Algora | Algora TV

Open source interviews #33 - Daniel Roe, maintainer of Nuxt

Algora

Watching now

00:00 Hey everyone, welcome to the Open Source Foundry Podcast.

00:03 Joining me today is Daniel, Core Maintainer of Nuxt.

00:06 Daniel, thank you so much for joining.

00:08 It's a pleasure to have you here.

00:09 And I would love to start with some information about your personal background

00:13 all the way to your involvement with Nuxt.

00:17 It is a real pleasure to be here.

00:21 Personal background.

00:22 I've got a lot of different things in my background, to be honest.

00:27 So let's see. I'm American. Originally, I grew up just outside of Washington, DC,

00:34 before I came to the UK to go to university where I studied law and loved that. So I was really,

00:41 really into law, particularly, I think I liked arguing, so

00:45 debating and that kind of thing. And so that suited me very well. Then I worked for a church

00:53 for a year, and then I went back to university to study theology for a few years. And then I

00:59 led a church for five years up in the northeast of England before I started a creative agency

01:06 with my wife and father

01:10 focusing on clear written communication.

01:13 We then added more and more technical stuff

01:16 to what we did.

01:18 And I then pivoted into software as a service

01:21 and got more and more involved

01:24 in contributing back to Nuxt, which was the framework we used at the time.

01:29 And then when I ended up shutting down the SaaS startup,

01:34 I was asked to focus full-time on maintaining Nuxt.

01:39 And I've done that ever since.

01:42 So it's been a bit of a weird journey, but I certainly loved every step of the way.

01:49 Wow. Okay. Yeah, that's an incredible background, actually, and very unique.

01:55 What is it that drew you to these career chains or a few of them?

02:00 I mean, you said you studied law, right?

02:01 So what made you decide to sort of like leave that field?

02:07 Well, in none of those cases was it a decision to leave

02:11 so much as it was something that I wanted to do.

02:16 So for example, law, I know I would have loved it.

02:21 I loved university and I loved what I was doing

02:25 in terms of moot court

02:27 and things like that and practice.

02:30 But that particular decision

02:34 was more about me thinking

02:35 that I was being more motivated,

02:40 I think, by what other people

02:42 would think of me

02:42 than by what I would actually enjoy most and be best at doing. So in terms of leading a church, church leader versus being a lawyer,

02:55 I felt, well, actually, I wonder if I'm making that decision just based on what I think I'd be

03:02 better at doing and doing well, or whether actually I was

03:06 letting other people's views of me influence that. So it was when I came to that conclusion that I

03:12 then went to study theology and think a little bit different about my life.

03:19 So that was, I think, a really healthy thing. Later on, when i pivoted again from church work to starting the creative

03:28 agency the main motivation was to help my parents come to the uk so as americans they couldn't just

03:34 move here um so we decided to start a company and uh build a creative agency that would then

03:43 give them the right to come to the UK. We could work together on it

03:46 and hopefully build something that would last

03:48 and be valuable.

03:50 And it was indeed.

03:51 It's still running.

03:52 The original team have still kept going,

03:55 which is a delight to see.

03:57 But again, that was the reason.

04:00 Every time you change,

04:01 like something big like that,

04:10 I don't know if you've experienced something similar, but definitely for me, there's a sense of mourning.

04:12 So you mourn what you had before.

04:16 You also, there's also, you know, what do people think of you?

04:24 So, you know, are you a failure in terms of a failure at the thing you were doing previously now you've switched to

04:25 something else did you fail was that like how much of that do you take on personally so i i think

04:31 there are a lot of things to um it's not just as difficult like it's it's more difficult than just

04:39 pivoting it's not just oh i used to do this and now i do this and there's some new challenges it's also what do i think of myself how do i feel other people think about me how do i come to terms

04:50 with that um do i feel like part of my life up to now has been a waste of time do i feel like

04:57 did you know there's lots and lots of questions along those lines, which are tricky.

05:06 Wow.

05:09 I feel like we have an hour-long podcast just discussing these things. Five minutes in and we're talking about deep, deep questions of identity.

05:14 That was full-sign.

05:15 Oh, my God.

05:17 And as you said, I guess it relates to how much your identity has been ingrained

05:22 with the things you do.

05:24 And that is that

05:26 is crazy I hope we can dig in just a little deeper I mean is there any maybe

05:31 just taking a step back making it a little more light-hearted are there

05:36 skills that you built along the way while in law while doing theology that

05:43 have been you know transferred to how you manage

05:46 the community today within open source or lessons that you carried over maybe so i mean absolutely

05:52 i do think that there are there are skills and things that are relevant sort of across

05:57 um those different career paths so uh i mean so yes that's one of the things that you spotted there, that in church life, a church isn't an institution. A church is a group of people, I believe.

06:27 a church does. It's not that someone says, this is what we do. And, you know, it's not like that.

06:37 It's a body. It's an organic thing, a group of people who work together. And I think there's a similarity to open source work in that open source work doesn't work on a company basis. It's not that there's someone who is in charge and just says,

06:47 this is what we do.

06:49 Leadership is important in both of them.

06:51 So it's,

06:52 I'm not saying that everything just sort of happens without any,

06:55 but it's not as,

06:57 it's not about just telling people what to do.

07:01 Like there is,

07:02 it's a much more complex,

07:04 more organic thing where things like

07:07 trust and um and value and um appreciation and like there are a lot of things involved

07:16 and that are that are i think and so yes there are absolutely parallels and i think particularly

07:21 organic nature um i also personally feel that,

07:26 so my view of open source is very much that it is about mutual gift.

07:31 So I mean,

07:32 I give and I also receive other gifts from people.

07:37 So those gifts look like lots of different things.

07:41 So other people give to the, I mean,

07:44 I felt this from the day I was contributing

07:46 to Nuxt, you know, building my SaaS startup, you know, I'm giving back bug fixes and features and

07:54 other people are also giving in bug fixes and features and we all benefit. You know, there's no,

07:59 it's not a contractual relationship. It's not a, we're not sort of, well, I'll contribute if

08:05 you contribute, you know, it's, it's just gift. And I certainly feel that as a maintainer

08:11 as well, you know, I give time, I help people. There's no condition on that. I don't do that

08:17 because people sponsor me or because I do that because I want to help people. And at

08:23 the same time, people sponsor me and they don't sponsor me because I am,

08:30 it was an arrangement, you know,

08:32 I'll sponsor you if you do this.

08:34 I don't have any sponsorships that are conditional.

08:38 And yet it worked.

08:40 Now, I mean, I guess this is actually goes quite,

08:41 quite to the sort of Al Gore question and concept.

08:46 But that is something I've felt

08:47 for a while. Freely you

08:49 receive, freely give.

08:51 Which is actually

08:52 actually a quote from

08:57 or quoted in the

08:59 New Testament as being something that

09:01 Jesus said. But it's certainly how

09:03 I feel about open source.

09:05 The free receiving

09:08 and the free giving.

09:10 And for me, that's

09:12 a very significant motivation for why I'm

09:14 involved in open source. That sort of

09:16 mutual giving, which

09:18 I really

09:20 love.

09:23 I'm not going to lie, I feel very

09:24 inspired listening to you speak.

09:27 And I kind of want to let you just go on and on.

09:29 And then, you know, maybe do a playback while I'm going to bed, you know, meditating.

09:35 Thanks so much for sharing this.

09:37 And are all these things that you usually talk about or that your community, you know, asks you about,

09:44 do you sort Do you try to

09:46 not really make these things

09:48 public about your background or

09:50 they manifest in your everyday life

09:52 and your community?

09:55 I mean, I don't

09:56 think I sort of keep

09:58 anything...

10:01 Maybe you talk about them a bit.

10:04 I guess often probably it doesn't come up

10:06 but it's

10:08 I think

10:12 initially

10:14 I probably chose not to

10:16 talk about my different

10:18 background if only because I wanted

10:20 to be taken seriously as

10:22 a developer and sometimes

10:24 I think you can feel,

10:26 particularly when you only have, you know,

10:29 I learned Vue by sort of LaraCast's

10:32 sort of free course on Vue.

10:36 And, you know, you think you don't necessarily,

10:39 like, I don't know, you feel a little bit embarrassed

10:42 if you're not coming from a background of huge experience,

10:46 but you're talking to people who have a huge amount of experience.

10:50 So definitely it's the imposter syndrome

10:54 that I think a lot of us experience tells us,

10:56 oh, you better not let on, you better not say it.

10:59 You should probably focus on just the technical details.

11:03 No need to bring up anything. But

11:07 I'm definitely much more comfortable talking about it now. And in fact, I would want to

11:11 talk about it to represent other people who might also have a less common background in

11:20 terms of coming into open source. Although I say that is less common, I'm not sure it

11:24 is. I think a lot of people have got different backgrounds i say that is less common i'm not sure it is

11:25 i think a lot of people have got got different different backgrounds and it is lovely to see

11:29 that because i think every different background brings different sets of skills and life experience

11:34 and i mean that's part of how you have a healthy community is if you have people with

11:39 different capabilities and backgrounds and gifts and if things things are too, like, homogenous or monochrome,

11:48 then you really lose out, sort of,

11:50 not just from a community spirit point of view,

11:53 but also what you're producing with a project

11:57 doesn't have the same life and direction to it.

12:02 I really appreciate that uh having this conversation so

12:05 far and uh you know i i as much as i want to continue digging into these topics i'd love to

12:11 transition and ask you a little bit about your everyday life today uh running knox what it looks

12:16 like what are some things maybe that people can anticipate what's on the roadmap. Yeah. Well, right now it's an interesting time because we

12:26 are about a year

12:27 since I started leading

12:29 the project. And one of the things

12:31 that I wanted to do early on was

12:34 to commit to a firm

12:35 roadmap and release schedule.

12:38 So we

12:39 said that we would have a

12:41 patch release pretty much

12:43 every week, give or take. A pretty much every week, give or take.

12:46 A minor release every month, give or take.

12:49 And a major release every year, give or take.

12:53 Which means we're coming up to Nuxt 4 as a release.

12:57 And so a lot of what we're talking about as a team and thinking about is,

13:02 what are the deliberate API changes that we're

13:06 going to make for next four?

13:08 Because we don't,

13:10 we don't want those to be big and we don't want those to be difficult to

13:13 migrate to.

13:14 So we want to provide all the tools that somebody will need.

13:17 And whether that's code,

13:19 code mods to help or tools to guide people,

13:21 or even just options to opt into previous behavior.

13:26 We want that to be a very seamless process.

13:29 But still, these are API decisions and design decisions.

13:32 So what are the deliberate breakages?

13:35 Bearing in mind, we basically get a chance to do that about once a year if we stick with

13:39 the original plan.

13:41 So that's a lot of what my focus is on right now.

13:45 Most of the cool features that you might think

13:47 of in association with

13:50 Nuxt don't need a major release.

13:52 Because cool features are enhancements.

13:54 So those are minors.

13:55 That we've hopefully been shipping pretty consistently

13:58 over the last year.

14:00 And if we hadn't been doing it every month,

14:02 we'd probably have bigger

14:04 minors. Though I think think our miners are not that

14:06 but yes we do

14:08 have some nice features planned

14:10 but a lot of what I'm focusing on

14:12 right now is that major release

14:14 coming up

14:15 What would you say is your biggest challenge

14:17 running Nuxt

14:20 and the community which sometimes

14:22 might have conflicting

14:24 interests or trying to pull the project towards one direction or the other Nuxt and the community, which sometimes might have conflicting interests

14:25 or trying to pull the project towards one direction or the other?

14:30 What's the biggest challenge in maintaining Nuxt?

14:33 So I think a big challenge is context.

14:39 So, for example, Nuxt has so many interweaving parts.

14:43 So there's even just the different packages in the monorepo from the CLI to how we test things,

14:51 to our integration with Vite, to our integration with Nitro, the server component of Nuxt,

14:57 to the sort of plugins that integrate with Rollup and more.

15:03 And then obviously there's a whole sub-ecosystem.

15:07 We split up a lot of the utilities we use in Nuxt

15:10 into a GitHub organization called NJS.

15:14 And those exist and are used by Nuxt quite deeply.

15:20 So figuring out this behavior is happening,

15:24 let's just say it's a bug.

15:26 Where is it coming from?

15:28 You know, it might be that it's really an upstream library quite far.

15:34 It might be that it's somewhere in the repo.

15:35 And so the biggest issue is making sure that when triaging or looking at an issue,

15:43 you're able to place that in its context

15:45 even before it can be resolved.

15:50 And there are obviously a huge volume of issues

15:52 and requests for help coming through.

15:56 And so basically handling,

15:57 I think that's the most bottleneck part

16:02 of the NUX community.

16:04 A lot of other things are working quite well.

16:08 But yeah, finding people who know the context,

16:11 that's a very difficult thing.

16:15 It feels to me like there's a lot of movement,

16:18 a lot of dynamism,

16:20 and very little pulling in opposite directions in NUX.

16:22 So a lot of things we offer, we have choices.

16:25 So you can opt into one behavior or another.

16:28 And we experiment with those a bit.

16:31 And if you want your own different behavior,

16:33 there's nothing to stop you from building that yourself

16:35 because it's quite extensible,

16:37 which means that it's not so final.

16:41 So if you say, oh, we don't think this behavior

16:43 should be built in from a

16:46 vision roadmap

16:48 point of view, it doesn't necessarily

16:49 mean that people are

16:52 unable to

16:54 proceed. They can still

16:55 use the behavior they want,

16:58 which is nice.

17:00 But yes, obviously

17:01 where you have to make a decision,

17:04 even if it's as simple as what

17:05 side of the road you drive on, it's not, it doesn't matter.

17:10 And decision making, do you feel like it rests a lot upon your shoulders?

17:21 Is there more communal decision making process? What does it look like

17:26 in practice for such a big project? I don't think it's just a need to make decisions.

17:33 I mean, it could be, but I think my point of view is, so I have the responsibility to make

17:41 the decision, but when you have such a talented team,

17:48 as Sebastian, who wrote it in the first place,

17:53 or Pouya, or Anthony, or Alex, or Lucy, or Harlan,

17:56 or Julian, or whoever,

17:59 why wouldn't you get that viewpoint? And I'm very happy to change my opinion.

18:04 So it often ends up being more collaborative.

18:07 If people have strong opinions,

18:09 then obviously we try and resolve those first.

18:13 At the end of the day, I'll make the call,

18:16 but that's not because my opinion is right by any means.

18:20 For people who are not too familiar with the community,

18:24 the project or the ecosystem surrounding it, the ecosystem that Nuxt has created,

18:30 is there some information you could share with us, a big picture about what it's looking like today?

18:37 About the community?

18:39 About the community, the ecosystem, the attraction of it, the size, the organizations involved.

18:47 So, well, and I guess I should probably say what Nuxt is. It's a framework for building web

18:52 applications with Vue.js as the front-end library and built on Nitro as the serverless framework

19:00 that covers the backend. It has a real focus on developer experience and also extensibility.

19:09 So you shouldn't need to make any changes.

19:13 So best practices should be built in,

19:14 but you should be able to change almost every part of it.

19:17 And this was really part of Nuxt from the very beginning.

19:21 And that is, I think, what has helped make community.

19:26 the very beginning. And that is, I think, what has helped make community. So we have this concept of a package called a Nuxt module. So a function, effectively, that can hook into different parts

19:34 of Nuxt's build process and even add runtime code or change behaviors or defaults, whatever.

19:44 And that is incredibly powerful.

19:46 So Nitro, for example,

19:47 was originally implemented as a module for Nuxt.

19:51 And I think almost everything you could do,

19:54 you could do with the module.

19:56 So that means that people can build their own integrations.

19:59 We have a directory on GitHub,

20:01 which is just full of, just has YAML files.

20:10 That is a sort of index to all the Nuxt modules that want to be added to it.

20:12 And that's also on nuxt.com as well.

20:17 You can browse nuxt.com forward slash modules and see some of what there is.

20:27 So third parties like CMSs or deployment providers or error tracking, or whatever, can build their own integration.

20:31 And people who have good ideas can also do the same.

20:38 And often, even features that ultimately are added to Nuxt at the core level start out as a module,

20:40 so people can test it and see.

20:46 So that has, I think, made a huge difference to Nuxt and to the Nuxt community,

20:54 because people can, when you think about Nuxt and its evolution and growth, I think you really, the community are a very significant part of that. Without people building modules and pushing

21:00 forward, Nuxt wouldn't be pushing forward. You forward. We're not a company or enterprise-driven project

21:07 like some of the other frameworks that are out there.

21:11 We are very much a community-driven project.

21:14 And so, yeah.

21:16 So the community, I think, is very healthy.

21:18 We have lots and lots of modules.

21:24 And a lot of people are using Nuxt.

21:27 I don't know really if I should share this,

21:31 but just today we hit an all-time high

21:36 in terms of projects and sessions using Nuxt.

21:41 So we have anonymous opt-in telemetry,

21:51 using Nuxt. So we have anonymous opt-in telemetry, which just tells us that a dev session started with Nuxt.

22:05 And so today we had over 10,400 projects using Nuxt. The dev command was started. So that's pretty cool. I was really excited to see that.

22:07 So it's healthy.

22:10 There are a good number of folks using it.

22:13 Exactly.

22:14 And I'm wishing you a lot more growth for the community

22:17 and the project moving forward.

22:19 That's incredible.

22:20 And you just mentioned about the fact that there's know, there's not an enterprise behind Nuxt.

22:27 It's just a community. And I guess that brings us to our next conversation topic, which is about the sustainability of the community and the organization.

22:42 are going towards the commercial direction,

22:44 creating companies behind, getting funding,

22:46 so that they can also use to,

22:50 so that people can work full-time on the project.

22:53 You mentioned earlier how a lot of people are sponsoring you.

22:56 Not a unique concept in open source,

22:57 but definitely prevalent.

23:03 And so how is all this looking like today?

23:07 How have you experienced it personally, being at the helm?

23:12 And what could other people learn from this? And maybe what challenges are also inherent with this format?

23:18 Well, I am maybe not the most representative example

23:24 because I'm not only sponsored by individuals,

23:27 but I'm also sponsored most notably by Nuxt Labs,

23:31 which is a company that was created

23:34 by the same people who created Nuxt.

23:36 So Sebastian Chopin and his brother.

23:39 And they provide a whole raft of services.

23:42 So they fund a number of Nuxt developers across the ecosystem, like me, like Anthony Fu and Booyah.

23:51 And they provide services like consultancy, and they have some software as a service products as well.

23:59 So there's Nuxt UI Pro, there's Nuxt Studio, and so on.

24:04 There's Nuxt UI Pro, there's Nuxt Studio, and so on.

24:11 And that was basically created in order to try and fund open source. So the purpose of it was the funding of open source.

24:15 And I'll tell you, I really, really respect how Sebastian has done that

24:20 and how open source doesn't take a second seat in terms of his vision.

24:27 So it doesn't influence the direction of the framework, for example, to serve the needs

24:33 of NUX labs.

24:33 It's sort of the other way around.

24:37 It's, and I should say, I am able to spend most of my time working on NUX.

24:44 I'm pretty much a full-time maintainer.

24:48 But I think that is totally unusual because I think a lot of times it does take this sort

24:56 of, it takes either popularity, so seeing a project become incredibly well-known, just

25:03 so that there are so many people who might

25:05 donate or sponsor the project. Or it takes a company or something that is driving adoption

25:13 or sponsorship or revenue in some way to make it possible. So you look at Tailwind CSS, for example,

25:20 which is a phenomenally, hugely used open source project. And it's basically funded

25:25 through Tailwind UI

25:27 and some of the other

25:29 products that they put out.

25:32 Or you look at

25:33 Next.js, which is funded

25:35 as far as I'm aware, primarily through

25:37 Vercel, the company.

25:39 And, you know, there are

25:41 others out there.

25:43 I'm not sure that's entirely correct.

25:45 Of course, you mentioned Nux Labs already and how they're supporting, of course, the Open Source project and its priority.

25:55 Have you personally taken up any consulting work or, you know, has this been part of your work?

26:02 has this been part of your work?

26:04 So I do

26:06 consult as well with

26:08 companies and

26:10 individuals too.

26:13 I'm sort of, I'm a bit limited

26:14 in terms of

26:16 time because

26:18 leading the NUX

26:20 team is probably quite a

26:22 significant thing.

26:25 But I do really value consultancy

26:28 and doing things outside of the core.

26:31 If only because, and I would say this

26:32 to anyone building an open source project,

26:36 that actually seeing how people use it

26:37 and building things with it,

26:41 which is another reason why I try and build stuff with NUX,

26:44 really changes how you interact with it, which is another reason why I try and build stuff with Nuxt, really changes how you interact with it.

26:47 So you see the need or the bad DX decision or what needs to be fixed much more clearly

26:54 when you're actually using it on a day-to-day basis

26:57 than if you're just focusing on building that project.

27:01 So I often put out like tiny tiny

27:06 projects they're not huge

27:08 but it is really useful to see

27:10 and the amount of time I've spent for example

27:12 on my own blog totally

27:14 outweighs the importance of the blog

27:16 my website isn't that important

27:18 but you know I try every

27:20 edge case scenario

27:22 and so on and it's really nice

27:24 to have something where you can tinker

27:25 and just try and focus on solving some issue or figuring out some need.

27:32 And, you know, it's when you're micro-optimizing.

27:34 We all micro-optimize, don't we?

27:36 When you push something far enough, you micro-optimize at the end

27:39 and actually be able to take those micro-optimizations

27:42 and bring them back into the core of Nuxt as well.

27:44 It's quite a good thing. Quite a good thing to be able to take those micro-optimizations and bring them back into the core of Nuxt as well. It's quite a good thing.

27:47 That's sweet.

27:48 So it sounds like a lot of times some implementation,

27:53 something new that was built custom,

27:55 maybe for a customer in a consulting format,

27:58 makes its way into the core project sometimes.

28:06 Yeah. into the core project sometimes. Yeah, so I think, for example,

28:08 I was able to use my website to do a lot of debugging

28:11 with NUC server components

28:13 and figuring out how you would want to use them

28:16 and what the best way of...

28:22 the best kind of API from a user point of view,

28:26 what was it like to use them.

28:27 That's certainly been the case.

28:31 I mean,

28:32 the vast majority of my time is

28:33 actually just triaging issues and

28:36 fixing bugs and building features and

28:38 nuts. But yes, absolutely

28:40 stuff does filter in from

28:41 testing things out.

28:44 And it's definitely a privilege for

28:46 organizations to have you and collaborate

28:47 with you on these

28:49 kind of custom

28:52 projects. No, absolutely.

28:54 Just one last

28:55 follow-up question here, because

28:57 for a lot of open source maintainers who

29:00 there's no

29:01 commercial entity behind their project,

29:04 they're working on this on their own free time.

29:07 Do you recommend this sort of format for people

29:10 who would like to work more full-time on their project

29:14 and make it sustainable for themselves and their community

29:17 to take up this, take consulting work with, say, an agency

29:24 or a startup that is using their technology.

29:28 Is there any word of caution or something that is maybe a problem that people should keep in mind

29:36 compared to not the many other options?

29:40 But yeah, this format, what more could we say about it?

29:43 So, I mean, I think consultancy can work very well.

29:47 So when I did more consultancy, I was paid much, much better, for example.

29:54 So from a purely financial point of view, it can work very well to be doing consultancy as opposed to full-time open source.

30:06 And I think it could be quite a good bootstrapping opportunity if you start out that way.

30:11 But I think a lot of the things about consultancy,

30:16 there are a lot of things, for example, that I don't like about consultancy.

30:20 So consultancy is a very specific, it can be very specifically about your time,

30:31 So consultancy is a very specific, it can be very specifically about your time, which means that, so I'm very people orientated.

30:33 So I hate billing people.

30:37 I don't like asking much of them.

30:39 I like to do things as fast as I can.

30:48 And so that means that I might have like five minutes of available time. And that is very frustrating because you contact switch, you fix the problem.

30:51 And then I don't know, you build five minutes.

30:54 The whole thing is not necessarily a great, great experience.

30:59 You also have the whole, you basically, if you're a consultant, consulting you,

31:06 and you don't have a platform to somehow get people to ask you to do things.

31:10 You have to do marketing, and obviously you have to do all of your bookkeeping and your accounting and your invoice creation, obviously time track.

31:15 There's a whole business that you need to implement, which is really tricky.

31:21 And some people are great at that and love it,

31:23 and focusing on like the bones

31:25 infrastructure. I mean, I've got to say, I have invoices I probably should send out. I have,

31:31 in fact, an invoice I probably should have sent out several months ago. I just, I hate doing it.

31:36 I really hate it. So I'm probably not a very successful freelancer in lots of ways. But

31:44 so I don't know. It depends on who you are, maybe. a very successful freelancer in lots of ways.

31:46 So I don't know.

31:47 It depends on who you are, maybe.

31:50 Thank you for the perspective.

31:54 Do you think part of this, I mean,

31:58 the need from businesses is recurring and big,

32:01 and there's only so many people like yourself,

32:04 but there's certainly a lot of experts in the community. Do you think

32:05 eventually some members of the community, some Nuxt experts could take up some of these

32:11 projects, some of these, you know, contracts? Could there be a sort of like, quote unquote,

32:18 you know, little internal marketplace of some sort? Is that something you ever considered?

32:23 Or you think it's a bad idea?

32:25 If that made sense.

32:27 For next consultants,

32:30 for people to find them

32:31 and group them.

32:34 I think

32:35 if there's a way to do that, where it

32:37 is...

32:40 So, I guess

32:42 a lot of it is about

32:44 trust.

32:47 So if you're hiring someone,

32:49 and I've absolutely been on this side of things,

32:51 obviously hiring people to do work,

32:53 it's very difficult to figure out if they know what they're talking about.

32:55 And even if they do know,

32:57 how reliable are they now on a day-to-day basis?

33:01 How efficient are they at working?

33:03 Can you leave something with them?

33:05 Do they need to be managed?

33:08 And ideally, a consultant doesn't need to be micromanaged.

33:12 You give them an objective, you set parameters, and that happens.

33:15 There's no need to check up, check in, and so on.

33:21 And unfortunately, you know

33:25 you can

33:26 there are people

33:27 out there

33:27 who are

33:28 who aren't

33:29 necessarily very good

33:30 at what they do

33:31 or even who are

33:33 straight up scammers

33:33 it's very

33:34 it's very very tricky

33:35 if there's a way

33:36 to make a marketplace

33:37 work

33:38 in a way that

33:39 I think that could be

33:41 really useful

33:42 to the ecosystem

33:43 absolutely reduce the friction make discovery available keep it outcome based I think that could be really useful to the ecosystem. Absolutely.

33:46 Reduce the friction, make discovery available,

33:48 keep it outcome-based,

33:51 things like that.

33:53 Hooking things up with GitHub

33:54 accounts and

33:55 portfolios of work that can be

33:58 verified or authenticated somehow.

34:00 I think there could be

34:01 a huge opportunity there.

34:05 Duly noted on our end.

34:08 Absolutely.

34:11 Thank you so much for these insights here.

34:15 A little quick change of pace.

34:17 You are quite active online on Twitter.

34:19 I wonder if you could share a little bit about how you've experienced it, how it has helped the project, any advice for people that are not too active.

34:30 And then maybe if there's some people you really like following that you might like to give a shout out to.

34:36 I mean, I don't know that many people wouldn't already be following him, but I would certainly recommend Anthony Fu.

34:46 He is certainly inspiring to follow

34:49 because you can just see the pace at which he creates projects.

34:55 He's fantastic.

34:59 I would recommend, who else would I recommend?

35:02 If you'd like a dose of positivity in your

35:05 feed, follow

35:08 Scott

35:10 Hanselman. He's not in the

35:12 Vue community, but he has an

35:13 invariably positive voice on Twitter,

35:16 has a lot of interesting things

35:18 he's up to.

35:21 I've been

35:22 I would also also by the way

35:25 it depends a little bit on

35:28 the kinds of things you're interested in

35:30 so

35:32 I would

35:34 highly recommend

35:35 if you're interested in TypeScript performance

35:38 follow Alexandra

35:40 says, Alexandra Shikara

35:42 she's very good on

35:43 giving you some

35:46 tips and understanding how things work

35:48 there.

35:50 I would also say

35:52 that...

35:55 What would I recommend?

35:56 There's so many interesting folk to...

35:58 Maybe someone that makes you laugh. I don't know.

36:00 But yeah, I'll let you mention it.

36:02 Follow Debbie O'Brien.

36:05 She's great.

36:07 She's on the Playwright Corps team and

36:10 I think

36:12 representing them as a

36:13 developer advocate. She's pretty

36:16 busy right now. I think her baby

36:17 she has twins. They were just

36:20 born in December.

36:23 But she

36:24 is an invariable voice

36:26 of warmth and

36:27 definitely will make you laugh

36:30 with some of the jokes.

36:32 Definitely follow her.

36:34 There's lots of folks.

36:36 Absolutely. Do you oftentimes

36:38 get together with people in person

36:40 in the context of conferences or other meetups?

36:42 Are things planned maybe for this

36:44 year?

36:46 Yeah, absolutely.

36:47 So I'm an extrovert at heart.

36:49 So I love it when I can meet people.

36:53 And particularly if they're people I've collaborated with

36:55 or know from GitHub issues or Discord or something.

36:59 So I'm totally in for that.

37:03 And that's one of the main reasons why I go to conferences.

37:06 So I just because, you know, otherwise you can,

37:10 working from home is great.

37:11 I love it.

37:12 And I really would not want to change it.

37:13 But at the same time,

37:14 the thing that you really lack is connections with personal connection with

37:19 people.

37:21 I've, I've even, apart from conferences,

37:23 I've even gone to Paris paris amsterdam just to meet

37:28 up with people on the team um and actually just spend a couple of days and uh do some co-working

37:34 and so on and it's totally worth it so absolutely i try and go to about a conference every month

37:41 that's my my aim um and so i've got a few already booked in

37:47 for for 2020 maybe we can meet up in one of those conferences hopefully uh i mean there's a lot going

37:54 on in europe so yeah hopefully soon um before we proceed to the demo of next are there any um

38:03 last things you might like to mention to people, advice,

38:06 people getting started with open source or to other

38:08 maintainers?

38:10 Any final notes

38:12 and remarks?

38:16 Well, do, if you

38:18 would fancy it,

38:20 I stream often on

38:22 Twitch, just

38:23 random stuff,

38:25 coding,

38:26 fixing bugs,

38:27 creating new projects.

38:29 So that might be worth mentioning.

38:32 If you want to,

38:33 I also have a YouTube channel where I post the past streams.

38:37 If you feel,

38:39 feel no pressure,

38:40 it doesn't work for everyone,

38:41 but I really enjoy it.

38:42 And I,

38:42 I enjoy being able to chat to

38:45 people as i code as i said i'm an extrovert so it works really well um for me so do you feel free to

38:53 join me if you have a one and i should say alex lichter is putting out some phenomenal youtube

38:59 videos on nuxt and how things work internally. They are really great. His thumbnail game is excellent.

39:06 And obviously, he's on the core team,

39:08 so he knows what he's talking about.

39:10 And that's definitely something to check out.

39:14 Thank you so much for the interview so far.

39:16 And whenever you're ready, we could transition to a demo.

39:21 Sure.

39:22 Great.

39:23 Okay.

39:24 Let me fire up an editor. So code. I will share my screen.

39:34 One more speak. Precisely. I will share my window. Okay. This is a new Nuxt application.

39:41 Is that coming through clearly for you? Mm-hmm. Great.

39:45 So I've just run Nuxy in it.

39:47 And it's created this folder,

39:50 which has got sort of the bare bones.

39:53 And actually, most of the stuff here

39:54 is just infrastructure.

39:57 The only code we have is app.vue.

40:00 I'll start the dev server,

40:01 which powers most of what Nux does

40:03 in the background.

40:05 So Nux does a lot of things under the hood for you.

40:10 So, for example, you'll see that I have this component, which is made available just as a welcome component.

40:18 I'll open this in the browser.

40:22 So you can see the welcome component renders this

40:28 thing. And you can also see that it has

40:32 some props, so you can actually change some things about this.

40:34 And you get type-safe access to that.

40:38 So I think it looks like there's this star GitHub

40:40 prop, which we could say

40:42 star us, something like that.

40:46 And that will change some text somewhere in here.

40:50 I'm not quite sure where it is.

40:51 But if you were to pass an invalid prop, that would throw you an error.

40:56 So Nuxt has stuff like that built in.

40:59 And you could do the same if you created your own component.

41:02 You could do something like test me.

41:04 Create a test component.

41:09 Hi there.

41:10 And if we go back and use that test component,

41:14 we could, well, one, your editor knows that it exists.

41:19 And when you put it in, it sort of immediately works.

41:22 So that's to automatically import that component

41:25 only where it's used in your app. So it's bundle split out if it needs to be. And then you also

41:30 get stuff like define props. And we'll say this one has a prop of message. That is then going to

41:39 be, oh, and if I say maybe required true, then I'm actually going to need to supply it at this point,

41:50 or I'll get an error in the message.

41:55 And we can actually view compiler expecting a comma.

42:00 That makes sense.

42:02 And we should also see that this is typed for us and so on.

42:06 So you get a lot of sort of good developer experience stuff out of the box.

42:11 The same goes with view composables.

42:14 So a lot about Nuxt, we've really leaned into the move from view two to view three,

42:20 which was basically moving from everything being built in,

42:24 having a sort of composable

42:25 logic concept so you could pull in what you're needing only when you need it and so uh so we do

42:34 that with not just the client side or the next built-in utilities but also your own code so you

42:40 could do something like use my stuff.

42:46 And you could create your own composable, do some logic, return something.

42:51 And then again, you can use that easily in your app.

42:57 So you could just do something like use my stuff.

43:02 And so that will be typed and available to you.

43:04 So it knows, for example,

43:06 how it comes back.

43:08 So you get this really nice

43:10 sort of all-the-way-through, end-to-end

43:12 process of

43:14 what is hopefully quite a clean

43:15 approach, rather than needing to put all the imports.

43:18 But TypeScript, and you're

43:20 added to know where they're all located, so when you

43:22 refactor things, things work.

43:23 Or when you click through, stuff works.

43:26 We also have this Nitro server, which is, like Nuxt, uses file-based routing.

43:33 So you can do stuff like, we'll call it Foo.

43:39 You could create an API endpoint.

43:41 And again, we could return something, some value.

43:46 endpoint. And again, we could return something, some value. And you can actually access that with this $fetch utility, which will actually be type hinted for you. So you can fetch that.

43:55 So we could actually just display it here. And that's going to be the data that the server

44:01 endpoint returns. In fact, in the server rendering lifecycle,

44:05 when you're doing fetches to your own API,

44:09 we don't even hit the network layer.

44:11 We'll just make an internal request

44:13 and it will emulate the network layer,

44:17 which basically makes this possible to be incredibly fast,

44:20 which is built for serverless.

44:23 Because we, again, every time you would hit the network layer with serverless,

44:26 you would have to potentially invoke a new instance of your Lambda.

44:31 So we can just avoid that entirely.

44:34 And we can even,

44:35 we even supply view composables that do stuff like prevent refetching on the

44:43 client so that there's a cache. So when this

44:46 data is fetched, we're able to

44:48 fetch it only on the server and not

44:50 refetch it on the client,

44:52 which is pretty useful. And we

44:54 can pass lots of

44:56 other request

44:58 handlers and other

45:00 utilities that you can do there.

45:01 That's pretty nice, I think.

45:07 Next, it's always had the concept of

45:14 plugins, so code that runs when your server starts. So you could do something like,

45:24 you could have an auth plugin, for example, and you can inject, I guess, a global singleton and you could

45:25 maybe have your auth

45:27 user and have a

45:29 login function or something like that.

45:32 And when you do that, you then

45:33 can access it with

45:34 useNapsAppAuth

45:37 and you would then be able

45:39 to again do the same kind of

45:41 stuff where you

45:43 basically are able to

45:45 handle

45:46 things that are in that plugin.

45:48 We provide

45:49 callbacks, so you could do something like

45:52 on Nuxt ready, so once everything is

45:54 all loaded and your app is interactive

45:56 then we're going to load a third party

45:58 script.

46:01 We provide

46:02 sort of, obviously Nuxt is

46:04 server side rendered, so you can do things like use we provide obviously Nuxt is server-side rendered

46:05 so you can do things like

46:07 use SEO meta

46:09 and just put some

46:11 stuff in like

46:13 title and description

46:14 and that will

46:16 make a pretty easy

46:19 way of interacting with things

46:21 you can have

46:23 layouts and root middleware. You can have layouts and root

46:26 middleware, and

46:28 you can have

46:29 your own modules that integrate

46:32 with Nuxt.

46:33 There's lots and lots and lots of stuff.

46:36 I'm just talking. What's interesting to you?

46:38 What else should I show you?

46:40 What does

46:42 Nuxt do?

46:43 What does it not do? What does it not do?

46:47 Well, we try to make it. I mean, I would use Nuxt personally in almost any situation.

46:53 Even if you don't need a view, we actually have lots of experimental features.

46:59 We can even turn off, for example, what is it?

47:04 Oh, I think this might now be in features.

47:07 You can even

47:08 turn off

47:08 scripts entirely and have a purely

47:12 HTML rendered

47:13 app at the other end.

47:16 And that's

47:17 interesting.

47:20 How would you

47:21 describe the learning curve

47:23 of Nuxt?

47:25 And I mean, it looks not scary at all now that you're walking us through it.

47:30 But for a junior engineer, someone maybe in college, going out of it, a beginner,

47:37 what are you hearing from people in terms of their first experience with Nuxt?

47:42 terms of their first experience with Nuxt?

47:52 So I'm hearing, I mean, I think I'm hearing, I'm always happy when I hear both positive and negative stuff.

47:53 I mostly hear positive stuff.

47:55 Obviously, there's a danger of self-selection there.

47:58 The things that people tell me tend to be more positive.

48:03 I hear different things based on whether people use NUX2 or not.

48:07 Because I think a lot of the things about Vue 3

48:10 that are, I think, good decisions,

48:13 people also might not like if they are coming from Vue 2.

48:17 So sometimes there's a difference based on that.

48:21 And if your first experience is with NUX3,

48:24 then things will be more intuitive

48:28 than if you are coming from a different way

48:30 of using Nuxt in Nuxt 2.

48:33 Generally speaking, the feedback is really positive.

48:35 And I'm really, really happy

48:36 if people give me feedback that isn't.

48:39 But the very best way, of course,

48:41 is when people don't just give feedback

48:43 on is this good or not, but when they become part of a solution.

48:47 So they have an idea for improving it or thinking through an API that might be a better idea or even just take responsibility for it.

48:58 There are obviously, there are sometimes, you know, people who come in as sort of with ultimatums.

49:04 Okay, if you don't change this,

49:06 I won't use the project.

49:08 That's not a great way to interact.

49:12 Because the whole thing that's motivating me

49:14 and most other maintainers

49:15 is that we want to help people

49:16 and we want to create a project

49:18 that people like to use.

49:19 And it's like you're going directly

49:22 for the jugular.

49:23 You're like, yes, you thought you did something good. I'm going to tell you it's terrible.'re sort of going directly for the jugular. You know, like, yes, you know, you thought you did something good.

49:27 I'm going to tell you it's terrible.

49:28 Unless you change it, I won't.

49:30 I'm going to reject you.

49:31 Like, that's a very sort of...

49:33 I think if you don't realize that open source is a very personal thing

49:37 and that, you know, it's a very, very emotive...

49:41 One of the reasons people burn out in open source

49:45 is because they pour so much of themselves

49:48 into it because

49:48 it's very much of them.

49:51 So when you reject a project,

49:53 it can come across very much like you're rejecting a person.