Episode 293 - Daniel Roe and Matias Capeletto
[00:00:10] Richard Littauer: [00:00:00] Hello and welcome to Sustain, the podcast where we talk about sustaining open source for the long haul. Who are we? Where are we coming from? Where are we going? I think my guest said that my voice sounded like it was about to make a pitch, so why am I speaking with an accent? Very excited to have our guests on here today.
There are no other hosts. It is just me solo hosting. I am your host, Richard Littauer. Hello, everyone. And I am joined today by Matias and Daniel. You may know Matias as Patak, who is how I know him. That's Matias Capoleto. He is a steward of npmx and also on the Vite core team, possibly Vite core team. I really need to learn more about current web technologies.
I feel like I've been falling behind. And then we're joined also by Daniel Roe. Daniel Roe is a member of technical staff at Vercel, [00:01:00] also used to be one of the core team leading Nuxt and a steward of npmx as well. That's right, this is an npmx episode Daniel, how are you doing today? 
[00:01:10] Daniel Roe: I am doing well. It is a summer's evening in Edinburgh, which is a lovely place to be enjoying a summer's evening.
I have a glass of wine. I'm looking out at the leafy green trees of Bruntsfield Links, and I'm talking to two very nice people, so I'm having a nice evening. 
[00:01:30] Richard Littauer: Truly the most beautiful of things. There's a wonderful Norman MacCaig poem about Bruntsfield Links, which I will send to you afterwards. And calling in from Hungary today, Matias Patak.
How are you doing? 
[00:01:41] Michael Capeletto: I'm doing very well. Thanks a lot for the invite finally to talk with you. And yeah, here in Hungary was also a very nice day. We did, like, 24 kilos of pickled cucumbers for the winter, so we are in winter preparation in the middle of, like, 40 degree here or something. So, nice. [00:02:00] Nice. That sounds 
[00:02:00] Richard Littauer: amazing.
[00:02:00] Michael Capeletto: I, 
[00:02:01] Daniel Roe: I was gonna say, is it winter or winter-ish where you are? 
[00:02:04] Richard Littauer: It is not. It is spring. It is spring. It is 7:00 AM. The rosy-fingered dawn is rising over the ocean, which I can see as I look out from this podcast. I've already gone for a run. I already had a whole Community Morning Session at 5:00. I got up at 3:00 today.
Wow. Let's move on to more fun stuff. NPM X, what is it? 
[00:02:22] Daniel Roe: Well, good question, and probably we, we didn't do anyone any favors by naming it NPM X. Uh, I've had a lot of questions about that, but, but, you know, it's, it's better than all the alternatives. NPM X was originally a faster alternative view for npmjs.com.
And for people who don't know what npmjs.com is, which is probably, uh, quite a lot of people, NPM is a package registry, uh, for probably the main one if you're using JavaScript or TypeScript, and there are three things that NPM does. It's the registry itself. It's also a CLI tool used for installing [00:03:00] packages and interacting with that registry, and it's also a website used to view the registry.
Yeah. And there are alternatives for the package manager. You can have, uh, PNPM or Yarn or Bun or Deno also supports the installing it and, and there are more, Nub and Ob and plenty more You can have a different registry. So there's Volt, for example, which just came out. There's JSR. But I felt there was a bit of a gap when it came to websites or browsing the registry, the npm registry.
And in particular, I was ... Uh, sorry, I, I maintain Nuxt, which is a framework for building full stack web apps. And one night, very early in the morning, I should say, not quite as early as your 3:00 AM start today, but I woke up, I think it was 4:00, and I checked my company Slack, and I was nerd sniped into starting to plan how I could make a really fast website for browsing npm using a new feature that I was going to release in Nuxt.
And so I, I dived in, [00:04:00] posted I think later that day on Blue Sky to find out what people's frustrations were about npm, and pretty soon I think I knew this was something that wasn't just a me thing. It was something that a lot of people felt there were frustrations they were experiencing. I posted again, I messaged Matias and a few others, and quite without any kind of fanfare, certainly without mentioning it, I think we had something like 70 PRs by the end of the second day-
by people who just wanted to- Sorry. That's amazing. I know. 
[00:04:31] Richard Littauer: When 
[00:04:32] Daniel Roe: was this? 
[00:04:32] Richard Littauer: What year was this? 
[00:04:33] Daniel Roe: This was this year. This was in January. 
[00:04:35] Richard Littauer: Six months 
[00:04:35] Daniel Roe: ago. Oh my 
[00:04:36] Richard Littauer: God. 
[00:04:36] Daniel Roe: I didn't post the address or anything. This was just DMs, and I guess people mentioning it to other people. I had an MVP, I guess, but everyone who contributed made it better, and everyone had their own thing that they wanted to make better, and we were sort of mob programming.
But I think it was about a week we had 1,000 PRs. But I, I should say that very early on I asked Matias if he'd like to co-lead the project with me. And so we are [00:05:00] co-stewards of npmx, and it's a huge project now in terms of the people who contribute, and that is, I think, what mostly matters when it comes to open source projects.
It's the people. And npmx is so many people. There are so many of us. I think we have a core team of about 10- 12 ... maintainers of about 20, and in terms of total contributors to npmx, I think we have well over 200 who have contributed in code. But I think we have more like 600 in our Discord server. How accurate are those numbers, Matias?
[00:05:30] Michael Capeletto: Pretty good. I see we are- 
[00:05:31] Daniel Roe: Pretty. 
[00:05:31] Michael Capeletto: Pretty good. 
[00:05:32] Daniel Roe: Okay ... 
[00:05:33] Michael Capeletto: already at, already at more than 300 contributors. Let me check quickly 
[00:05:37] Richard Littauer: That's incredible. I mean- 
[00:05:39] Michael Capeletto: Yeah, 300, 320 ... outside of, 
[00:05:40] Richard Littauer: outside of like... 320. 
[00:05:42] Michael Capeletto: Mm-hmm. 
[00:05:43] Richard Littauer: Outside of OpenClaw and, like, random Hugging Face type things, I haven't heard of a project that grew that fast, that quickly.
Just because I wanna wrap my head around it, can you tell me what was the technical advantage that npmx [00:06:00] offered over npmjs.com or something? Like, what was the reason everyone's like, "I need this right now"? 
[00:06:06] Daniel Roe: Well, I'm really, really tempted to say that it was Nuxt, and that Nuxt is obviously the best framework for building for the web.
But honestly, you could probably build something fast with any framework. Okay. I think the main thing is that npmjs.com was probably nine years ago, and hasn't been profoundly updated since then. Okay. But there are a lot of new technologies that make the web fast today, and there are a lot of frameworks that are pioneering the front end as, like, the leading edge of making things fast, including, I would like to think, Nuxt, but also plenty of others too.
And I think it would've been possible to write an npmx in another framework as well. We really don't mean to throw shade at npmjs.com. It is a single website that was written nine years ago, which is still serving its purpose, and that is a phenomenal testament to the people who wrote it. But it is dated for that reason, and Microsoft haven't chosen to invest [00:07:00] time in it, and that is a decision of, of management and, and so on.
And it does mean that we have the opportunity to do really cool things from the community and build the NPM registry browser that we all want. 
[00:07:12] Michael Capeletto: Yeah. I think one of the features that- Is not that special, but like probably a lot of people were talking about was dark mode, because like after nine years, people were like switching from their IDs to these like white massive blob, and that was like one of the top features.
But yeah, like it was fast, like the index is using Algolia, for example, instead of like the default, and that is already a lot faster how to get the results. And quickly it got so many features, as Daniel was saying, that with community members that are starting to contribute things like the graph, for example, of downloads, that instead of being like this little thing on the top, you can actually click it and you have like all this expanded beautiful graph that you [00:08:00] can explore and put whatever that you want.
You can compare different packages. There is so much built into it, and I think that that was really important and one of the reasons why after that initial move of like 1,000 PR probably in two weeks, that was, uh, this kept going. And it was because a lot of people felt, wow, like we're gonna stop complaining, and we're gonna start to actually take into our hands and like do the features ourself.
And they got to a project where they were welcome. They were like treated as peers, and they was like quickly onboarded into... Like by a lot of the others, it's not even us, because like you see the welcome channel, and it was like 700 people welcome by a lot of different peoples there and talking and like really a lot of human connections.
And it was very special for like a lot of us that were a little bit maybe lost that connection with all the noise [00:09:00] from companies telling people that they're going to be replaced, and they are not needed anymore. And this group of people kind of assembled and created something that it was important for them.
So that I think it was part of why this worked so well. And like two things to mention there, like you mentioned like apart from OpenCLO, there weren't other project that came to mind at this speed. And if you take the new projects that appears in Q1 of this year, from these new projects in GitHub, this was the top project in activity and contributors.
And if you see the top 10, all the others in the top 10s are AI projects. So this was the top one and the only one that was not AI related. That is one of the things, and the other one is that we were having so much fun that we were not sleeping at all, a lot of us, and we were starting to say to us, "We should sleep more.
We should take care." Like Actually do an open source project where we really take care of [00:10:00] the people. And we decided in week three when this was going exponential to actually take a vacation, and we closed Discord, all the channels, even the welcome channel. The only channel that, that we left open is our garden channel, that is like this random channel where people can discuss about off-topic.
That is the only channel that was kept open. We didn't merge PR, we didn't do anything for a week. It coincided with Daniel and me having, like, a vacation in the mountain too. We went to ski and snowboard, and it was very good. Like, we came back, and it was, like, getting together again and continuing the project.
But that week in the middle there was really, really interesting. Like, we even removed the link to Discord and instead of, like, going to Discord you were going to a page in the website saying, "We are on vacation. Came back in a week." That was like, can you do that in open source? Like, can you actually put a hold- 
[00:10:54] Richard Littauer: Have you heard of anyone doing that before?
Like, what, what inspired that move? 
[00:10:57] Michael Capeletto: I think it was a little bit random and lack of [00:11:00] sleep maybe, or it was, uh- Yeah. I love 
[00:11:02] Richard Littauer: it. I think it's great. So. 
[00:11:04] Daniel Roe: It was Matias' idea, and I was a bit annoyed actually. Because Matias and I were both going on holiday and- ... you know, one of the things that I love doing on holiday is coding.
Mm. I really like it. Coding for me is fun. Don't tell anyone, but I would still do this even if I weren't paid to do it, and, I mean, that's how I started, you know. This is fun. And so I, I do it on holidays and I do it while watching Netflix, and it's great. It's amazing. And then Matias is like, "Now we're going to have this week where we can't do anything on npmx."
And I was so annoyed. I thought, "You're the thief of joy. What are you, Matias? Why? Why are you doing this?" But it wasn't for me really like, oh, it cleared the space for me to... you know. It, it didn't. I guess unintentionally it did, but much more important than that was it planted a flag in the sand and said, "This is the kind of project that we're about.
This is what open source is about. It's a marathon, not a sprint. It's not about the grind culture, [00:12:00] it's about the joy culture." And that is, I think, one of the reasons, not because we did the holiday, but the thing behind it For a while now, companies have been saying to their employees, "What matters is how much you produce.
Produce more. You've got to use AI because that will help you produce more." How many 
[00:12:16] Richard Littauer: tokens you 
[00:12:17] Daniel Roe: max? Yeah. Maybe you don't even need to review each other's PRs. Just have AI review your PRs, and that way you can merge faster. You can produce more. And the thing that is missing from all of that is connection with people.
So then npmx comes along and partly unintentionally, and partly very intentionally because Matthias tends to create a warm and welcoming and lovely place wherever he goes. npmx was all about the connections, and we were chatting and talking about what we would do, and the Discord server was buzzing, and everyone had an idea, and then-- and people were saying, "Hey, welcome," and, "How did you come?"
and, "How did you hear?" because the link isn't public yet. And it was all about connection. And so the code was a side effect. And yes, we produced an insane number of PRs and [00:13:00] commits and issues and all of that, but that was a side effect of people having fun with their friends that they were meeting for the first time online, which was exactly the opposite of what everyone had been telling them was the important thing for a long time.
And I don't know about you, I've been close to quitting open source in the past or maybe even th- quitting tech entirely. And it's always I've recovered from that in that moment by doing something with people that's fun and free and not constrained, and where you have the permission and the ability to do things without people stopping you and constraining you and taking your agency away.
And one of those was, in fact, with actually Matthias and Anthony and Kevin and a few others, which was a client for Mastodon called Elk. But just something like that, it has the power to rescue you from a moment of burnout when you are being forced into this awful mold that your spirit isn't meant to fit into.
Just having a moment [00:14:00] to be free and spread your wings and actually enjoy what you do again makes a big difference. And certainly, it was a lot of fun for me, and I hope for others too. 
[00:14:10] Richard Littauer: I appreciate that immensely. I've had similar experiences before when, you know, when I get tired of reading and I don't really wanna read, and then I find a book that I absolutely love, and it's like, "Oh, yeah, this is what I was doing this for.
This is why I like this." I don't mean to be a doomy and gloomy person. I naturally tend towards those things. 
[00:14:26] Daniel Roe: In 2026- I know, weird ... how could you be thinking about doom and gloom? What, that's very strange. That 
[00:14:32] Michael Capeletto: requires 
[00:14:32] Richard Littauer: so much 
[00:14:32] Michael Capeletto: imagination. 
[00:14:35] Richard Littauer: Have you had to deal with any conflicts with anyone yet in the community, or is everyone still staying in their lane, being very kind all the time and thoughtful all the time, even without sleep?
[00:14:47] Daniel Roe: We've absolutely had conflicts. Matias, were you- 
[00:14:49] Michael Capeletto: I was going to say that in any time that you do something significant, like that people care a lot about, and in a big group of people... We are talking about 700 [00:15:00] people in this Discord, 350 people, like, committing stuff and having ideas, and I think it will be weird or that there isn't any kind of conflict.
But, like, what is important is how we approach together and how we can move. And this is why, like, governance is important, having a code of conduct is important. There was very little banning needed, and actually we are learning a little bit from other experiences. And before the alpha launch, we created another Discord that is working as a lobby, and, like, it's going to...
intended to be, like, a Discord for users long term that is separate from developers that trying to protect a little bit this developer community. It happened in other places where once you reach a moment where you have maybe 100 to 1,000 time more users than the people developing the project, that is normal.
Any kind of, like, successful library that you will [00:16:00] do, like, it will have more users. And at one point, it start to feel like all the public channels are about support, whatever you name them, and, like, people start to retract to more private spaces, and, like, it really changed it. So there were like... I- I think that the normal stuff which we are an inclusive- Community, and we have our values.
Like, the community took in their hand, like doing nodels, for example. That is like our version of like the doodles from Google. So like if you go to npmx in like certain dates, you're going to see like the npmx logo with the colors of different communities and inclusion. And we did like a lot of that, and some people didn't like it, and they don't need to be with us developing or using it, and they can just use whatever they want.
So we had a few talks where we put a point that everybody that shares our values is welcome, but it's not that [00:17:00] everyone is welcome by itself. There is a common base of values that you need to share with us to be involved in the project. 
[00:17:08] Daniel Roe: The thing people always lead with is why does this need to be political?
And people say that, "Why can this not just be a great browser for the npm registry?" Or something like that. And that totally misses the whole point of everything, because it's not about just the software. Isn't that everything that we've said so far? This is about people, and this is about open source, which is about people collaborating and building together.
And the moment you say this is about people, then it's not just that inclusion is a good thing as it would be even if you were just building a product. But actually, it's not just a good thing, it is actually the main point. The point of having diversity, the point of having inclusion, that is what open source is about.
And honestly, if someone thinks that's [00:18:00] politics, I would actually pitch it back and say it's value. Values should drive politics, not the other way around. And one of the most important values that we should care about is inclusion. And yes, that should affect who you vote for, but that's not party political thing.
But it should affect what party you vote for. Sorry. Yeah. But the point is, for an open source project, it matters. Um, it's one of the absolute sine qua non. This is essential for an open source project if you want that open source project to be a good, healthy, safe space. It needs to welcome people. It needs to value their contributions.
It needs to empathize with different users of the code and application. And one of the best ways of achieving that is to have a diverse group of people. 
[00:18:41] Michael Capeletto: Yeah. I think in the case of npmx also, it's, as Daniel said, this is important for any project, but in the case of npmx, this is a browser for the npm registry that is like the biggest packages registry in the world right now.
And this is about finding [00:19:00] what are going to be your dependencies, who you're going to depend on, who is the people that is behind. We are creating a browser to browse the, basically the open source npm ecosystem, and it is important that r- also is reflected in the way that we build it. NPM X is also a conversion of several communities that came together to do this product, and this is also why it was, I think another reason why it was kick up so fast is because like a lot of us knew each other.
As Daniel s- said, like we shared project before. There was a lot of people from the Vita ecosystem, a lot of people from the Nax ecosystem, from the View ecosystem, from like other different communities. One particular community, for example, that we also were part of and was heavily involved is the E18E, that is Ecosystem Performance, a play with the A11Y for short.
And this ecosystem performance community is like a community of [00:20:00] people that systematically goes over the biggest, most used dependencies in the NPM ecosystem and try to make the dependencies leaner, try to improve the performance, try to modernize it. Let's say that this library is using a dependency that is not needed anymore because the node already supports it, you know?
So like they go and they help the maintainers modernize it, and if the maintainer doesn't want because of any reason, they try to help other libraries, alternative libraries, or create new ones to actually modernize because like then you need to still have an option there. And one of the collaborations that we did early on was that if you go to NPM X, now that we have a lot of users using this registry, we can help them in ways that npmregister.com was not doing.
And if you go to any library that has not been maintained very well in the past or like they are better alternative and that library is already [00:21:00] locked in time, then there are modern alternative that maybe it's half the size, maybe it's faster, then we show a warning with a link explaining what are the alternatives.
So instead of just relying on like the popularity signal, that is a trap because like something that is popular, if not will just get more popular even if there are better alternatives, we start to show other signals. And that is like just an example of like collaboration between people. And if you want conflicts, there is a lot of conflicts there directly because you have like the people on the other side of that table saying, "This library, maybe there is a better alternative," that aren't happy with us.
Some of them are 
[00:21:42] Richard Littauer: still, you know, happy. I get to know a lot of people at NPM, and I feel bad for the current NPM web team, you know? 'Cause like this is a really cool project, and they probably are hamstrung by funding or other people above, and I hope that they take what they can, but I hear you on that 
[00:21:55] Michael Capeletto: Even with the current team, they have been, like, doing more features lately.
It [00:22:00] was quite good to see. Like, they added dark theme. They added now is, uh, like, how it's called? Like, the station publishing. They have been doing, like, things. And as Daniel said, I think there is a lot of engineers that really, really want to improve the website, and decisions are coming from other side that are not in the engineers itself.
[00:22:21] Richard Littauer: So there's a lot of threads I wanna pull on right now. 
[00:22:23] Michael Capeletto: Yeah. 
[00:22:24] Richard Littauer: One of them is I really wanna get a sticker that says that all science is just library science, because I really think that this registry, like, finding stuff is the most important thing, and we're seeing this across every ecosystem. So that's why ecosystem.ms by Andrew Nesbitt and Ben Nichols is just amazing, right?
Because it helps you sh- show the scope of things. But it's the same problem we have in the sciences general. It's the same problem in Python, where a lot of the recent podcasts have been Python. I'm kind of curious whether you're thinking about extending npmx to other ecosystems so that you can also search for Go libraries or other package [00:23:00] managers through this incredibly fast website.
Is that something that's come up? 
[00:23:04] Daniel Roe: Interestingly, we've had a couple of clones in terms of the design and look and feel, which is a really nice thing, by the way. When something like that happens, you're like, "Ah, someone else likes it." There was a Windows Chocolatey. Anyway, there's some websites that are basically npmx but for other registries.
I mean, it would be possible. Using something like the API that Ecosystems provides, we could display different packages. I think one of the challenges, though, is what's the purpose somebody has when they go to a website like npmx? So typically they are operating in a single ecosystem. They are in the JavaScript and TypeScript, say, ecosystem.
And so there are other registries in that ecosystem, and I think it would make sense to support them as first class citizens alongside npm. So you have npm and JSR and anything else. That would be a definite benefit, I think, more than supporting Go, for example, so that when you're searching for [00:24:00] library for computing hashes for my project and, you know, you're not seeing Go and like you shouldn't be seeing Go in the list.
In fact, we even wanna make it more specific, and this is one of my personal bugbears from my early days reaching into web development where I'm like, "Okay, I wanna build something. Great. I'm gonna do something with PDFs." And then I search npmjs PDFs. Great, there are like five libraries. Okay, let's try this first one.
Oh, wait, it doesn't seem to work in the browser. It's relying on some Node stuff. You're like, "Ah, Node. Well, I guess npm originally was all about Node. Okay, so that's not gonna work." Well, actually, it's really difficult sometimes to find that the package works in the environment you want to use it in, and that can be anything from the Node version that you're running or the fact that you're not using Node at all, you're using Deno, or the fact that you're actually compiling and you're deploying in Cloudflare Workers, or whatever it is.
You have a set of constraints, and wouldn't it be just great to be able to take those constraints and impose them on the registry so that when you're searching and [00:25:00] looking at stuff, actually you see like, "Hey, this package isn't gonna work for you. Just so you know, these are the ones that will." And so I think that will be a killer.
More useful than going broad. And it's always tempting to go broad We could rule the world. We get all the packages will come to npmx. I'm still up for that one ring to rule them all. But more useful, I think, is to narrow it and actually help people narrow it themselves and help people focus on just a little slice of the ecosystem that will help them.
[00:25:28] Richard Littauer: One of the reasons I stopped being a JavaScript developer and stopped doing web stuff is because I just got so annoyed at constantly having sites not work because of modules that weren't in the environment that I wanted to be, and not being smart enough to figure it out and feeling dumb all the time for dealing with things that were just never going to work.
So I would vastly support that. I want to go back to something earlier where you were talking a bit about politics. It's 7:50 in the morning right now. In two hours we'll hear whether Count Binface has won. Let's go Count Binface. But there's a British election right [00:26:00] now. You can read about it later. It happens in a 
[00:26:01] Daniel Roe: couple hours.
Richard, I just want to say I envy every single one of those people who don't know what's happening right now and as
a result are going to go and discover like one of the best experiences of their life as they uncover the history of Binface. You have just brought joy and just, I realize you're about to tell me how important that- ... but this is one of those wonderful moments in 2026, you know, and people discover Count Binface.
Um- It's 
[00:26:28] Richard Littauer: true. It's also following a long history of British satirical, you know, like, figures. I mean, Boaty McBoatface, et cetera, et cetera. There's something about Britain that brings these people out, which really just makes me very happy. Do Google Binface, and we will have it in the show notes, which you can find at podcast.saintossess.org.
Thank you for pointing out how great that is. It gives me joy. Politics. Money. Yeah, money's great. How are you funding yourself? It, it seems like it's entirely funded right now by [00:27:00] goodwill, sleepless nights, vacation, possibly the Wombats Developer Union, which used to be the acronym for NPM. What are you doing for the long run to make sure that this project is sustainable?
[00:27:10] Michael Capeletto: That's a good one to actually put in the table, and maybe we can discuss how to actually do that together. It's something that for the whole ecosystem of, and everyone doing open source to figure it out, and I think that this is why you have, like, the name of this podcast, and you have been running with this podcast with so many episodes trying to discuss this topic.
But in particular, there are people like Daniel, for example, that's employed at Vercel, and it was very interesting as a project for Nuxt in the same way that Elk was very interesting for Nuxt because it gave us a way to put ourself in the user perspective and, like, really try use the tools that we're building, also from the Vite perspective, for example.
And it gave us something that is very interesting as a result is that if you want to see a modern [00:28:00] Nuxt application, if you want to learn about Nuxt, go check out npmx because the code has been written by it, by, like, people who are, like, the plugin authors, by the Nuxt modules authors, like, all our maintainer friends came and, like, integrated their own stuff into npmx.
So it's really, really interesting to see as a code base, say, as a l- learning resource. So we have some people that have time to do it without needing to get money directly from npmx. But we clearly need more for the long run, and it will be extremely, extremely helpful for the project to get enough money to be able to sustain, like, one or two developers full-time to maintain this thing if it is useful for a lot of people.
And we are in that process trying to get some grants, trying to get some way. Maybe this is important for some particular company that, like, wants to partner, try to really, like, make this happen. The people is there if we will have the money. Like, the people is there to [00:29:00] actually create an even better resource, like, that is better long term.
There is a lot of features that we would like to implement that it is waiting for that particular moment. We have some companies sponsoring for at a level that allow us to rent an Airbnb to do offsites for the core team and, like, go to conferences and, like, if someone wants to do a talk about npmx in a conference, help them get there, and maybe print stickers and, like, T-shirts and, like, these kind of things.
We have companies supporting us. You can go to the About page of npmx and see, like, a bunch of companies that are already giving us money and supporting this vision at the level that allow us to get together, go to conferences, but we will need more. And I think that is a, like, bigger discussion to have how we get this for npmx and for a lot of other projects.
[00:29:57] Daniel Roe: There's so much to say about sustainability in open source, [00:30:00] and I think often people talk about it in the wrong way as though open source is a business, and I don't think open source is a business. But I do think that there is a degree to which if you think of an individual open source project, there is a degree to which it is like a product.
It needs Marketing. In that, it needs people to hear about it if it's going to grow. It needs vision and direction and executive function. It needs people to do the work. It needs community management. You need to talk to people and help them and welcome them and onboard them if they're going to be contributors.
You could take the life cycle of a whole company and plot it onto any open source project that you think about. But most people do not have within themselves, or at least they don't enjoy, doing all of those things. So in some cases, there are these amazing people creating incredible work that nobody ever hears about, and it sits on three stars on GitHub.
And then there are these people who are amazing at hype, and their [00:31:00] libraries are not necessarily that great, but they seem to get all the credit and the sponsorship and there's lots of different things. But one of the things that when we come together we can help with is what if, as a group of people, we can actually start to fulfill more of those different things.
So maybe we can reach out and find funding and pass it on to someone else who would never want to or go through the process of trying to find funding. But they have incredible work and they're prepared to do amazing things for the ecosystem. That's something that you can do when people come together, which you can't necessarily do if each of those people tackles the whole problem, which is company sized, on their own 
[00:31:40] Richard Littauer: I think that's an excellent place to basically stop because that's pretty much sums it up.
Open source is the best. It has room for everyone, and it needs everyone for various projects. That was really, really good. If you could direct new people somewhere, where would you send them? 
[00:31:57] Daniel Roe: I would absolutely send people to the npmx [00:32:00] Discord. Come to builders.npmx.dev and I promise someone will say you're welcome.
And if you wanna get stuck in, it's a great place to explore and test things out, whatever that looks like for you. You'd be very welcome. And I, I should say the same is true, I hope, in Nuxt, if that's something that you wanna play around with. I have an open diary. You can book a 10-minute chat with me anytime at ro.dev/chat and give me a bell and you can chat for 10 minutes.
It's not much, but if there's anything I can do to help you, I will. 
[00:32:30] Richard Littauer: I would also like to point out that both Matthias and Daniel are advisors for the Software Stewardship Lab. If you haven't heard our previous episode with Vlad Stefan Harbuz, you should definitely check it out, where we talk about a new nonprofit that's dedicated towards research on open source that's just launched in the UK, but globally in terms of its area of research.
And you could also join that Discord as well, where we can chat with other fun people like Matthias and Daniel. Matthias and Daniel, thank you so much [00:33:00] for coming on. Don't leave yet. At the end of every Sustained podcast, we have a section called Spotlight, which I continue to pronounce in a bad Boston accent, where we spotlight other people's projects, things which we think just need a little bit of light shed on them.
Traditionally, the host goes first, which I guess means me. My spotlight today is I have a book called With a Bird: A Reader on Avian Kinship, which was sent to me by Errol. Thank you, Errol. A lot of essays on birds, and it's really quite lovely. But while I was getting this book, which I'm enjoying, I have a brown clock, which I use as my alarm clock.
It's digital. It has a digital face, but there's no syncing. There's no Bluetooth. The buttons are really, really easy. Craig Mod, one of my favorite artists, recommended it. He's a writer in Japan, and it basically keeps my phone out of my room 'cause I can just use this clock to wake up. I wanted such a thing for years.
It's very well made. It's relatively cheap, and I will put the link in the show notes. Highly, highly [00:34:00] suggest. Daniel, what's your spotlight? 
[00:34:02] Daniel Roe: This might be well known to folk, but one of the things to look forward to every day is playing a little game called Clues by Sam. It's at cluesbysam.com. It's a logic game.
It's entirely logic. You can only solve it by deducing the next correct move. It'll give you a little tutorial, and there's one every day, and it is so much fun. I know my day's gonna go well when I sit having a cup of coffee and doing my Clues by Sam, so highly recommend. It's not open source as far as I can tell, but it is a glimmer of light.
[00:34:32] Richard Littauer: I like that. A glimmer Matías, what's yours? 
[00:34:35] Michael Capeletto: I will go, uh, still a little bit in tech a little bit. But given that we didn't discuss that much about AI in this particular episode, in open source right now, that is something that is very present. I wanted to shout out to Asian Scan. That is a tool by Matteo Gabriele that it has been helping us a lot in a lot of projects.
And this is a, [00:35:00] a very good example also of like people getting together and working on... Like Matteo started it, but now like other people are giving feedback and like working on it. And the idea of this tool is that it's automation tooling for detecting if someone that send a PR to the project, it is an Asian or not.
And that is very important to avoid talking to an Asian. That is going to be like a little bit more difficult to detect lately. And there are a lot of people that right now are sending 10, 20 PRs in an hour to different open source project for some reason, maybe like clout or like whatever, and like that is a lot of noise to work with, and it's really making it more difficult to trust that you're talking with another human being, and like to create that connections that we like really talk it through this episode, and to see the noise is hiding the conversation between the humans.
And this tool has been helping us a lot by, yeah, fighting back a [00:36:00] little bit there. So like it's good to check out. 
[00:36:02] Richard Littauer: I love that. I've started writing to people like who email me or cold email me being like, "Can you explain why you used this word?" As a Turing test, 'cause I'm starting to get lots of like AI written spam, including to say, "I would like to come on Sustain."
And I'm just like, "You... I'm not even sure you've listened to an episode or can listen." I will have to look that up. Thank you, Matías. Listeners, I hope you have enjoyed this podcast. If you have, there are several things you can do. One, you should go to npmx.dev and join in, sign up, join the Discord, have a good time.
You can also go to sustainoss.org and join in, sign up, have a good time. We have a discourse. We have a podcast page, which is surprising. I, I know. We have a podcast, podcast@sustainoss.org, where you can find the show notes for this episode. You can also go to our open collective where you can donate, which would be really, really cool.
These episodes cost some money because our editor is so amazing that like I can't exist without him, and I need to pay him, and he is not AI. Paul. Hi, Paul. Love you Besides that, [00:37:00] you can also donate to npmx. I think that's probably an option at this point. If not, it should be soon. You can go ask them about it.
You can share this podcast with your friends and enemies, or just disinterested strangers. That would be really, really cool. Sharing is caring. You can also like it on whatever app you're on, you know, Apple, uh, Spotify, whatever you're using these days. Please just press the heart button. I know it feels stupid, but it does help.
You can share it with one friend. You can send hate mail to richard@sustainoss.org, or podcast@sustainoss.org, which will go to all the other hosts as well, and you can suggest guests. We would always love to have more people. Daniel, Matthias, thank you so much for coming on. I hope you have beautiful evenings, and I really appreciated this podcast.
Continue to make the world a slightly better place with dark mode and faster websites. Good work. 
[00:37:50] Daniel Roe: Likewise. Thanks a