After our recap episode on 2024, it is time to look into 2025 - and share our predictions and even some rather spicy hot takes what developers can expect in 2025. Of course, Michael and Alex once again have the Nuxt team Lead Daniel Roe on the show to get some insights on his takes as well!
Find out more about their thoughts on Vue Vapor, Nuxt 4, Conferences, Alien Signals, Vue 4, Nitro and many more topics that will be (most likely) relevant in this year.
While the focus is mainly Vue and Nuxt, topics around general Web Development are covered too - from AI to Open Source and the job market in 2025.
Links marked with * are affiliate links. We get a small commission when you register for the service through our link. This helps us to keep the podcast running. We only include affiliate links for services mentioned in the episode or that we use ourselves.
Chapters
Welcome to DejaVue
Daniel's Injuries and Accessibility
Fake teeth and other foolery
Biomodding your teeth
Is 2025 the year of Nuxt 4?
React Metaframeworks and the year of Nitro?
Course Announcements
No more Vinxi in 2025?
A prediction from the future
Will 2025 shift Open Source Sustainability...
...driven by the WordPress incident?
Why hasn't there been a WordPress fork?
More amazing Conferences coming in 2025?
Vue Vapor in 2025
The Year of AI Agents?
Alien Signals Adoption in the JS World
Vue 4 coming in 2025?
A Unified JavaScript Toolchain in 2025?
The Developer Job Market in 2025
What are you predictions for 2025?
Daniel's info and Alex' last prediction
After our recap episode on 2024, it is time to look into 2025 - and share our predictions and even some rather spicy hot takes what developers can expect in 2025. Of course, Michael and Alex once again have the Nuxt team Lead Daniel Roe on the show to get some insights on his takes as well!
Find out more about their thoughts on Vue Vapor, Nuxt 4, Conferences, Alien Signals, Vue 4, Nitro and many more topics that will be (most likely) relevant in this year.
While the focus is mainly Vue and Nuxt, topics around general Web Development are covered too - from AI to Open Source and the job market in 2025.
Links marked with * are affiliate links. We get a small commission when you register for the service through our link. This helps us to keep the podcast running. We only include affiliate links for services mentioned in the episode or that we use ourselves.
Creators & Guests
Host
Alexander Lichter
Web Engineering Consultant • Founder • Nuxt team • Speaker
Host
Michael Thiessen
Full-time Vue educator
Guest
Daniel Roe
Nuxt Team Lead and Independent Open Source Maintainer
Editor
Niki Brandner
Audio Engineer and Video Editor
What is DejaVue?
Welcome to DejaVue, the Vue podcast you didn't know you needed until now! Join Michael Thiessen and Alexander Lichter on a thrilling journey through the world of Vue and Nuxt.
Get ready for weekly episodes packed with insights, updates, and deep dives into everything Vue-related. From component libraries to best practices, and beyond, they've got you covered.
Alexander Lichter:
Hey, everybody, and welcome back to DejaVue.
Michael Thiessen:
Your favorite Vue podcast. You maybe just don't know it yet. Although, this is now the 2nd year that we've been recording this podcast. Well, I guess, we started in 2024. Correct.
Michael Thiessen:
Now it's 2025. So we've got we've got a few episodes going. So today, we're going to be sharing some of our predictions for 2025. What are we gonna see in Vue? What's gonna happen in the whole ecosystem?
Michael Thiessen:
Things like that. So you might hear some spicy takes, maybe some boring ones, but, yeah, we think it'll be pretty interesting.
Alexander Lichter:
Absolutely. And but who is we? I mean, you might have heard 2 familiar voices already. First of all, there's Michael Thiessen, the lovely co host of this podcast. How are you doing, Michael?
Michael Thiessen:
I'm doing pretty good. And, with me, my co host, Alex. How are you doing?
Alexander Lichter:
Also pretty great. Can't complain. The weather could be worse. That's always a good indicator, but no. Actually good.
Alexander Lichter:
Lots of new resolutions to to keep up with. For example, of course, going to the gym, maybe some more bouldering. But talking about bouldering, we have a lovely guest here, same the same guest as we had before in our lovely 2024 recap edition, the wonderful lead of the Nuxt, team and open source maintainer and avid boulderer, Daniel Roe. Daniel, how are you doing?
Daniel Roe:
I am I am well. Thank you. Well, I mean, I'm slightly well. I actually got a bit injured injured a few minute a few like, about an hour ago. I popped to the gym.
Alexander Lichter:
No.
Daniel Roe:
And, yeah, managed to managed to cut my hand up a bit. So it it'll it'll heal probably quite quickly. It's not gonna hold me back too much, but there's a small amount of pain.
Alexander Lichter:
Get we'll soon. I hope this doesn't impact your coding abilities at all. Of course.
Daniel Roe:
Well, you know, hands, you know. I I don't know if well, I mean, one of the things about accessibility, peep people quite helpfully say, you know, we are like accessibility is about, making provision for for making the web, making our apps accessible for people who are, who might, for example, be temporarily needing that help. So, you know, if you suddenly lose the ability to type because you have a hand injury, then suddenly you'll start to be very, very grateful that people made something accessible, that you might not have cared about before. So it's, yeah. Anyway, I can I can, I think, still type?
Daniel Roe:
And I
Daniel Roe:
Adn Ican definitely prognosticate. I can wisely stroke my chin. I might not be wise when I come up with any predictions, but, but, yeah, I should be good.
Alexander Lichter:
I mean, it's
Michael Thiessen:
I hope you get better soon.
Daniel Roe:
Yeah.
Alexander Lichter:
Yeah.
Daniel Roe:
Thank you. Thank you. Actually, I got I got even more I got more injured last night.
Daniel Roe:
I I was at a bike accident. So, basically, I I like, at the moment, I feel like just bits of me are sort of in in in in shambles. I I'll be I'll be totally fine, I feel. Is that Like, I even I even have, like I I I clogged my head. Like, I slid on the asphalt.
Daniel Roe:
So I've got sort of, like, cuts on my face, and my other hand has got lots of stuff. Anyway, this is probably not why you invited me on the podcast to talk
Michael Thiessen:
You have all your you have all your teeth, though. Right? That's that's
Daniel Roe:
Well, you know, I mean, I wear false teeth, of course. You know, that's the that's I always have. You know, it saves on this kind of thing.
Alexander Lichter:
Just for the podcast. Just for the audio podcast. Very important.
Daniel Roe:
This is this is not this is not the case. But, I I do I do like, do do you always need to tell people when you're joking? Like, there was that time that April fools joke that I I made about, something like I'm taking a break from leading the Nuxt team, something like that. And then once I have a cup of tea, I'll be I'll be, you know, as ever as you mean again. But I but loads of people never read to that sort of second tweet, and, I I I felt very guilty.
Daniel Roe:
I I felt like people have had a bit of a worry that they might have built their website on the wrong framework. Like, there was there was a a sort of real implications of that, so I always feel really bad. So I not sure there's any real implication of me saying I have false teeth. There's not no problem with that.
Daniel Roe:
Oh, I could tell you, actually, I know people who have taken out their healthy teeth and replaced them with new teeth
Alexander Lichter:
Why?
Daniel Roe:
That are artificial and therefore will never wear out. And the idea is that, you know, it's it's like an upgrade. It's like a body body mod upgrade, that will just,
Michael Thiessen:
like man.
Daniel Roe:
Set you set you for life. So, like, now now you don't need to worry about your teeth ever again. They're just permanently perfect.
Michael Thiessen:
Interesting. That's That
Daniel Roe:
is not something I have
Michael Thiessen:
ever I understand, though.
Alexander Lichter:
It must have been expensive as well. Like,
Daniel Roe:
I'm sure it must have been expensive. And, also, by the way, do you know how it works? Really? I guess this is probably all gonna get cut from the podcast, but
Alexander Lichter:
It won't, no.
Alexander Lichter:
We have it all in, and there's a trigger warning just in case people would not wanna hear how you
Daniel Roe:
You probably don't wanna hear this. The way that it works is they file your teeth. They don't take them out. They file them down to, like, little nubs, and then they attach the new teeth that they have created some in some lab to these, like, little nubs of teeth that you have remaining. Doesn't that just
Michael Thiessen:
That's odd
Daniel Roe:
Doesn't that just sends, like, a a a weird feeling up your spine? I don't know. I I it creeped me out.
Alexander Lichter:
It it does. Totally. It does. Fully. But I think
Daniel Roe:
Fully. Absolutely. Like, you don't feel at normally, I'm I'm very much, you know, hey, body mod. I I I like, I'll upgrade myself. Yes.
Daniel Roe:
I'm gonna stick a chip in my brain the first chance I get. But for some reason, filing down your teeth, to attach, you know, fake teeth, like, I don't it doesn't doesn't do it for me.
Michael Thiessen:
It's like one of the decisions you can't go back on either. Like like, you decide, oh, actually, this was a bad choice. I should I wish I could go back. Well
Alexander Lichter:
Too bad. Right. But, I mean, chip in the brain is also something that doesn't sound too reversible necessarily. So that's still a higher tier for me. But nevertheless, all the all the feelings that people might have now, I hope you won't have them for all the hot takes and the predictions that are coming up now right now, which is where we slowly should move to as we are here in a Vue podcast.
Alexander Lichter:
Yeah. But I always love the little sidetracks.
Michael Thiessen:
Yeah.
Alexander Lichter:
So so in in 2025, a lot of things will happen. I'm more than sure about that.
Alexander Lichter:
In the 2024 edition, Daniel already told us more or less that there will be at least one thing happen. So a very boring prediction. I just wanna start straight away and say, 2025 is the year Nuxt four will be released.
Daniel Roe:
Yes. Alex, I I think that prediction is basically that's, like, a 100% chance of happening. So 100%. Yeah. A 100% chance.
Daniel Roe:
But, well, it like, basically, as long as I remain alive, which might you know, given my bouldering and cycling accidents, it you know, that's not 100% guaranteed. But I think as long as someone from the court team remains alive, I feel like we're we're probably well gonna release it. Like, I I I say I I'll tell you this. Even if Nitro doesn't release in 2024, 25, we are still gonna have Nuxt 4 because there is no way. There is no way.
Daniel Roe:
There is no way.
Alexander Lichter:
Yeah. I mean, we we covered it luckily in a in a
Daniel Roe:
Yeah.
Alexander Lichter:
Recap episode abroad. So if somebody missed it out, definitely Yeah. Check that out.
Alexander Lichter:
There are chapter marks for everything. So just jump there or listen to the whole episode, of course. So, yeah, a very, very simple, but of course, a thing to bring up 2025 as the the year of Nuxt 4. So, yeah.
Michael Thiessen:
I've got a related and also, like, another boring prediction, which is that 2025 will be the year of the new Mastering Nuxt update for Nuxt 4.
Michael Thiessen:
And, so we've got some other things in the works. You might see a few more other things related to mastering Nuxt, but I can't say much more than that. So I'll I'll leave that there. I'll dangle that there for you.
Alexander Lichter:
Interesting. Interesting.
Daniel Roe:
Can I just say, like, this what's the significance of having Nuxt 4? So the what does it mean that it's the year of Nuxt 4? Because in a way, like, we're saying that the features is not about features. So to say that it's the year of Nuxt 4 means that it will be the first chance people will have to see what a major version upgrade looks like for Nuxt, which means that 2025 will be a year of evaluating the strategy and vision of Nuxt. So people will be able to make up their minds.
Daniel Roe:
So enterprises will be able to say, are we happy to build future stuff on this? Are we happy to commit to it if this is what it's like? So it's a very pivotal one for Nuxt, I think. Maybe less pivotal for people using it, I hope. So the the the idea that it's the year of Nuxt 4, probably, it's going to be significant, I think, for the future of Nuxt.
Daniel Roe:
Maybe significant for enterprises. Mhmm. I would I think that and a lot of the stuff on our road map is stuff like multi app, module federation, stuff that's enterprise level. Yeah. Maybe you could revise the like, to say rather than just be the the year of Nuxt 4, it's the year of evaluating the future of Nuxt.
Daniel Roe:
It's a year of enterprise adoption of Nuxt, I I hope. So those might be some some tweaks to that prediction.
Alexander Lichter:
Do you think that Nuxt 4 will drive the adoption of Nuxt JS, especially like in the enterprise sector? Like people see, hey, stable version, don't have to migrate for half a year or longer. Don't need any help. Just can push a button. That's it.
Daniel Roe:
I think that there's a lot of uncertainty, from the 2 to 3 upgrade. And a lot of the reason people are asking so much when all next 4 come out is that they want to start the migration or plan migration in their company's budget. So of time budget. Like, they'll have sprints, you know, new features are being released, stuff's happening. When do we we need to allocate some time to upgrade?
Daniel Roe:
We don't want it to be 6 months. So people want want to know that. Now once once it's clear that it's a matter of, like, a day or a week at most, you know, like, that's going to be hopefully mean that people are gonna feel a lot happier about their decision to be on next. That's what I what I hope. Hopefully, that will mean people be more eager to build new stuff on it.
Daniel Roe:
And and when I say, you know, people will be built you know, it's not raw numbers. It's because the people who use Nuxt are the people who contribute and build Nuxt. So the best thing for the future of Nuxt and seeing, you know, great features and good bug fixes is to have an active group of people who use it. And that's true also from from enterprises as well who will be hitting, like, edge cases that nobody else hits and adding features that maybe nobody else needs. But, you know, it's so it's it's a significant and important important piece of the puzzle.
Michael Thiessen:
I think there's also a bit of, like, a psychological thing with, like, a new major version even if, you know, there's not you can't do anything new necessarily with Nuxt 4 versus, you know, what we have now. There's like this okay. It's the next it's like the new year where, like, you can make your New Year's resolutions at the beginning of December. You could decide to start going to the gym or start bouldering or whatever. But, like, it's this whole, like, fresh start kind of thing, and you think it's easier to talk about and, like, I don't know.
Michael Thiessen:
It's just, like, more exciting in a way in some I don't know. I think there's maybe some sort of psychological thing there too.
Alexander Lichter:
Good points there.
Michael Thiessen:
At least for me.
Alexander Lichter:
Should we should we go on with another hot take? Daniel, Michael, anyone who wants to go first or another another prediction. It doesn't have to be a hot take. It can be a prediction too.
Daniel Roe:
I think it's going to be a year of reaping. And, I think that there's a lot of discontent. This is ongoing in spicy. So looking at the state of JS, well, any of the survey results for the last year, there's a lot of negativity about the current React experience with frameworks. And so I think this is going to be a year where we see explosive growth and stuff like TanStack start, which I think we'll see huge I think people are really excited about it.
Daniel Roe:
Remix also might I I think it will basically be a React framework challenger year. Might not be the case, but I I think I think we'll start to see it. Basically, the, sentiment yeah. People sentiment, developer sentiment, I think, lags the actual change in behavior by some considerable amount of time, like a matter of year or 2. And I think that we might start seeing it from this past year.
Daniel Roe:
People unhappy with, you know, they don't like the app router, or they find that the the deployment story is is limited. So I I say I I think I think I'm right in saying that will be we'll see stuff like Remix and TanStack start. And I hope that means that it's going to be the year of Nitro because obviously, TanStack start built on Nitro as is SolidStart and other frameworks too, like analog. Angular is considering adopting it, which is pretty exciting. So I think that, you know, that is maybe a side effect.
Daniel Roe:
So it's not like individuals are saying, hey, I wanna adopt Nitro for my React app. But I think people who are looking to use a different framework for their React app might find that they're using Nitro under the hood, even if it's not something they're explicitly aware of. And I hope that means there's gonna be a lot of, you know, forward movement, momentum development, excitement around around Nitro and people discovering it and and being pretty excited about it. So that was 2 hot takes in 1. But I think so.
Daniel Roe:
I I think what you'll see is the the beginnings of growth, but I don't think you're gonna start seeing a reduction in people using Next.js because that takes a long time for that to happen. People who already built sites on they're not you don't migrate stuff across because you you build new stuff in in a different thing. So that might be something we start to see.
Alexander Lichter:
I think a lot of good points there. I remember when I checked the state of JS survey, I also of course, everybody saw the downwards trend with positivity in React, same with with Next.js. There was a small number of people, I think, like, 1 or maximum 2% that, like, the the Next.js usage got reduced.
Alexander Lichter:
So not significant significant, but at least before, it was always trending upwards. Like, more people used it. So in a way, we I think we see some change there already, and I agree. Like, everything that has 10 stack in front of it, no matter if it's TanStack start or TanStack query forms table, whatsoever, is super popular, not only in the React universe. Right?
Alexander Lichter:
That's the best part. But especially since they start as a as a React based framework with the router and everything. I think that will, yeah, that will go, quite quite well. And I'm I would wish, like, Tanner and and the whole team all the best. I think it's a great thing that happens React Universe, where it was always very, let's say, Next.js Vercel Dominated, especially the last year's. So having a bit more choice and also push towards framework agnostic libraries is pretty nice.
Daniel Roe:
Yeah. And, obviously, a lot of the, the stuff that people are going to be, interested in in TanStack start will be usable in Vue as well because the query will definitely be usable. And it might be interesting to see a router also. There could be some kind of, Nuxt router integration.
Alexander Lichter:
So That would be neat. Interesting. And, yeah, all the things they do with, like, I don't know, hydrating specific parts of the application, like, or scroll restoration for different parts. It's like, it's all super interesting what's happening there. And, yeah, I hope that, we as the the Vue and Nuxt Ecosystem can also benefit from these, let's say, results out there.
Alexander Lichter:
That would be great.
Daniel Roe:
Yeah. It feels interesting. It feels fun.
Alexander Lichter:
One word there in terms of nitro is I really hope that prediction comes true, and this will be the year of nitro. I mean, we've seen it also in certain posts. For example, that we find a little post of a a year review where it was mentioned that, like, okay. Nitro is not only used in Nuxt, which is true by a while now, but now more and more people see that. So I think that's great.
Alexander Lichter:
Though on the other hand, I think it will only be really the year of Nitro if there are a few things done, let's say, differently, or the focus, has to shift a little bit in terms of, like, more content around what Nitro's capabilities are. People could say, this is like marketing, quote, unquote. I would just say extensive docs, examples, cookbooks, and so on. Just say like, hey. Okay.
Alexander Lichter:
You can do all these cool things on all different, like, deployment providers, no matter if it's on Cloudflare or if you use, I don't know, your VPS or Vercel, Netlify. You can do all these amazing things, and this is how the framework works here. The 100 most common cases, what you can do, and and have, like, people pick it up, not only being seen as someone that's like, okay, this is used to build your meta framework, but also this is used to replace current server frameworks to make things more modern. And I really hope this will come true, plus the whole part of well, I think that even applies for UnJS in general, that more people can devote time to, like, contribute to UnJS and to Nitro that can be issue triaging. As you just mentioned, documentation examples.
Alexander Lichter:
So I really hope that will happen as well because then I really think that the year of Nitro is is 2025 indeed.
Daniel Roe:
Yeah. So someone needs to create a course, don't they? Oh.
Alexander Lichter:
That's true. Some someone. That's right. That might be. When Nitro v3 is out, there could be someone on this podcast, maybe even who created a course.
Alexander Lichter:
But I heard there's also someone else who wanted to create a course moving from an framework that's, has reduced and reduced positivity over to Nuxt as well.
Daniel Roe:
Oh, I feel so guilty about that. So you're you're talking about React to Nuxt. Actually, I got an ask me anything question the other day about it. The problem, of course
Michael Thiessen:
Do you have a prediction for this?
Daniel Roe:
Do I have well, I predict it will be it will be launched this year because my plan is to, to build it, like, to record the sessions, the lectures, the lectures, videos, like, with with a initial cohort of people who are going to give feedback and direction. So basically say, hey. I need some React devs. I'm gonna teach you, like, next. So here we go.
Daniel Roe:
I'll give you a video every week, guaranteed. Tell me how it goes. What are the topics you want to know more about? And I'll keep them coming. That's that will be the sort of the promise.
Daniel Roe:
And so once that starts, I'm in. I've got no choice. I've got to got to build it. So the main thing for me is just to get the info infrastructure ready. So I've got, like, a payment, a way of being paid and hosting the videos, which I'll probably do on Cloudflare video.
Daniel Roe:
Yeah. And,
Alexander Lichter:
yeah, that kind of thing. Yeah. I think, Michael, you also can, sing a song about that. Like, building your own course platform to launch your courses.
Michael Thiessen:
Yeah.
Daniel Roe:
It's crazy, isn't it? Like, it's impossible to use an existing SaaS link. I'm I'm guessing there probably is some kind of SaaS that already exists, but it probably takes a percentage. Probably, I would guess.
Michael Thiessen:
Yeah. Yeah. There's some and then they you can use or whatever. Yeah.
Michael Thiessen:
They're probably
Daniel Roe:
not exactly what you want.
Michael Thiessen:
What you want.
Alexander Lichter:
Yeah. Exactly. That's the problem.
Daniel Roe:
Like, we're very demanding.
Michael Thiessen:
Add this little thing in there and you're like, what? You can't. So Yeah. But then you have to deal with all of, like, the, you know, the the authentication and payments and, you know, all that kind of stuff. And that's the part that takes the longest as as with, anything.
Michael Thiessen:
Because the actual, like, amount of features you need for a course platform is fairly minimal. It's just, like, all of the boilerplate is actually pretty substantial because you do need, like, authentication and, yeah, payments, and you have to manage, like, maybe you wanna have multiple courses in the future. And so you have to maybe use do that. And then are you gonna, you know, all this? Yeah.
Michael Thiessen:
There's lots of stuff.
Alexander Lichter:
There definitely is. There definitely is. I I think as a lovely reference, there is a lie like, not library. There's a repository out there. Even better, it's open source.
Alexander Lichter:
It's called course dash builder by badass courses. And whoever, know, like, probably all of the or, like, a lot of listeners out there know Matt Pocock, for example, who's also on the podcast already. His total, TypeScript course is also based on the course platform he's using on the repository. Sadly, of course, it's React based, but that's still, I think, kind of a good starting point to see what is all, like, possible, maybe needed, maybe nice extra, and and also worth taking a look at, especially because it's open source, all made by, I think, Joel and the team there to bring like an amazing course platform builder, so to say, to the people. So also worth looking in there.
Alexander Lichter:
It's also like they say it's a real time multiplayer CMS for building, deploying opinionated data structure of developer education products, which sounds quite substantial. So, yeah, worth worth checking out for sure. Maybe, the 2 of you, and maybe me at some point when Nitro v 3 is out around that could, learn a thing or 2 there for their their course platform. But, yeah, I'm I'm definitely looking forward to to all courses that will come up, actually. So, yeah.
Daniel Roe:
Especially yours. I'll be very interested, if you if you produce a a course on Nitro, that would be phenomenal.
Alexander Lichter:
I will most likely. You
Daniel Roe:
really should do.
Alexander Lichter:
Here here everybody But you already started.
Daniel Roe:
I Are you waiting for you waiting for the release?
Alexander Lichter:
I wait for the release. I also think I I put quite some substantial Nitro content out there already so people can get started with. And then it's like, of course, I started with, like, a roadmap to see what I wanna cover, how I wanna do this different angles.
Alexander Lichter:
And, there's also one other thing I'm waiting for, and this is a really spicy hot take. And I don't wanna offend anyone, but I really hope 2025 is the year that Vinxi won't be needed anymore.
Alexander Lichter:
So this, yeah, to, like, explain a little bit because for the Nuxt users, it's totally relevant, and I think for the analog users as well. But testing start and solid start users so their their frameworks are based on Nitro, but not directly. So there's a layer, let's say, an unopinionated, opinionated layer in between called Vinxi. And it's basically wrapping Nitro or, like, a lot of features of Nitro and providing certain functionalities. And in an, I think, ideal world, quote, unquote, this would not be needed at all.
Alexander Lichter:
So I really hope that there is a way to see that, also maintainer of Vinxi could maybe join forces, and we can all work on 1, like, say, one bigger project together, to make things work out and all like, fit all the community needs in.
Daniel Roe:
Yeah. I think the the question is, like, whenever you have a it feels weird to have a layer over a framework. So, like, as in I guess the equivalent would be someone building a layer on top of Vite in order for Nuxt to use Vite under the hood, that kind of thing. So you might want to do that if there's a deficiency in Vite or something that needs to be. But it feels like the better solution, as you say, is to contribute to the underlying, framework and add fix whatever the the thing is.
Alexander Lichter:
But that's, of course, the other thing.
Daniel Roe:
That's definitely a spicy take, though.
Alexander Lichter:
It it is a spicy take. So I even the author, I would say, kind of agreed. It didn't talk to Nikhil specifically about it, but there was a was a video that was posted and someone on Twitter asked, like, okay. Does it mean that Vinci will go away? And Nikhil was like, ideally, yes, but unlikely.
Alexander Lichter:
Something like that. I'll link the the tweet below so you can read it up.
Daniel Roe:
So he would like to see it go away?
Alexander Lichter:
I mean, in an ideal world, it seems like. If we can reach that or some kind of compromise there, it would be great.
Daniel Roe:
It's always hard though as an author of anything. Like, you don't want like, it's really hard to I don't know about you, but I I I often I do not want my stuff to to go away. Like, I I want my library to stay. And then then I find, oh, there's another library that does the same thing. Maybe it should go away.
Daniel Roe:
It's really hard. It's basically a like, it's ego death. It's sort of the it's like I have to sort of say no to, like, the thing that I might be known for this. I won't be known for this. It's gonna go away.
Daniel Roe:
Or you enter a sort of a time where you you sort of fight or or like contend, compete, for and, you know, and there can be lots of good reasons for that. If you think your thing is better, there's a better approach or whatever. But now it's really, really amazing humility when people are able to say, I'm going to not try and keep my thing. I'm gonna sort of collaborate instead. And so it's always impressive to see.
Daniel Roe:
I always admire people like that.
Michael Thiessen:
I'm impressed by, Eduardo as one person who I've seen do this a bunch of times where he's like, don't use pina if you if you don't have to. Like, try try to avoid using it. And it's like, how do you get to that point where you can actually like, oh, I created this thing that I spent so much time on, and I want people to, like, get value from it, but also, like, don't use it. It's like, I I appreciate that, though.
Alexander Lichter:
I think I I like to mention, some kind of tendency. It's called NIH. Maybe some people know about that. Short for not invented here. And it's beyond tech.
Alexander Lichter:
Just like, okay. People are less likely to, like, use some external that can be, like, things. Right? Like, products, research, whatsoever. So, like, okay.
Alexander Lichter:
If that wasn't made here in our company, we'd rather not use it and build our own, and that kind of leads to the standards xkcd, as we know. Like, oh, there are 20 competing standards. Let's just make our own now. We have 21. So in a way, it's really tricky because if there's nothing that exactly fits your use case, then it's very tempting to say, okay, just build this, especially it's modern.
Alexander Lichter:
It's also under, let's say, quote unquote, my slash our control. But ideally, then making the step either before or after saying, well, there's something that works equally well. We just need to add a feature or, like, use it differently and can do the same, if not even more, can always, like, be a good solution, at least to check out because I'm the big fan of having less code. And also that means less code to maintain as long as the other thing that's out there that I might co maintain then is is steering the right direction or a similar direction.
Michael Thiessen:
So I've got another hot take for you all. Actually, I'm borrowing this one from Justin Schrader. So I don't know.
Alexander Lichter:
From the future.
Michael Thiessen:
But from the future, we will talk to him in about a week. Yes.
Daniel Roe:
You will publish the episode in a week.
Michael Thiessen:
We are all
Alexander Lichter:
No. Michael has Michael has a little a crystal ball, seeing the future. I'm hoping.
Michael Thiessen:
I'm predicting a prediction that he will make. So this is getting meta now.
Alexander Lichter:
It's a 5 k move. Impressive. There we go.
Michael Thiessen:
And that now 2025 will be the year of meta frameworks in general where, we're seeing most of the innovation and excitement and movement happening at the meta framework level, not at the framework as much. So Justin pointed this out in the future. And I think I think there's probably a lot of truth to that. I think it's, you know, Vue has had not as much in terms of, like, releases and things like that, and and that's great because it's we want it to be stable and all this. Nuxt is also stable, so maybe that's not, like, the point that I'm actually trying to make.
Michael Thiessen:
But, like, there's not as many features that Vue needs to have to be added, and it's harder to do stuff there. And we are seeing the movement and extra stuff happening is all of the things around what Vue is doing, like with Nuxt.
Daniel Roe:
Well, I mean, I like I like this prediction. It's it sounds good prediction for Nuxt. I mean, it I think, basically, there are natural limits to what a framework can do because a framework is confined, in many cases, to runtime code. So it's therefore it's not totally the case. Vue, you know, has compiler, like a compiler plug ins.
Daniel Roe:
It can do a lot of optimization at that level, and it's tried to reach out a little bit. So with supporting props being defined in TypeScript. I'm doing some cogent based on then it has to resolve the TypeScript config of the project it's running in. Like, it becomes very complicated very, very quickly. But there's a lot of stuff it can't touch.
Daniel Roe:
It can't touch the rendering of HTML. It can't touch the, like, the optimization of the project as a whole. And that's why, you know, VitePress even existed, VitePress, I guess, originally existed, as just, like, an opportunity to really implement some of the the things that you needed because Vue wasn't enough. You couldn't just do it with Vue. So I I I totally see that.
Daniel Roe:
I mean, this isn't so much a thing in the future. It's a thing, and it's been true for the last couple of years, that with the proliferation of deployment services and ways of hosting apps, serverlessly and on the edge, you know, it's not a new thing, but that it demands frameworks. I think there's been this pull. It's something that is very much shaped what we've been doing at Nuxt. But I think also, it's no surprise that it's gonna pull everything else out.
Daniel Roe:
The React is the one outlier. Right?
Alexander Lichter:
With, like, the server components built in the the framework and not the meta framework. Yeah. Exactly.
Daniel Roe:
And so that's the React, it feels like, has aspirations to be, Like, if you have a sliding scale from something that's purely runtime code, just a library for rendering some stuff in the browser, and something that is, you know, much more of a framework, React really wants to be on the framework end of that, I feel like. I mean, someone even said, oh, meta frameworks are called that because React wants to be a framework. I'm not sure if that's true. I think meta frameworks, it's not just that they have one framework, like Vue or React in them. They also have other frameworks, like Vite or Webpack or PostCSS.
Daniel Roe:
Like, there's lots of other things involved, other libraries. And so but I don't know. React definitely has a claim to be sort of on the framework end of things.
Alexander Lichter:
Now, at least.
Daniel Roe:
Yeah. Now. And maybe that's a good thing, but maybe it's also a bad thing. I I mean, it was one of the people can struggle to hack. How do you implement React?
Daniel Roe:
How do you build on top of it? You know, what does your server component look like?
Alexander Lichter:
Yeah. It's I think also it was very controversial for people just said, I want my single page application and, like, also that not being that easy with, well, React meta frame I want specific React meta framework, so Next.js, at the time before. And, also, like, people just straight up suggesting, hey. You should use React with a framework, which means people, like, are more pushed into, hey.
Alexander Lichter:
You should use SSR and not sure you can still build your plain SPA. You better use the meta framework though to make things easier and still keep a, let's say, a foot in the door if you need some SSR capabilities instead of, like, rewriting your stuff done. So in a way, I think what the, let's say, quote, unquote, rest of the ecosystem does with, like, yeah, you have meta frameworks. You can still build SPAs in it. I think that makes a lot of sense.
Alexander Lichter:
And still, we don't see that big of a adoption, in other ecosystems using meta frameworks compared to React. Maybe because of that push, but also even before that, we've seen a lot of people just straightaway jumped into Next and not plain React first. So I I also think that will not change for React, but maybe that will change in terms of the year of meta frameworks to to join that prediction. Hopefully, it will change in 2025 a bit. The meta frameworks becoming more of a thing that will be adopted.
Daniel Roe:
Rock on the year of meta frameworks.
Alexander Lichter:
Anyways, Daniel, you have another prediction?
Daniel Roe:
I think, we're gonna see a rethinking of how open source works from a maintainability and sustainability point of view. It's something that we've seen some very interesting moves on this last year. So GitHub is really consistently trying to invest in open source. They've obviously GitHub Sponsors has meant a lot, has done a lot over the last few years. There's also, stuff like Sentries, Pledge, and others, Stackblitz, and so on.
Daniel Roe:
Platforms like thanks.dev. Other like, obviously Open Collective and Patreon, all these things exist as well, but people looking to try and automate things or create funds for open source. I mean, I don't know. This is the kind of thing that you predict a lot a lot of people have had hopes and expectations for. It feels like there's enough rethinking or asking the question of it that it might be.
Daniel Roe:
And I think one of the things that will drive that is actually WordPress. So the very, very, very bad and very public implosion of Matt Mullenweg, who basically went rogue and and started throwing around accusations that that seemed to me to be without base and acting in an incredibly damaging way to the contributors to WordPress and to, to other members of the WordPress community, all over the question of whether people are allowed to make a profit on top of the work that he and others have put in to an open source project. And so that question I mean, I think it that what he did was a sort of existential threat to open source. But I think one of the things that it might do is it might end up making us ask the question about how we have a healthier model of open source and how we have a healthier model of open source sustainability. So I this is an edge prediction.
Daniel Roe:
I I don't feel this is a sort of 100% kind of thing at all. I think people might be asking the question. They probably should be asking the question. There are enough big organizations already that are voting with with money, like, you know, Sentry and StackBlitz and and VoidZero and and others. It feels like there's enough there that something could happen.
Michael Thiessen:
I'd agree. I think, like, I really, really hope that something does change because it does feel broken. And I don't think I know enough or I'm in embedded enough to see, like, what possibilities there are necessarily. But just like I've talked with, Ryke from Directus, and he's very passionate about the whole open source and, like, funding models and things like that. And Justin Schrader, we talked to in the future with Formkit, and they're trying to figure out how to make this work.
Michael Thiessen:
And, like, we see people like you, Daniel, like, all these open source maintainers who are out there doing, like, really good work and yet, like, trying to find ways to actually fund what they're doing. And if you were doing that same work at a Google or a Facebook or any other company, you would be, like, very well compensated. And yet here you are, like, taking this massive sacrifice financially to do this other work. And it seems like this weird imbalance. And I do wish that there's some way to to address that, to, like, close that gap of of allowing people to actually work on open source, but actually, like, sustain themselves and and actually make money and not just have to beg online practically to to get people to to donate in, like, big corporations that are using and profiting off of their their work.
Michael Thiessen:
And somehow we've we've limped along in this current state. So hopefully we can, yeah, make that better.
Daniel Roe:
I mean, I wanna be very clear. I'm very fortunate. So I'm I am sustained. I'm not gonna hand out my cap asking for people to contribute. But that is by no means true across the ecosystem.
Daniel Roe:
I have a lot of opinions about this. I have a lot of opinions also about what we get wrong about it from the point of view. Even when we're trying to talk about fixing it, what are we trying to fix? And why isn't stuff working? So is the idea that people should be able to earn a salary by contributing to open source?
Daniel Roe:
How does that work? Or is the idea that people should just be able to to sort of maybe not earn a salary, maybe earn a stipend or something that basically enables them to work on open source, but that it's not directly tied to work? Stuff like maintenance contracts is typically I think it's it's very, potentially, very unpleasant place to go. Anyway, they would I think it would take over the whole episode. I think there's there's a lot of
Michael Thiessen:
Mhm
Daniel Roe:
lot of, A lot yeah. It it would be interesting to talk about because I think there's a lot of, you know, people don't give money to open source based on the value that open source brings. It's just not a thing. People are there are projects that are incredibly valuable and nobody supports them.
Daniel Roe:
And there are projects that receive disproportionate support. Well, I mean, disproportionate to the value they bring. They bring value, but if they get more support than than you would expect, you know, there's there's reasons for that. And there's not bad. There's there's reasons.
Daniel Roe:
So it's there's lots lots to talk about.
Alexander Lichter:
Absolutely. When you started with WordPress though, I was wondering, and I think a lot of people are wondering, why hasn't there been a fork of WordPress yet? Like, okay, we have all the drama and some kind of existential threat to open source, but there hasn't been a WordPress fork. So that's also quite interesting because usually, if people start, hey. I have a different opinion than you and people are like, okay.
Alexander Lichter:
If we can't find a way, it's open source. Do whatever you want. You can fork it. But even if a project that's that big or maybe because it's that big in terms of WordPress, that didn't happen.
Daniel Roe:
Maybe. It is a very interesting question. I mean, it's also not done. So although, I mean, I've got to say that no matter what the courts say, it's difficult to find a to see a future that doesn't involve Matt needing to step down or step out of leadership. I mean, this I wouldn't be happy being involved in the project if he remains because he has done such awful and he's the the idea that that he's not stepped down, expressed remorse, made any apologies, He's doubled down every single time.
Daniel Roe:
So it feels like that is is, you know, that that could change. But right now, you know, I think they need a a different they need some different leadership. And it will be interesting. Yeah. Maybe, you know, because the WordPress is it's got community word camps, and it's got, so it's got support, and it's got, obviously, infrastructure of the plug in system, and there's the agencies who have clients on WordPress and and all of that.
Daniel Roe:
It's not a simple thing to do, but Absolutely. I I think there will surely there will be something.
Alexander Lichter:
Oh, that's that's interesting take. But do you think there will be a fork?
Daniel Roe:
I think probably people are devoting their energy right now to trying to get a change. I would guess. I don't know anything about this, but I would guess that behind the scenes, people are trying to rework the leadership structure and the governance structure of WordPress. And probably the reason we're not seeing forks from bigger community members and more influential people in the WordPress community is because that's where their energy is going. I suspect if they don't manage to make that work, then we all see something.
Daniel Roe:
So the fact that we haven't seen something yet isn't a sort of weird people don't care, but probably it's evidence that people are doing something else instead behind the scenes.
Alexander Lichter:
That makes sense to me. Yeah. That's I think it's a good take. I like that, and I'm curious what will come out of that for sure.
Alexander Lichter:
Michael, you have another one for us and our prediction for 2025.
Michael Thiessen:
Sure. I'll go with not very controversial one because it's already starting to come true. So a few weeks ago, I I sent out this this tweet where I had some predictions, and they were mostly just, like, off the cuff. I didn't really give it a lot of thought. One of them that I put in there was more amazing conferences worldwide.
Michael Thiessen:
And just a few weeks ago, the, Mad View Conference in Madrid, popped up. So there we go.
Daniel Roe:
Yeah. You made that happen clearly.
Michael Thiessen:
I Obviously. Manifested it. I willed it into existence. So say what you will.
Alexander Lichter:
Yeah. I think that's a, that's a very interesting one. I think I read the tweet and, yeah, that MadVue is coming up. I'm really like to see that. Plus also a possible 2025 edition of a PragVue, who maybe a few people, listeners have heard about before.
Alexander Lichter:
Let's see if, and how these conferences will go. I think it's pretty interesting because from what I have experienced as someone who went to a lot of conferences, I mean, Daniel, you too. And Michael, you've also been to quite some. But, like, seeing most of the Vue conferences in in 2024, I felt that here and there, especially with, like, budget cuts, and and so on and educate not only education, but also, like, level of our budget cuts, There were less and less people at a lot of conference, not all, of course. There were still, like, full rooms, and quite some.
Alexander Lichter:
But I felt a few conference also struggled with getting the right amount of attendees, especially in the the Vue ecosystem somehow. For example, like Daniel, I both went to Vue.js DE, which was 2 days this year and not just 1. And the Angular conference, which was right after that, was way bigger in terms of numbers. Might be because, well, Angular is more, let's say, prominent in the German speaking area, or maybe because it has different standing, different community ties, whatnot. But I also think that in a way, there might be other like, there would be a change in terms of how conferences market themselves or, like, more local things, maybe more smaller things.
Alexander Lichter:
I would love to see more, also different approaches. I would love to see it coming true. I'm not sure it will besides of MadVue, but I would love to.
Michael Thiessen:
I'm trying to get to more conferences as well. For a long time, I've just been to the one in Toronto because it's in my backyard practically. And most of the other conferences are over in Europe, which is a little bit more difficult for me to get to. But now I have a baby girl, so it's hard to, get away for, like, almost a week at a time. So we'll see where where things happen, but I'm all for it.
Michael Thiessen:
I have to be selective with the European ones. I'm hoping that an Australian one comes up because I have
Alexander Lichter:
That'd be lovely
Michael Thiessen:
to see people out there, and I'd like to have a good excuse to go out there and meet some of the Australian folks. So
Alexander Lichter:
Harlan, do you hear us?
Michael Thiessen:
Yeah.
Alexander Lichter:
Yeah. I I would love to see that too. Like, like Laravel Australia is a thing, for example. So, we're not having a views Australia or so? I don't know.
Alexander Lichter:
Yeah. Take the name, please. Take it and make it happen. I would love to.
Michael Thiessen:
The one in Japan, I haven't I haven't been.
Alexander Lichter:
So that was lovely.
Daniel Roe:
Definitely. I mean, I I have questions about how much will be specific to Vue in terms of conferences. How much might be more generic? Because one, one thing of course that attracts conferences. So I think conferences are relevant for new developers who are exploring something for the first time.
Daniel Roe:
Also when thing people are doing cool stuff and pushing at the edges. So, like, you have a new library, Trace or Penia or sub view 3. Let's talk about what composition API is and, you know, and and sort of exploring something that's new. The more something becomes established and there is just it's there, you know, how many years will we have talks of here's what the composition API is? And so, like, as something becomes mature in there do you know what I mean?
Daniel Roe:
Basically, we have to have the, and here's a prediction as well.
Daniel Roe:
2025 is the year of vapor, Vue Vapor. Then we'll see conferences talking about Vue Vapor. And that that's going to be an interesting thing. Right?
Daniel Roe:
As you sort of have people talk and get like, what is it? What are the patterns for using it? When do you want to use it? When do you want to use Svelte and Vue in the same app using Vue Vapor?
Alexander Lichter:
Wait. What?
Daniel Roe:
And I'm sure you know about this. So basically, so transitioning smoothly into my prediction of Vue Vapor, there's been a lot of work on it. It's happening behind the scenes in a number of different places.
Daniel Roe:
I think things are gonna move forward. And I think probably even quite quickly, we're gonna start to see stuff working or accessible. And it's gonna start out as probably not gonna work in a bigger project. It's gonna need some compilation. It's going to work probably in VitePress, out of the box to start with.
Daniel Roe:
People are gonna play around with it. But for example, kazupon, who, is the driving force behind Vue, and Nuxt i18n, originally, has been creating, like, a Svelte Vapor that basically how uses the same approach. So you can because so the concept, of course, of vapor is can we compile away the Vue runtime or have a very minimal runtime so that basically we don't need a virtual DOM for these components. We can directly manipulate it. Well, that's very similar to Svelte.
Daniel Roe:
And actually, there's nothing Vue specific about it at that point to the compile output. So normally, you wouldn't want to have more than 1 UI library in your project because it's extra code. But if you're compiling down to effectively imperative functions that just directly manipulate the DOM, it doesn't matter. You're gonna have the same behave. If I have 2 components, it it's going to be the size of those 2 components.
Daniel Roe:
It doesn't matter what frameworks they're written in. And so kazupon is exploring, like, well, Svelte. Can we write in Svelte and compile it to Vue Vapor? So why not? So I think Vuevapor is going to potentially be if if people build on top of it, in the same kind of way that we're starting to see, happening behind the scenes.
Daniel Roe:
Vue Vapor might end up being, you know, actually, well, better than Astro, you know, better than what Astro promised. Actually, something that allows Interop at a very granular level in existing frameworks. So that could be pretty cool.
Michael Thiessen:
Sort of like using Vapor as like a machine code for the DOM in a way where it's like everything kind of compiles down to the same assembly code or byte code or whatever. You know? Yeah.
Alexander Lichter:
That's super interesting. Yeah. Like, you know what I mean? Before I'm really happy that I didn't say anything because before you brought it up and I I was aware of the experience from kazupon I was I was even saying, I mean, vapor mode will surely come out this year. I think we're all clear on that.
Alexander Lichter:
Like and of course, not a 100% guarantee, but same guarantees as Nuxt 4, I suppose. But I would have even said a vapor mode will, like, not be significant for most of the people. That might be a hot take just because, well, if you're not doing, like, large data tables or some kind of data visualization with a lot of SVGs and changes and charts that move super fast or, like, gaming or whatnot in the browser with a lot of elements that move super fast. So if you build your typical software as a service application or, like, your typical, I know, admin dashboard and whatnot, it won't really matter how fast the, let's say, UI renders because it's quite fast already. And instead, it matters, of course, how optimized your, I don't know, API calls are, how well caching is, and so on.
Alexander Lichter:
So I think a big, like, a big group of people that are just building stuff, and I don't wanna say boring stuff, but, like, more, let's say, typical things won't see crazy benefits out of that for their stats.
Daniel Roe:
I agree. I agree.
Daniel Roe:
But I think what what is going to be interesting is going to be, Justin's point about meta frameworks. So can we automatically optimize things for users? Now the Vapor is one compile target.
Daniel Roe:
The Vue virtual DOM is another compile target. Can we have others? So can we have stuff like that is purely static? Can we have things that are do you know, like, I I feel like server components is effectively becoming a compile target in Nuxt? Other kinds of compile targets that we can optimize for, where we basically say, this should be this kind of thing, because that's gonna be most performant for you, maybe.
Alexander Lichter:
That's interesting take. Yeah. I like both seeing that. I hope there will be more of that in 2025 and more exploration than that. Nice.
Alexander Lichter:
Anyone else having another hot take for us, another prediction? Something not so spicy, maybe. Anything like, hey. This this is what should happen in 2025. Might like might likely will.
Daniel Roe:
Okay. I'm gonna give you an AI hot take. Oh. Oh, prediction. I think it's gonna be of the AI agent.
Daniel Roe:
So, there's there's a lot of stuff that's l been LLM powered this year, and there are a lot of explorations in into agents and agent frameworks. So, of course, the idea of an agent is not just, LLM or AI help that is directed by you, but it's it's able to do things more on its own. So maybe you give it some powers, some abilities. It can buy some stuff. It can research some stuff.
Daniel Roe:
It can contact like it. I don't know. It can can do some things, and then basically let it loose. Imagine, you know, a virtual assistant that does stuff, just does stuff. That I think is going to be a common project people build.
Daniel Roe:
People are going to be building a lot of these. We're gonna see them built in Nuxt and in other projects too. And I think it we're gonna see some frameworks for them. And I think it's also gonna start affecting us as developers. We're maybe going to I'm not talking cursor.
Daniel Roe:
You know? You're still directing that. I'm talking maybe stuff that is more, you know, sua sponte. Be interesting to see.
Michael Thiessen:
I've seen recently the people at Langchain released this email agent. And with that, this idea of I think they call it ambient AI or ambient agents where the idea is it's like if you had an assistant looking at your email, you don't have to reach out to them and say, what emails did I get today? Or you're not the one initiating. It's just like they're triaging your email. Maybe it replies and says, oh, Michael will get back to you within a business day on this issue and maybe some, like, automatically puts into spam or maybe there's really important ones and it, like, escalates that back to you.
Michael Thiessen:
And it sort of, like, just sits there and kind of just, like, takes action as it needs to and and is proactive versus, like, yeah, you directing it like we're used to with with cursor. And I think for like, I can see this happening maybe with, like, sitting on different repos. Like, we have code review AI and and things like that, but I think it might get more sophisticated.
Alexander Lichter:
Really interesting. Yeah. I think, like, AI will definitely grow more and more, though I'm also wondering when the tipping point will be. People say, yeah, probably not in in a couple of years. People say probably this year already because we can't just, like, feed more data into the models, like, make them bigger and see better results.
Alexander Lichter:
So from that level on, I also wonder when is the peak in terms of, like, AI models. Not even talking about agents. They were just saying, yeah, can we do, like, AGI very soon, or is it more like we need other methods to to build these models? Of course, being no ML engineer or AI engineer or whatnot by far, I would just curious
Michael Thiessen:
There's been a lot of, a lot of developments. I think, yeah, the model size has sort of topped out, but what they're doing, they're doing a bunch of interesting things. So the one is the whole reasoning angle with, like, o one and o three coming at some point. And there's other models like Google has Gemini, has, like, a reasoning angle, and then there's deep seek. There's, like, a whole bunch of these companies are now doing that where it, like, takes an existing model, but lets it basically think more until it's and people are speculating on what happens.
Michael Thiessen:
Maybe it's like, I don't know, some sort of, like, search process. It's, like, running itself multiple times or doing something under the hood where it maybe it'll take a few minutes instead of 10 seconds to answer the question, and that can make a big difference. And they're also taking these big models to actually generate new training data to basically make the smaller models more potent instead of just, like, taking, you know, data from the Internet. They're specifically crafting, like, very targeted data, and so that's another interesting thing. There's always, like, new developments people think, oh, that's it's topped out, but then there's some new breakthrough that happens.
Michael Thiessen:
And, like, I think it was meta a month ago, came out with this idea of, so right now, LLMs sort of think in tokens that are approximately, like, one token per word. But their whole idea was what if we had these tokens as, like, a whole sentence or a whole idea, and they found that they got, like, huge performance gains if they switched to this method. It's a huge space, and it's impossible to keep up with everything. And I'd, like, try to sort of pay attention to what's going on. But, like, every week, there's, like, so much stuff.
Alexander Lichter:
Oh, yeah.
Michael Thiessen:
But I think there's a lot of room left, so I don't think that it's topped out yet. It'll top out anytime soon. That's my AI prediction for 2025 is that it's Interesting. It's not gonna top out.
Alexander Lichter:
Alright. We are slowly but surely wrapping up. I think I will bring in 1 or 2 more that are rather easy, not too spicy. Maybe one that's really spicy. Let's see.
Alexander Lichter:
But I think an easy one, and we talked about it in the the recap of 2024, alien signals. And I think alien signals will be adopted by other frameworks because everybody is on signals, and they wanna good get good performance, at least adopting some parts of it. And now Johnson is once again on a different approach, yes, right now, when we're speaking, of once again writing things differently about, like, I think, memory leaking versus memory safe signal systems, something like that. I will link the the posts, from Johnson on that. So but I think, really, like, Eigenstates was adopted in different languages already.
Alexander Lichter:
So we've seen things in Dart and Lua even, and it might also be adopted by other JavaScript frameworks. I'd love to see that. Everybody nodding,
Michael Thiessen:
The only thing I could think of for some reason while you're talking about that was that Aileen signals would be a great DJ name.
Alexander Lichter:
Yeah. That's true.
Michael Thiessen:
If anyone's out there.
Alexander Lichter:
So if Johnson will ever start a DJ career, we would have it. That'll be fun, though. And okay.
Alexander Lichter:
And and another one, I think that Vue will not get a major this year. It's not a crazy surprise take.
Alexander Lichter:
What do you think about it? Major for Vue this year, possibly?
Daniel Roe:
Do I think there's a major thing this year?
Alexander Lichter:
Major version.
Daniel Roe:
I do not think there is.
Daniel Roe:
It's an interesting question. I mean, I think people are starting to talk about it. I think I wouldn't rule it out for 2026, and I wouldn't rule out work on it, and I wouldn't rule out even an RC. So Same. I mean, nobody is is talking about it right now, though.
Daniel Roe:
And in the sense that there is no there's no feature or breaking change in Vue that anyone is talking about being a reason to do it.
Alexander Lichter:
I mean, there's there's some ideas floating around that also, like, even Evan told us last year in the podcast, but it's the thing, like, there's some ideas, some things try trying things out, seeing what's possible, but yeah. And maybe these are even things you can bring in as experimental, first of all.
Daniel Roe:
Yeah. Yeah. Which is nice pattern, isn't it?
Alexander Lichter:
Oh, yeah.
Daniel Roe:
Bring something in experimentally.
Daniel Roe:
Try it out.
Alexander Lichter:
I heard that's, in progressive, one would even say.
Daniel Roe:
Yeah.
Alexander Lichter:
I got one more, which is I cannot find a way to frame it very spicy, but, actually, it's not very spicy, I think. Mhmm. We will not see a UniFile JavaScript tool chain this year. Very important. But I think we will see more stepping stones towards that, which means Vite with Rolldown included will be shipped this year.
Alexander Lichter:
And we already have a beta, so it's I think it's not not too too questionable.
Alexander Lichter:
But, yeah, that's definitely something. Well, maybe we even see more rustification around that space. But Vite with Rolldown, I think, is the the minimal level. I said, like, that will that will happen.
Daniel Roe:
Very cool idea. I I don't know what's gonna happen with with, with Vite and the whole void 0. I I'm sure yeah. You've obviously talked to Evan about it. It'd be really interesting to see.
Alexander Lichter:
Oh, it will be. It will be. Any other predictions or hot takes or opinions?
Daniel Roe:
I think, I mean, on the hiring side of things. And I think, basically, these things are cyclical, and they go up and down. And I would expect to see increasing demand for developers, but I think a lot of it is going to be in fields like AI and LLM. And I would think that's probably quite a useful thing to develop skills using building with LLMs. So even if you're building even if your experience is full stack or even if it's front end, understanding a little bit about how these things work under the hood.
Daniel Roe:
It's not just AI. Like, it's not just magic. What's going on? What's happening? I'm not talking about becoming, like, a machine learning engineer, but just what's happening and how to use it.
Daniel Roe:
I think that is probably something that's going to become a huge hugely useful. I don't know to what extent AI is gonna be about, like, crypto. And at some point, just a a lot of startups are gonna disappear. I'm I'm sure that will happen. Yeah.
Daniel Roe:
Not not in the same way that it happened with crypto, but, I do think there's gonna be a a correction. But, again, that kind of thing doesn't necessarily affect individual developers in the sense that you find a job somewhere else. The company disappears, though. And so it it'll be interesting. Yeah.
Daniel Roe:
It'll be interesting. But I I I expect to see sort of healthier markets for for hiring. What do you think?
Michael Thiessen:
I hope so. I hope I also hope that for Vue and Nuxt, there's a growing market because I often see people say that they love to use Vue and Nuxt in their side projects and hobbies and and stuff like that, but they learn React because they wanna get a job. And that's where the the market's at. And I'm not in the market, so I don't know what it's actually like. And so, that's all that I have to go on.
Michael Thiessen:
So I, yeah, I hope that it becomes more the case that people can get employed using Nuxt and Vue. And, I think that's that's going to happen as, I mean, if 2025 is the year of enterprise Nuxt, then, I think that'll go a long way to to helping with that.
Alexander Lichter:
I think both points are quite interesting. I think with AI and let's say productivity improvement in a way, we will maybe see less hiring in that space or like at least hiring because, like I said, I don't know. Before we needed 10 developers, and now we need, I don't know, 5 of AI, just like just imaginary numbers, but, like, less developers plus AI help to build the same thing productivity wise. Not talking about people code reviewing and so on, but, like, pure building. So we might see less people being hired in general because the productivity goes up.
Alexander Lichter:
On the other hand, we could also say, like, okay, then people just build more and wanna, like, have just faster, like, better better velocity. So So that could go either way. But I I I really hope for for everyone out there, and I talk with a lot of people from communities saying, yeah, in jobs, job market, especially through all the global situation with the wars we've seen with, like, at least also for for Europe, big energy crisis and everything. It's not easy, and I hope it will get better for sure. As Daniel, as you said, it's like an an up and down.
Alexander Lichter:
And we've been for a long time in the market, it was just, like, for developers, it was so nice. Like, you could you have a lot of projects you could join as a freelancer. Like, if you're looking for a new job, you have no issues finding one. Even, like, pick your your tech stack whatsoever. I probably won't get to that easily, but at least moving back towards, like, let's say, more contracts, more more job opportunities.
Alexander Lichter:
And when we take a look, and maybe that's a nice circle back to the state of JS we talked about earlier, Vue is this, like, framework that's used second most in words. It was React, then Vue, then Angular, then Svelte, and then I think LATE and so on. I I can't fully remember, but it's easy to check. So I really hope this will go a long way. Plus, with all the positivity and retention we've seen in the view results of the state of JS survey, which I've also talked about a little bit.
Alexander Lichter:
We also talked about it in in the episode before. I have a dedicated video on that where I go through that a little bit. I I hope we'll keep this positivity and retention, because if if you showed one thing in 2024 and stability, and I hope it will continue. So fingers crossed for that.
Alexander Lichter:
Alright. And I think that's a that's a good point to wrap up. We're, have, well, we have presented lots of, opinions, predictions, and and hot takes. Now we want to know from you out there, what are your predictions for web development for for Nuxt and Vue? We haven't even talked once about web components.
Alexander Lichter:
Maybe that's a prediction by itself, which we won't go into that. But, yeah, let us know what what you think about, well, our takes and predictions, but also your own. Maybe there's something we we didn't cover. Definitely let us know. Daniel, thank you so much for joining once again on the lovely day to day podcast.
Alexander Lichter:
Always a pleasure having you here. And, if if people don't follow you yet, where can they do so?
Daniel Roe:
Well, you can find me on Blue Sky or on my website, roe.dev.
Alexander Lichter:
I'm there.
Daniel Roe:
And I'm probably everywhere else as well, GitHub and email and Discord. But, it's been a real pleasure. Thank you for having me.
Alexander Lichter:
Of course. And oh, that's that's a good point actually. Prediction wise, I really hope Blue Sky will explode a bit more. So that would be, maybe might be great if more more people could join, over there so I could slowly use less and less and less x. So people, if if you don't have a Blue Sky account yet, that's your chance.
Alexander Lichter:
Follow all of us there. We're all there. Links, in the phone and in the description for everyone. And, yeah, that was our 2025 episode. Of course, we will do we will do a review of our predictions at the end of the year and see if we were fully wrong, fully right, a full recap.
Alexander Lichter:
So stay tuned for that. But until then, of course, there will be new episodes, there will be new streams from Daniel, new courses and blog posts from from Michael. I also bring out some new videos.
Alexander Lichter:
If you haven't already check out the latest DejaVue episode, if this isn't the latest, because afterwards, we'll talk with lovely Justin Schroeder about Formkit. Also, some hot takes maybe here and there - AI is also topic, open source in general, and and much, much more.
Alexander Lichter:
And check out the older episodes. And, if if that's the last one, then, hopefully, yeah, see you next week, and, let us know all your thoughts, in the comments on social media. Until then, see you soon, everybody.
