinionated decision That has helped the Sanity, and I bet we'll see that in 2024,

Node will ship TypeScript support. I think I think even shipping it as a loader is is the right call, like you said, because then you're not having to essentially

It's it's a less of a a big leap that way. You know? And it it doesn't have to be in there forever or, like, it can be in there forever. But if there's something new that comes along, then it's Not necessarily a big deal to move along with it.

Other predictions here. On the Cloudfly Node, we talked about, rid will there ever be a standard router? That somebody sent me there actually is a proposal for, URL pattern matching, rid Which is basically, like, forward slash episode or like, on the Sanity website, we have a pattern match, which is forward slash show forward slash A random number 0 through 800 and then forward slash the slug of the URL. And those URLs are matched re Deno in SvelteKit Wes you hit that URL. Right? You've ever written Express Jazz or Hano Jazz Route, you'll have written those as well. So rid. There's now a proposal, and I'm predicting that we that will be approved and implemented this year because, again, we're all writing route matchers, Whether it's through file even if you're using like the folder structure routing Mhmm. The way that you name your folders We'll be standardized by this, and I think that is so cool that you could then name your folder specific

things in any framework, and it will be the same. And that API is the URL pattern API. Is that what you're referring to? Yeah. Did you find it? Well, I found a URL pattern API that is shipped in Chrome in 2021, but no support in Safari or Firefox. And and that's browser rid. I don't even know what what node support looks like. The URL pattern standard.

Yeah. Last updated about 2 weeks ago. So there might not be sometimes Chrome does ship things in order to rid Just test them out. I'll link to the actual GitHub, that shows you everything.

That's pretty cool, though. There's a whole Way for matching protocols.

Last 1, BUN releases full Node compat.

Jared Sumner, author of BUN, tweeted this out Their only goal for next year is to get full Node compat.

And it's kind of interesting that everybody is gunning For Node compatant, I don't see that as a bad thing. I see it as okay.

Eventually we're going to be able to say it's like, oh, what browser do you use? You use Firefox, Safari, or Chrome. We're going to say, hey, like, what are you running your app with? Fun, Deno or Node? And there's a lot that needs to happen in the nitty gritty. There's a lot of little edge cases here and there, but I think we're getting pretty close. Yeah.

It is interesting that everybody's Targeting node compat and nodes shipping. Right? Hey. Yeah.

Yes. There's kinda 2 standards. Right? There's the winter CG standard, which is the the The open web standard that we're trying to standardize on, and then there's also the node standard, which is we also wanted to work with node apps that people have Already written. Yeah. Totally.

Next one is tooling.

So, you know, I think there's gonna be some major space here in the linter and formatting space. You know, we've already seen quite a bit of movement here.

Biome being the one used formerly named Node, Being the one that people are actually use using to some degree right now. And another one which we just heard about, o x c, being a really fast one in Rust. However, I did some digging, and they have the OXC project, I think, is kind of dead in the water because They explicitly state they have no intention of adding HTML support.

So are you gonna install 2 different linters And formatters on your system, like Node for HTML support and one for not, I would just stick with Prettier for everything.

Rid So, Ouid, you know, we'll see about Biomm and what they end up supporting because I know their support they have a a nice little support matrix that they're filling in, Which I'll have a link to where where they don't support HTML, but I don't think they've explicitly stated. It says, HTML super languages support once these parsers are available so they're clearly going to add HTML support. Biom Wes. But OXC explicitly states that it's not on their road map, and they have no plans. So if we're gonna do a head to head OXC versus Biome, I think OXC might It has some juice right Node. Yeah. But I think Biome is is the one to win out amongst these 2. Isn't it wild how,

For what? 3 years? Maybe we need to give context.

There was a project called Roam, which was meant to be an all in one Build tool, linter, code formatter.

What else? All of that stuff in 1, and It it never took off, you Node, and Wes they like officially they ran out of funding and they officially started to shut it down, some One of the what's the story? One of the old employees or somebody who had worked on it basically forked it and renamed it to Biome And sort of continued on with the progress. And I feel like I've I've heard about it, like, more than I ever heard about Roam in the 3 years in the last 2 weeks. Don't do you feel the same?

Yeah. You know, there's there's some drama there, Wes. I don't know how much we wanna get into. Mama drama.

Rid. Yeah. Some drama around, who took funding and where that funding went.

And, I think I think a lot of people are, slightly frustrated with the the Rome leadership, I I will say, in terms of Hey. They they had received a lot of funding, and basically nothing came out of that funding.

And this is the same individual that was behind the whole controversy regarding Babbel payments, so I don't wanna get too much into it. I don't know the whole story. All I know is that There is some some drama, and that's probably why you're hearing about Biome now because the the folks who took over Biome were the ones who were very passionate rid about the solutions to these problems. Yeah. And they had worked really hard on these solutions before having, you know, money Get in the way of things. So I think that's probably why. They're they're passionate about it. Now I will say if Biome wants to succeed, they gotta change that logo, man. I I don't know who I gotta talk to at Biome, but they gotta change that logo. Yeah. They had

Prettier put up a big Bounty for writing a formatter in Rust and Biome came in and swooped it, which is amazing. So like they built it. Yeah. It might be the year. It might be the year. Yeah. I'm surprised. Like, if you would have told me when this came out, I was like, This is not going anywhere, you know, and I think I was wrong about that. How about this? I'll I'll have a prediction. Buy them will change their logo.

Okay. That's my prediction.

I like it.

Other tooling, rid I predicted this last year and Wes incorrect. I'm gonna predict it again. We will see a new TypeScript type checker written in Something, Rust, Go, one of these lower level languages. We do have many TypeScript tools, but they simply parse and strip types. There's nothing existing that does does the actual type checking for you.

It's only written in Node.

And for large projects that can get very, very sluggish.

So it's very, very complicated. I do not envy the person that has to dive into all those little edge cases figuring out how TypeScript's type checking works with all these infers and whatnot.

But I bet we will see something with, like, a couple of caveats, maybe, like, no pnpm support or only support for new decorators, but somebody's gotta someone's gotta help us out here. Yeah.

Yeah. Totally.

You next 1 you have here is that lightning CSS pops, And I'll let you defend that Node, but I'm going to say Lightning CSS does not pop. That's going to be my prediction. Alright. So Lightning CSS is

rid A CSS parser, transformer, bundler, minifier for CSS. Right? It's written in Rust super fast, And it's by the author of Vercel. Js, which has long been one of the best to build tools. It sort of Did a lot of stuff Vite did way before Vite was even around.

And the reason why I say that is because it's Going to be built into some of the upcoming tailwind stuff.

And once it hits that, obviously the installs are going to go to the sky.

But also people are going to realize, oh, it's not just auto prefix or not auto prefix. It's not just post CSS in this space.

There is other tools that can transpile and polyfill and all that good stuff with our CSS.

So, yeah, I'm I'm sitting there. I think it needs more spark in that value proposition.

I think lightning needs A spark of something to get people more excited about it. Yeah. You people need to have more of a issue.

That's the other thing. It's, like, are people Yarn people complaining about slow CSS, same with the TypeScript stuff. Are people complaining about slow TypeScript builds? Sometimes, but rid. Is the average developer pained that way?

Who knows? Enough to rip out what they have Yeah. To take a chance on something that maybe rid A lot of people aren't using currently. I don't know.

Yeah. I I think they're you know, I I fully believe And the and the dev behind it to to make it pnpm, but to me, there has to be a little bit something juicier there. Interesting.

I think another thing is you'll hear more about RS Pack and Turbo Pack.

The Web Pack folks are really causing a lot of, You know, conversation online about performance, bundling versus not unbundling. Vite. Is is Vite fast or is it slow? You know, there's been a lot of conversation here, and, of course, it it feels like the the people who talk the most about RS Pack are the people who are working on RS Pack. But I do think that because there's so many Webpack people out there, I think these tools, these Rust based, bundling tools that Yarn used to Webpack APIs, rid I think they're going to continue, to grow or see much more usage because as people really start to implement and have fully working solutions for what they're already using, but have it be instantly fast compared to how slow Webpack is comparatively.

I think they will be, sold easily. This to me, there's enough of a spark here that and and besides, once once Other projects have blend this stuff to a higher degree. You'll just see RS Pack and Turbo Pack take off. And, likewise, I think, Veet 6 will come out. But just like Veeet, it's just gonna be filled with more steady improvements to Veeet overall, and it'll continue that value prop of being just a a great tool. I don't actually, here's a here's a not a bold prediction, but here's a prediction. I don't think Veeet will release anything that will knock your socks off in 2024,

but I think it's still going to be good. Rid Yes. And I'm trying to think, like, what would be Yeah. The the what's the v killer feature? You know? Like, it's it's fast enough rid For us, we have a very like, a relatively large syntax. F m is relatively large site. You know? And certainly, there's websites that are a 1000 times bigger than that and have a 1000000 different mono repos in it. But I'm Just trying to think, like like, what what are we missing? Yeah. I can't I'm not imaginative

enough

to think about what a feature. I am very happy to not talk about rid tooling for an entire year. And be like, you know what? Everything's fine. Let's talk about the actual code and the things we're building. You know? I hear you. Yeah. Yeah.

Very much. Yeah. Who cares about Wes SMRE? That stuff. Just let me write the code to have it work, ship it, go. Yeah. And likewise, you know, the next section is CSS. I think here's a big swing. I think contrast color, which formerly named color contrast. I don't know why they changed it. I think color contrast is better, name, but who cares? Contrast color will land in Chrome. I think that's easy. Now I could say It could also land in Safari. I think Safari is one of that seems like an API they would love to to support. So, I think a big Swing would say that contrast color lands in Chrome and Safari, but, you know, who knows? I think that's a an important API that we need in the browser. Also, relative color. I believe this will land in all major browsers. Why? Because it's already in 2 of them, and we gotta get it. I relative color will save me So many lines of CSS. Relative color may save me more lines of CSS

than any other CSS API has included. Rid I am very, very excited for that to hit all of the the ones.

Scroll animation. So we're gonna see 2024, the year the, rid I don't actually, I'm gonna predict this. I'm gonna predict scroll animation does not hit in 2 production browsers. That's your you said it will hit rid production browsers. I'm gonna say it's not.

Although, Apple Apple loves their scrolling animations, though. Yeah. We don't disagree rid agree on things very, but I found myself disagreeing with you quite a bit in this episode, which I think it makes for for, better predictions, considering we're not just gonna be, oh, we were right. Now we get to say, Scott was right. In fact, scroll animations did land in 2 productions browsers because, as you said, Apple loves their scrolling transitions. I know that. Yeah.

Rid. That's true. Yeah. It is funny. We should say which ones because all the time that I say will land in 2 production browsers, I think you can read that as it will land in Chrome and Safari, Wes Of course, it's already in Firefox. So I'm gonna predict that

just this is just a very generic one, but 2024 will be the year of CSS discovery Because 2023, we got so many new CSS APIs, we can't even list them all. We've probably did like 8 shows this year just on new CSS APIs in general.

And the way that those APIs work, we've explained them. But there's these techniques that you can use them together for UI enhancements for neat coding ways.

We're still finding them out, and I think that's going to be big this year. It's just like, oh, here are some rid patterns that you can use with these new CSS properties to make better, more accessible Yarn mode, light mode websites. Yes.

Yeah. We think about how long we had to find every little pattern or interesting rid thing and see I mean, how how long did it take to for somebody to figure out the lobotomized OWL selector? I mean, there are there are people out there who will discover Some wild stuff with the CSS APIs that we have not even begun to uncover. And, Wes, you're already doing great work in that space, but I think people will continue to really push the boundaries there. And I agree. I think some of the APIs that are ripe for exploration, subgrid.

Nobody is doing anything interesting with subgrid right now, mostly because it just dropped and, you know, nobody's really talking about how incredible it is.

I think we'll start to see a lot of patterns for subgrid come out. I think we'll see a lot of people using linear, the linear function for easing curves. That's going to be a big one. Animation composition is a feature that nobody is talking about, but it has support in all major browsers.

As of now, I think it came in June.

I went back and looked at my I did a talk called Fresh Style at Render Atlanta that was all about Here are the new CSS APIs, and can you use them today or not? And guess what? I went through that talk To say, alright. Which one of these are coming? And most of them dropped Wes 2023, and I gave that talk in, like, June.

I think it was June. Either way, The second half of the year brought, like, just about half of the the conference talk I did dropped in the browser. So I think these tools, animation composition, linear, subgrid, even container queries, CSS grid will continue to find secrets. You Node there rid. I I learned CSS grid hacks every single day. So, yeah, there's gonna be a lot of cool stuff. And we're also getting things, like, you can animate them now. You know? Like, just little things

Make it so much nicer.

Ecosystem. Let's talk about, like, browsers and the way that we consume the web, what that will look like. So I'm going to say Safari will ship 3 pieces that make a Wes web app, things that you might think, oh, I didn't think they would ever ship that. You know, one example was like input type of file where you could take a photo.

Rid. The Web Share API. I know that it's still a bit of a mess and Scott just went through the pain, but there Wes some major bugs in the Web Share API earlier this year and They Scott fixed, so I bet there's going to be like 3 progressive web app things That I like I don't think we'll probably ever see notifications on Bos for a PWA.

I doubt it.

I think it's just too obnoxious. I have it on desktop Every time I accidentally allow some website to send me notifications, I Wes like, goddamn it. You know? Like, it's too abused. Yeah. But What users actually want notifications

beyond, like, a calendar app? Like, I I Wes, I don't want anything to notify me. I don't want anything to notify me straight up. Yeah. Yeah, me too. And they're just abused for marketing. Super annoying.

So I bet there's going to be some additional things in that space that allow us to They do it in the browser, and that's a pretty bold prediction given how much people gnash their teeth at Safari and they're intentional rid. Their intent they say they intentionally

removing something. I'm gonna say that the, in in regards to maybe some of the comments about the share API, I don't think the share API is getting any better anytime soon.

That API is a mess because it's beyond It's beyond, like if you use it on desktop, it's it's basically worthless no matter what you're using. So there are some caveats there, but it's almost Not even worth it to say. And then you get into the whole thing of, no. I gotta figure out I gotta I gotta reliably track if I'm on a mobile device now, which In itself is hard to do. User agent's unreliable. There's not a lot of great great ways to do it. In fact, most places are like, oh, if you wanted to check if it's a mobile device, use the device Pixels to deter no. That's not a that's not a good reason to do this because now you're leaving out iPad support.

Because these these mobile browser or mobile OS can handle the share mechanisms way better than any of the desktop browsers can. Mhmm. You can't rely on it. You just gotta kind of ignore it on desktop, and that's kind of a bummer. No. Hold rid. One more thing I forgot with the PWA that shipped this year is you can use Apple Pay on

Firefox on Ios. Mhmm. So like that That was big. And I bet I hope like, can we have Apple Pay on desktop applications as well? As someone who sells stuff on the Internet, I would absolutely love that.

Yeah.

Another trend, I think we're gonna continue to see Fox usage go down despite every nerd on the Internet telling you that they only use Firefox and everybody only uses Firefox.

You ever Node that, Wes? I posted TikTok about anything about Chrome, and the comments are like, who still uses Chrome? LOL. It's all Firefox. And, like And then meanwhile, the next headline is Firefox usage slips to 2%. You know? Yeah. The, like, out of touch nerd is very frustrating

because They don't realize that the entire world now uses computers and mobile apps and Vercel, and That is a bigger much bigger segment than you. And there's just some some areas of users that are insufferable. I'm a Firefox user. I hope that I'm not the insufferable Firefox user, but, like, Android there's a lot of Android users that are in that camp as Wes, or it's just rid Android has just a million oh, and then JetBrains users are the worst. Brains. Worse. Oh, a post anything about Versus Node. Have you considered using this pnpm Expensive, text editor instead of the free one that you're using. Yeah. I'm sure it's amazing, but I'm not gonna use it. I'm using Versus Node. I enjoy it very much. And I realize it's not a full blown IDE, but it does absolutely everything that I would ever want from an IDE.

Rid I don't know. But, yeah, this IntelliJ users Andrew, I'm probably pissing off a lot of people right now, but very insultful. Who else who else can we throw on this? Yeah. Hold rid the skewer pile. Yeah. Let's we've only hit about, probably, 40% of our audience here. Let's Yeah.

I I'm a I'll I will push back on the Android users, because as an Android user, I will tell you JS a former Android user.

I will tell you there was nothing more frustrating than a feature that you had for 10 years. Yeah. And then, and I Apple releases it, and then Bos folks are like, this changes everything. And they're like, Wes,

rid. But it does it does change everything because, like like, half the world can now use their device in a different way. And, like, I realized they probably should have rolled it out Earlier, I Apple takes forever on everything. I think it's not that. I think there I think there's a bit of a head in the sand mentality,

for Bos users who who don't look at other systems Wes I think the average Android user has a better understanding of iOS's limitations and features than the Bos user has of the Android features. IOS users don't know what they don't Node. And Yes. That's very frustrating. I totally agree with that. That's why they're mad. That's why they're mad. I'll tell you that.

You know, a web

the WebStorm folks are are mad because they paid for a product, and they want you to pay for it too. The WebStorm folks are, like, are mad because, like, yeah, they've had those features for a long, long time. And WebStorm's awesome, but, Like, you don't have to tell us that you use Webstorm every single time. It's like sometimes I watch, like, beginner repair tutorials, like rid. Like, soldering copper 101. And then there's a guy in the comments being like, I've been soldering copper for 76 years, and that's been What are you doing? What are you doing watching an intro video? I got some video.

Yeah.

Yeah. Shouldn't you be soldering? Yes.

I also predict Safari will ship.

Well, Node. Safari has Pasky support as well. So, yeah, I think that we will continue to see Pasky.

Rid I think it's a big year for Askew's. I'm really excited to get that episode out as well. I think Yarn will become a paid product, or there will be a paid tier of Arc. We already talked to them. We know that that's in their plans, but I think it's gonna happen in 2024.

I have no knowledge of this, just prediction.

Next prediction, I think we're gonna see a lot more WebXR in that same, conversation we were just happening because nothing happens until Apple does it. Apple releasing the Vision Pro, which even though people laughed at this thing, I think it's gonna sell a crazy amount, for what it is and how much it costs. And I think a ton of people are gonna have their eyes open to XR based content and types of awesome things you can do for that. For instance, I have an Oculus Quest 3 that has good pass through. It's not great pass through, but it's good pass through. And it it doesn't weigh too much. That way, you can wear it from for time and a time, and it comes with a virtual desktop software, Wes. So what I can do, I can put on my headset, connect to my laptop that's in front of me, And make a monitor that's 40 feet tall in front of me and use my actual keyboard in the house. I've done this many times, and it's pretty great.

And that's with a Cheaper, non Apple based device.

I know for a fact that this Apple device is going to be really good in the pass through department, so I think this thing will make people, like, really wake up to what you can do with XR, and I think there's a room in the web for that too. Let's talk about AI

Tooling around web development.

I think that this is going to be a big year for like, it's crazy how quick everything changed in 2023. You know, rid You go back to Christmas time of 2022.

I signed up for Midjourney and stayed up till 3 in the morning playing with it. Yeah. And rid. Here we are 12 months later complaining about Midjourney and then not being good enough, not having a desktop UI And saying, oh, these these other things are way better than it. It's it's like things that I could could never comprehend in my lifetime.

12 months later, I'm complaining about Node.

So I think we're going to have another big year for that type of stuff, specifically around the Generation of websites. So Vercel has vzero. Dev. You can take either a prompt Or a screenshot and it will convert it into the HTML and CSS that you have. There's another tool I've been playing with called Galileo AI. They added me to the private beta, which is you. Text prompt to design, and it's it's pretty it's okay.

Rid. The designs are a little bit limited. And if you look really closely, they all kind of look the same and they apparently can download to Figma. I downloaded a couple of them. It looks Looks decent. It's not not amazing, but it's certainly I never could have thought that this would have existed a year ago.

So I think these tools are gonna get started getting built. They're getting built into Figma. You know, they're gonna like, I don't V0.

Dev is great for a tweet, and I don't see a lot of people using it day to day like we are using the chats. And I think that's going to change this year Wes it's actually a tool we use in our day to day, very much like a Copilot autocomplete or rid. Chat gpt chat.

Chat gpt chat. Yes. Mhmm. ATM machine. ATM machine.

Small models that run-in the browser are going to be more commonplace.

So this past year has been open AI everything.

And then we're starting to see a couple other competitors pop up.

But I've been dipping into the, like, hugging face rid Via transformers JS, which is a little library runs in the browser. And, like, some of these some of these models are 8 megs and you can run text completion or you can do object detection or Mhmm. There's you can do depth detection with some of them. And And, like, I think people are gonna start to same with, like, libraries. You say, you know what? You don't have to send everything to OpenAI via an API request.

Maybe we can run some of this on the server or in the browser because the models are smaller and faster and they can be done rid For free. Word. What else? Easy to train your own model. That's one thing that I think is starting to get a little bit easier is, like, how do you how do you decide, like, You can train your own model. I know there's lots of tools out there, but I think it's going to get even easier.

Many tools are going to become Sherlock'd by OpenAI.

Sherlock was like a finder app for for OSX, and then Apple rolled out their own version of it. It was much better. Right. So that's if you hear somebody saying it's gotten Sherlock, meaning that your app might become

part of the OS. Yeah. Part of the OS or part of OpenAI.

Part of the chat.

Because like the other day I was I was talking to a couple of people on Twitter About, like, how to compare 2 images.

And I was like, oh, yeah, you Scott, like, create them as embeddings, and then you can compare the 2 vectors. And then rid. Somebody else is like, oh, you just send the 2 images to OpenAI, and it will tell you how similar they are. And it's just like, damn.

Like, all that rid Information I gathered on how to how to compare images visually is like I pnpm a lot of time learning that, and now JS the answer is, hey. You just send them to this Node. It'll tell you how close they are.

I feel like that's gonna be our reality for a lot of people.

All the time I spent learning x, y, and z. Yeah.

Oh, well. And then last 1 I have here on prem corporate AI becomes more of a thing, so you're going to start to see your friends who work at These big KPMG or something like that, they're going to have a rack in there. Again, not to say that doesn't exist already, but It's going to be more of a thing. People are scared about privacy. But if you can run a box on your VPN or within your own rid System that has your own AI. It's trained with all of your documents. That will be a major competitive advantage. Yep. Yep. That's rid. For our predictions, we'll do a tiny little hasty coming up in the future, which is your predictions. And then Scott and I will provide our own reactions rid Commentary to it because there's certainly some very good takes and certainly some very bad takes.

Curious to see what they are. Cool. Well,

sick picks and shameless plugs.

Do you have a a sick pick

for us today, Wes? Yes. I am going to pick a robot vacuum that we recently got. So for About 8 months of the last year, I've been having my sights set on this one. I've been waiting for my Amazon affiliate Bucks to add up.

I have my like uses page, and if you buy something on there, I'll get a couple of bucks from it. And Finally, we saved up enough bucks and got this thing. It's the Roborock S8 Ultra.

And I know you I I originally looked into it because you said you have a Roborock. Right? Yeah. I have the s seven,

So I have the last year's version of the same one that you have. Awesome.

And it's fantastic. Ours is named Lomax After, the character from weekend at Bernie's, because we love weekend at Bernie's. And we figured our our robot Kinda bounces around the house like Bernie from weekend at Bernie's, so we named it Lomax.

I like it. Contributed to Bernie. Yeah. So this thing does

Everything. So it basically the dock has dirty water, clean water and a vacuum bag. And the thing will go out and it mops your floors, Sucks up the dirty water and keeps it in some dirty water tank. And then every now and then it will go and empty the thing, rid. And then it it will empty the vacuum itself. So we've had it for probably two and a half weeks now, and we run it twice a day.

Whenever we're done cooking dinner or after dinner, we just say, hey, go mop the kitchen, and it goes and mops it and does an awesome job. Like, you think, like, Is it really going to do a good job mopping and vacuuming? And I'm here to tell you it's amazing.

And you only have to empty the water Maybe once every week, depends on how often you run it and how big your house is and things like that.

But It is so good. It charges itself. It empties itself. It mops itself. It has, like, vibrations built into it, so it'll try to, Like, it's not gonna scrub your floors, but if you run it often enough, our floors have been pristine.

And it makes a huge impact. You don't even notice it until you're, like, Until it stops working or Yeah. You Scott change the bag or something, and all of a sudden, you're like, oh, floors are filthy. Yeah. There's

everywhere and, like And it does the straight lines. So the we have carpet in one of our rooms. So the the lines in the carpet are like like the we we had the, like Looks nice. Super old rid. Roomba for so long where it just bumps into shit until it and and then bumps into shit until it sucks up a cord, and then, guys and there's a camera on it. It's Bernie from weekend rid of Bernie. Okay. There you go.

I've I've never seen a Weekend at Bernie, so I don't get that reference.

But, like, it has a camera in it, which People are scared of. Apparently, the older ones, you can literally drive the thing via the camera, but the camera on this one is not connected to the or so they say.

And it just does it, like, detected my wife's iPhone cord and It went all the way around to the curve of the cord. Like, it didn't just like, I'm not going to go there. It went all the way around the cord and, like, Scott as close as it could to it. So it's amazing. It's like, well worth the money. It Wes like $1500

or something crazy like that. But it's something you save up for, for sure. Yeah.

Yeah. Node more thing about how excited I am about this is the The the brushes inside of it, they are silicone.

And what happens with the brushes is you get hair rid stuck in the in the bearings.

And on the brushes on this one, you can take the ends off and just slide the Disgusting hair clumps off and put the put the bearings back on it. With the last pnpm we had, you had to stick a knife in there and try to cut out the hair Wes you buy New brushes every 6 months because you have to throw them in the garbage, so

big fan. Yeah. I like you know, from the repairability angle, Everything comes off of this. At least it does off of ours.

I can get, like, every little piece off of Yeah. And I think that is Lovely that they do that. Right? It's not a black box. You can open it up. You can fix any clogs. You can really, you know, really clean it. I I I take take mine apart and detail it, you know, once every 6 months or so just to keep it running well. Yeah. Yeah. Give it a nice wax job. You Node? So I I meant before I get into my actual sick rid Sick pick. I will sick pick weekend at Bernie's. It's I'll have to watch it. You Node? Came out 1989.

It has an incredible soundtrack.

It's It's a breeze to watch. It's 97 minutes.

Courtney and I maybe watched this movie, and I kid you not, maybe twice a year, since we we've been Married, maybe even since we've been dating. We watch weekend at Bernie's a lot.

It's a great movie. I'm a big fan.

Rid Sophie, you haven't seen it? How much weekend at Bernie's y'all? I'm going to actually sick pick, though, beyond weekend at Bernie's.

Protein powder that man, protein powder stinks. You Node? I I've been weight lifting a long time, and I always Hate protein powder. I hate the whole process of it. And because of that, I consume way less protein than I probably should given the amount of working on I'm doing. So I've recently found rid A hydrolyzed

Hydralized.

Yeah.

Protein powder that mixes really Wes. ISO 100 by Dymatize.

I I don't go on, like, protein powder YouTube that often, but, apparently, this is a very popular one now.

So In addition to being, like, a great mixing protein, the flavors are really amazing. You have Fruity pebbles, which I mixed with my oatmeal, and it tastes like Sanity pebbles. It tastes like I'm eating fruity pebbles. They also have cookies and cream, like a Dunkin' cappuccino.

Like, they have a bunch of, like, brand deals. This is the cake. The dime is oh, yeah. I have Cocoa Pebbles. This is the ISO 100 hydrolyzed. Yes. Yeah. You have this? Yeah. This JS I bought this in, like Oh, my gosh. It's It's so good. Like, it doesn't make you it doesn't stink, and it doesn't make you super bloated, and it doesn't hurt your stomach.

It's amazing. Really. Yeah. Okay. I'm I'm happy to hear that because,

yeah, I've tried a lot of protein powders. Like, it like, I'm excited to have a protein shake because it's rid.

I I typically dread them. So, yeah, check this one out if you're into that world. If you haven't heard of this before, you need a new protein shake. I was pleasantly surprised. Yeah. That's the one I take. I also have

just the the premade ones in my office. Mhmm. Like, if I can't run up to go mix 1, like, when I'm recording today, I'll grab one of those, but the

the custom Shook ones are the Bos. Yeah. Heck yeah. Cool. I'm gonna shamelessly plug the syntax newsletter.

You can find out about the rid Syntax newsletter at Syntax dotfm forward slash snack pack. We release, 2 episodes or 2 episodes, 2 newsletters a month full of Hot take ESLint, interesting things that we found on the web really highlights from these episodes, and we give you an opportunity to give more read. Feedback about what you like, don't like about the show, as well as, if you're interested in merch. We had our webmaster tea sell out Instantly, if you're interested in being the 1st to know about merch on the store Wes that webmaster tea is coming back, Head on over to syntax.fmforward/snackpack.

The link is in the show notes. You can sign up, and you'll get that rid kind of access. We don't bug you a lot. Like I said, twice a month, so we're not we're not annoying you. Yeah. And by the time you're listening to this, the Webmaster teas Should be back in Scott. They

any day now at the time of recording, this is in December before Christmas.

They are on a truck, so they're just about to be.

We ordered 200 of them because they sold out in like 6 hours last time.

Yeah. Right on. Shameless Bos. Westboss.com/courses.

Check it on Audi. You want to learn a thing or two in the new year rid And, use coupon code syntax for $10 off. That's it for today. Thanks, everybody, for tuning in. Catch you later.

Head on over to syntax.fm for a full archive of all of our shows, And don't forget to subscribe in your podcast player or drop a review if you like this show.