Castopod and the podcast environment – episode 1 – Y. Doghri and B. Bellamy – AD AURES

Interview with the Castopod team

Walid: Welcome to another episode of Projet Libre. Today, I’m really excited because we’re going to be talking about a podcasting platform called Castopod and its development team. I’m delighted, since I migrated to Castopod at the end of last year.

To talk about Castopod, I have with me Yassine Doghri and Benjamin Bellamy whom I’ve had the opportunity to meet, Benjamin on several occasions, and Yassine not so long ago. So listen up, Yassine and Benjamin, welcome to Projet Libre, I hope you’re all fit and well.

Yassine: Great, hello.
Benjamin: Hello, thank you.

Presentation by Benjamin Bellamy and Yassine Doghri

Walid: So, first question, I’m going to ask you to introduce yourself and tell us how you came to be involved in free software. Benjamin, the honor is all yours.

Benjamin: Hello, my name is Benjamin Bellamy, and I’m an engineer. Over the past twenty years, I’ve led projects in a variety of industries, including publishing, book distribution, online media, television and e-commerce.
And I’ve always been a staunch supporter of free software and open source, as well as a podcast enthusiast: so the road was paved, I’d say. This led me to set up Ad Aures four years ago with Yassine, a company dedicated to creating fair and sustainable ecosystems for the podcast industry.

Walid: Yassine, what about you?

Yassine: My name is Yassine Doghri, and I’m a computer engineer. I co-founded, as Benjamin said, Ad Aures with him four years ago and today I work as an architect and main contributor on Castpod, mainly. I did a few other projects for Ad Aurès on the side. There you go.

Walid: How did you discover free software?

Yassine: So, how did I discover free software? It’s pretty classic. As I’m a web developer, I started working on projects and using open source tools a lot. And so, one thing leading to another, little by little, I began to understand the ins and outs of free software. Because it’s not necessarily open source, free. I’m pretty much a values person and that’s how I discovered it. I’ve been lucky enough to work on an open source and free project, so I’m very happy today.

Walid: and you Benjamin?

Benjamin: I started computing in the ’80s on MS-DOS 3, at a time when, frankly, there wasn’t much talk about it, and I started to get interested, I think, when I actually started working. I started working in 98, at the time we were already using the Apache web server, we were already using a lot of Linux and there were starting to be distributions that prided themselves on being user-friendly, in particular Red Hat, which was doing a lot of work.
It was also around this time that we started to get Mandrake and company (Editor’s note: see interview with Gaël Duval). Well, it was still user-friendly for people who knew how to recompile the kernel, but it wasn’t yet as usable as it is today. Because today, anyone can use Linux. At the time, it was a bit tricky, especially when it came to running corporate services. But it worked pretty well, in fact it was quite stable. And it was a bit of a revelation to realize that, at a time when I was still totally formatted by the times, a Lotus Notes server on Red Hat was working really well.

Walid: We’re about the same generation, both of us.

Benjamin: yeah we’re old yeah.

Walid: before I introduce Castopod, I’d like to tell you a funny story: when I started the podcast, I didn’t know Castopod. I discovered Castopod by watching a talk Benjamin gave at FOSDEM. And by the time I was watching the talk, Benjamin had been listening to an episode, he contacted me and said “yes your podcast is cool”. I thought it was very, very funny.” There you go.

Benjamin Bellamy (source:

Benjamin : There’s no coincidence.

Walid : That’s it, and so I said to myself “damn, that’s stupid, I didn’t even look for a free podcasting platform when I created my podcast”. Still, I was a little ashamed.

Benjamin : So no, we’re the ones who should be ashamed. Clearly, we are, I said, we love podcasts and we are ardent defenders of free software but we are not the best at marketing, we’re not going to lie. Sometimes we miss it a little bit. That’s also why we’re very happy that you’re inviting us so that we can talk about what we do and then make as many people as possible discover our solutions and how they can or can’t help them. We realized by searching all the podcasts that talked about free software that there are very few that use Castopod, including those whose creation is after the creation of Castopod.

Free vs. open source

Benjamin : I just wanted to come back quickly because I think it’s not bad to do it as a preamble: you ask us how we discovered free software and you didn’t talk about open source. And I think it’s interesting to make the difference in the sense that globally today between what is open source, free software, free software, freeware, we tend to lump it all together, it’s not super important because very often free software is open source and it’s free, but it doesn’t necessarily go the other way around. And we’re a defender of free software, not necessarily open source.
And I wanted to take this opportunity to reiterate this a little bit because it seems fundamental to me today, especially when you call yourself a defender of free software, to say what it is. Open source software, to start with the simplest, just means that when you distribute software, you distribute it with its source code. So how does it work, what’s inside. That doesn’t mean it’s free at all. Typically, when someone makes a website, they will give the source code which is often going to be HTML and that’s not why it’s free. Since it’s free, it means that I have the right to run a program for whatever use I want.

You can’t tell me, you can use that to do that, but not to do that. I can use it for whatever I want. I can study it to see how it works and modify it if I feel like it. I am free to redistribute it, that is, to modify it and redistribute the changes I have made and thus improve it for the benefit of everyone. So where free, everyone understands what it is, is that you don’t pay. Open source means that you have access to how it works inside, it’s also technical. Free is a bit more philosophical, it’s really about freedom. I have the right to do what I want, what I want within the limits of the user license. But in any case I have the right to use it for whatever I want, to study it, to modify it and to redistribute it. And this is fundamental, especially today when questions of digital sovereignty are at the center of all concerns, where we ask ourselves a lot of questions about “but what are they doing with our data? Where are they going? What are they using it for?”, and to be able to study how it works and what’s behind it, that’s fundamental today. So why is there this conflation?
Because very often free software, well very often I think it’s systematic, we have access to the source code. It’s hard to see how you can make free software without giving access to the source code. And in fact, as soon as you have access to the source code, you tend to say it’s free because I can take it, I can copy it, I can modify it.
So there are still economic models related to free software and I think we’ll talk about them in the context of Castopod later. But in any case, these three notions, which are often confused in fact, are very different and that’s important, because it’s also important to understand why we wanted to make free software. We didn’t want to make free software because we wanted it to be free, we didn’t want to make free software because we wanted people to be able to have access to the source code, even if it goes through that, it’s because we think it’s essential that people are free to publish their podcast with the possibility of modifying, to know how it works and to redistribute if they make changes.

Introducing Castopod

Walid : To go further, you can listen to the episode with Benjamin Jean and also the episode with Raphaël Semeteys and Gonéri Le Bouder on the economic models of free software.
Now I’d like you to introduce me to Castopod. What is Castopod? How did the idea of Castopod come about?

Benjamin : So I’m glad you asked. What is Castopod? So let’s make a little technical preamble about the podcast.
What is podcasting across the Atlantic? In fact, it is a mode of broadcasting multimedia files that are essentially sound. What we call podcasts, we have a technical definition, because we’re engineers, is an MP3 file in an RSS feed: that is to say the possibility of publishing sound programs in MP3 format or other sound format, moreover, via an index file that will list all of my episodes, we talk about shows, episodes, in an XML file which is a computer formalism to list a set of elements. When we said that, we didn’t say much, except that it has a lot of implications. The first is that anyone can today, yesterday or tomorrow say, “I’m going to do a podcast and I’m going to upload my XML files to a server.” And listening apps, we know a lot of them, we know Apple Podcasts which is the most widespread historically, today there is Spotify, there is YouTube which has started to do it in a certain weird way, we may be able to talk about it later, there is AntennaPod, PocketCast, Anytime, PodcastPlayer, Podfriend, Podverse, I’ll mention a few, including open source ones.

Antennapod Logo

And all these programs are able to understand the famous index file, the RSS feed, and download the MP3 files. What is very particular about the podcast and which is very different from the internet that we have seen develop since 2007 is that the podcast is decentralized: that is to say that any listening software can go and get its MP3 files anywhere. So I don’t have a closed silo, I don’t have a closed platform. I’m in an ecosystem that’s totally open, totally interoperable, where anyone can publish a podcast and anyone can play a podcast. As long as we’re talking about the same protocol. This is often compared to SMS, where no matter my phone, my carrier, my sender or my receiver, everyone can exchange messages with each other.

On the podcast, this is still the case. This is not the case, typically, if I’m on YouTube : my YouTube video, I can only have it on YouTube and not anywhere else. I’m in a closed ecosystem. So the podcast stayed open, which means starting to choose, if I’m a podcaster and I want to publish a podcast, a podcast host. So I can choose self-hosted, I can go and get a solution from the market. I had looked at about a thousand podcast hosts out there, which is significant because they all have the same value, which is to say that all of these hosts all have their content that is going to arrive in Apple Podcasts, in Spotify in exactly the same way. There is no one that is going to be more or less visible than another. Among these thousand, there are still a hundred that are bigger than the others. But it’s a must. That is to say, I have to start by finding a host and entrust them with my MP3 files. It’s going to make it an index, an RSS feed, and it’s this RSS feed that I’m going to then entrust to all the platforms of the cases.

Castopod is a podcast hosting platform, so there is an interface on which I will be able to define the name of my podcast, its description, the language I use to express myself, possibly the category, do I talk about religion, news and technologies, is it for children. After that, I’m going to publish episodes. For each episode I’m going to put an MP3 file, I’m going to be able to put an image, a description, a publication date. I’ll be able to say do I publish every week, in series, there’s a whole bunch of parameters. And Castopod will generate this famous RSS feed which means that my episodes will appear on all the platforms where I want to be referenced. It’s me as a podcaster who decides where I’m listenable, where people can listen to me. If I don’t feel like being on Apple Podcast, I may very well not be there. If I want, I can be there and so I will choose the platforms on which I want to be heard.

Why did we develop Castpod? Because in fact, we were already working in podcasting and for a whole bunch of reasons, we were doing tests and we needed to do tests on a number of podcasts. And so, we were looking for an open source podcast hosting solution because we didn’t want, when we were testing 12 podcasts, to pay for 12 subscriptions with a professional hosting provider that charges podcast per podcast. We made a little list of prerequisites, things we needed. And then we said to ourselves, we need an open source solution because we want to be able to do what we want, we want it to have such and such criteria. And then we started to search, to look everywhere on the sites that referenced open source solutions, podcast solutions, and then to choose the solution that met all our prerequisites. And we couldn’t find any. And it was a long way off: that is to say that overall we had about ten points, we found one that ticked 2-3 boxes, but we were very, very far from the mark. So we were a little disappointed of course, we used a solution that still exists called Podcast Generator, because it’s PHP, there’s no database, it’s very simple, the code is very old, it’s not MVC, it’s very old, but hey, we used it as best we could. And then one day, I think it was at the Pas Sage en Seine festival in Ivry-sur-Seine, I met Ludovic Dubost from XWiki, to whom I talked about it and to whom I mentioned my wish list. He said, “Oh, but there’s a company, a company called NLNet who subsidize projects with European funding, European subsidies, as long as it’s open source and in addition they are very much a driving force on everything that is Fediverse / ActivityPub“. It turns out that we wanted a podcast platform on the Fediverse, we’ll talk about it later I think, and he said “well made a file of about ten pages and our project was accepted”. We got a subsidy of about 40,000 and a few euros, which is not nothing. Which is a far cry from the blow that the development of the platform represents today, but it’s still not nothing at all.

NLnet foundation

And above all, it reinforced the idea that there was a need, that there was a place, that it was not normal that there was no open source podcast hosting platform that met the functional and technical criteria that we have the right to expect today. And that if a European body was willing to finance it to the tune of 40,000 euros, well, there really was something to be done. And it wasn’t so much the amount of the subsidy that convinced us, but the fact that there are people waiting for more. And that’s how Castopod was born. So what started out as a wish list has become a side project and is now a full-fledged project to which we devote a lot of time, energy, passion, blood and sweat.
And we’re super happy with the reception it’s received and the fact that people are using it. And as such, I want to thank you doubly for inviting us and using Castopod.

Walid : Yassin, do you want to add anything?

Yassine : You mentioned it for the Fediverse, but I think the differentiating point of Castopod is that it is connected to the Fediverse. That is to say, it is integrated into a social network that is federated. Castopod integrates a federated social network. But here are the other hosts, there are none today that are also integrated into the Fediverse.

The technologies that power Castopod

Walid : What I’d like to talk about now is the technologies you use in Castopod. Technically, what is Castopod made of? Yassine, do you want to talk about it?

Yassine : Castopod is based on a PHP framework called Codeigniter. It’s not very well known, but it’s in the same style as Symfony or Laravel, although it’s a little more capable, a little lighter. So, we based it on that. It allows you to develop web applications quite skillfully, so it’s great to get to grips with.

Logo Framwork PHP CodeIgniter

And as a result, it’s a web application, so PHP, MySQL, HTML, CSS, JavaScript, which is the most classic. Knowing that we modeled ourselves on WordPress, the WordPress model, namely that today 70% of websites are on PHP, and a lot of WordPress today on the internet. And so we wanted to keep this idea a little bit, it’s that today Castopod can be installed on a shared hosting so at a few euros per month and so it allows users who would like to have a little Castopod of their own to do it without problem.

Youngest child : I’d just like to add, when we developed Castopod, we really modeled ourselves on the WordPress model, whether it’s the economic model or the technical model: that is to say that Castopod, it can be downloaded in one click from, and it’s a zip file, and you unzip the zip file at your host, Yassine said it, It can be shared hosting.

And then you follow a wizard, an assistant. You’re going to have to fill in the database info, two or three odds and ends, the name of your instance, etc. And it works. The idea was to remain as easy and least expensive to deploy. Of course, today you can use Docker, Kubernetes, Ansible, Yunohost. There are plenty of ways to deploy Castopod, but there’s always the option to unzip, install and it works.

And that’s without the need for a big virtual machine. If I have PHP MySQL, it’s going to work.

Castopod’s license

Walid : License question. What license are you using and why did you take this license?

Benjamin : So we’re under the AGPL V3 license, we asked ourselves the question before recording this podcast and we thought maybe it was in alphabetical order because AGPL was the first one. More seriously, as we told you, we like free software, free software, to put it mildly.

So we wanted a viral license and then we were also guided by NLNet. Most of the Fediverse software today, whether it’s Mastodon, PeerTube, is licensed under AGPL. In fact, we didn’t think about it for very long, that is to say, we didn’t really think about our brains

Benjamin Bellamy

Basically, anyway, today, when you have server software, it’s either GPL or MIT. We were leaning towards GPL and then NLNet reinforced this idea a bit. It was a natural choice. Maybe we didn’t ask ourselves enough questions, but today we’re pretty happy with this license. I think it’s in line with our values.

Walid : I didn’t know that NLnet was funding projects from the very beginning. It inspires me that we really need to do an episode on European funding because there’s really a lot to say.

Benjamin : So NLnet funds projects from the very beginning. It finances projects for companies and individuals. So I don’t need to have a self-employed SIRET to get NLnet financing. You have to know that if I’m on my own developing software, I can go and see them. Applications must be submitted once every two months and, depending on their workload, they respond either very quickly or a little slower.

And importantly, payment is made exclusively upon delivery of the source code.

Introducing Ad Aures

Walid : We talk about it in the episode on Peertube. Okay, we started talking about it at the very beginning now, let’s move on to the presentation of Ad Aures. I’d be interested if you could tell us a little bit about your company, what is it, what is it, what are your activities, what do you do on Ad Aures?

Logo Ad aures

Benjamin : Of course, so as I said earlier, Ad Aures is a company dedicated to creation. We like the podcast a lot because we are a consumer of it, we like the podcast because it is an ecosystem that is decentralized, on which everyone can intervene, which is not the case in other ecosystems. However, it’s complicated to make a living from it. It’s simple, I create my account and then I just have to follow. In the podcast, I have to find a host, if I want to monetize, there are plenty of solutions. Do I want to advertise? Do I want to go to a financing platform? Do I want tipping? Is that… And anything is possible. And it’s not necessarily easy to understand and then it’s more difficult to hold on to success stories. And the idea of Ad Aures was to say today there is this fabulous ecosystem with this richness of content which is incredible because there is also a freedom of tone, format, language that you can find nowhere else because I can change hosting whenever I want and keep my audience. My audience can change their listening software and continue to listen to me. Finally, interoperability really comes into play. But when we said that, we didn’t pay his rent. And creating content, you can create quality content for free because it’s your passion in your spare time, but it means that in my free time, there’s someone who’s going to pay me a salary and rent.

So at the end of the chain, there’s bound to be someone paying. And while we’re at it, there are people who want to monetize their content directly and we think that’s completely normal and even commendable. We need to offer them solutions. And so we, we claim to offer a particular solution, it’s not the only one, there are many others, and once again we think it’s very very good that there are many, which is an advertising monetization solution, which is not based on user profiling, that doesn’t use cookies, so anyway there are almost no cookies in the podcast ecosystem, because people use dedicated software, which is numerous and multiple, but which will be based on semantic indexing. So in fact, among the tools that we have developed and are working on, there is a semantic indexing engine that allows us to study content and understand what the concepts are evoked. I’m really talking about concept, not keywords. To do indexing and recommendation. That is to say, by calculating semantic distance, we are able to say that one content is close to another because the subjects are related, because they are close, there are many possible reasons. So obviously, it uses a lot of artificial intelligence.

And we can use this either to do discoverability, which is one of the major issues in the podcast ecosystem, or to advertise, i.e. to recommend promotional items, advertising products that are directly related to the content I’m listening to. And we have a lot of testing operations with different degrees of completeness on this, which allow us to advertise. We work a lot in affiliate mode because we find that it works really well, especially on the podcast which is a highly cultural product and then it also allows it to be the content that goes to the advertising and not the advertising that goes to the content, which is kind of the usual mode of operation that we have on the internet today. I’m not going to go into too much detail, but today we’re developing tools for podcasters that allow them to respond to all the issues that may arise, including modernization, which seems to us to be the sinews of war, somehow. And of course, all this works with Castopod, which is the hosting platform, which is the first link in the chain.

Walid : To continue, it’s about understanding both your positioning in relation to others, so that’s what you’ve just given a little bit, that is to say how do you differentiate yourself from others, what do you bring that is different, that makes people want to work with you. And then afterwards, what I’d like us to talk about afterwards is a little bit about your contribution to the podcasting industry. What do you carry as a value, as an expectation, etc.? What do you highlight?

Benjamin : This goes back to what we said in the introduction, which is to say that we are really convinced that podcasting must remain free and must remain decentralized. We really don’t want the podcast to close in a closed silo. And today, there is a real risk. And for that, the angle we took is a functional one, which means that we try to provide functional answers to the problems of content creators who operate in a decentralized world, in an open world. Where today YouTube has opened its platform to podcasting, that is, you can go to YouTube Creator and then enter the address of your RSS feed and your podcast is going to be injected into YouTube and behind you will have access to all the features of YouTube.

I have no problem with that, I think it’s fine as long as YouTube doesn’t have a monopoly on podcasting. And we know that if there’s no alternative to monetization, if there’s no alternative to how I’m going to be able to improve my discoverability, that’s inevitably what’s going to happen. So for us, it’s paramount to provide alternative functional answers to YouTube. When you enter your podcast into YouTube, there’s information that comes in, there’s nothing that comes out of YouTube. If you create your podcast from YouTube, it won’t be accessible to other podcasts. YouTube is a black hole actually.

There are things that go in, but nothing comes out of them. They suck up the value, but they don’t reshare it. They don’t reshare information, they don’t reshare content. YouTube as a company by the way doesn’t create any content, they decide what you’re allowed to say, what you’re not allowed to say, whether you’re allowed to say swear words or not, literally, but they don’t create anything, they’re just there as a referee and as a functional provider of a number of tools. Again, I have no problem with that as long as they don’t have a monopoly and it leaves a choice to be able to express themselves because if I want to say something that doesn’t suit them, that contravenes their general condition of use, that I can say it elsewhere because YouTube is not the law, T&Cs are not the law, the law is something else. And so for us it is essential to maintain a decentralized ecosystem where there will be free competition in the capitalistic sense of the term, that is to say that different companies, different players will be able to propose solutions, but they still have to offer solutions. So that’s really what we want to do, that’s why we offer monetization solutions, because we know very well that we can offer the best platform on the planet if the prerequisite is to have to work for free, it’s not going to work for everyone. If the features aren’t great, it’s not going to work for everyone. And that’s where you put a lot of effort and invest a lot of time and energy.

I think it’s a good time to talk about Podcast 2.0, or at least to introduce it, because what we need to be aware of when we talk about podcasts today is that for about fifteen years, the entire podcast ecosystem has relied entirely on Apple.

Apple literally held the podcast at arm’s length for 15 years, which is to say that they had pretty much the only podcast index, the only podcast directory, so the set of listening apps depended on Apple to have the list of podcasts that exist.

Benjamin Bellamy

They were very much in the majority: which is quite absurd in fact, having a totally free, open and decentralized ecosystem, on which we have a player that has a market share of much more than half. But they didn’t get any financial benefit from it, and we have to be grateful to them for that. They kept the thing afloat, which is to say, they continued to keep the index alive, to give API keys to people who asked for them. So you had to ask for permission, but hey, in any case they did it and they gave it. They were the ones who decided who had the right to read and write on the index, but they did it for free.
The problem is that for 15 years, when you wanted to make a functional evolution of the podcast, you had to ask Apple, since they were the ones who had the index, who were at the center. And that if I want to add a feature, I have to be able to put it in the RSS feed, and so it has to be referenced in Apple’s central index. And if Apple doesn’t want to, I could put whatever I want, no one will be able to see it and it won’t do any good. And Apple refused any functional evolution on the podcast for 15 years. When I say all, it’s not almost all, it’s all. There have been zero functional evolutions on the podcast for 15 years. Which is quite incredible because we’re talking about a period from 2000 to 2015, a little bit later. In this period, a lot has happened and even on the internet and the podcast has had no improvement, no functional evolution. You take a podcast from 2000, a podcast from 2018, it’s exactly the same. To keep the content of the RSS feed there have been no changes, there has not even been a new category. There are people who have said it would be nice to have a category for ecology. Apple said NO, we’re not touching anything. So today, if you’re doing an ecology podcast, you don’t have a category, which is a bit of a bummer because there’s a bunch of them today.
And so what happened is that in 2019, if I’m not talking nonsense, Adam Curry, who is one of the pioneers of podcasting, both technically and in terms of content (he is one of the first podcasters) got fed up and he decided to create Podcast 2.0, i.e. a specification that allows the podcast to evolve, backed by, which is a podcast index to say “well, we shouldn’t depend on Apple anymore, we need to have an index that is open source, open data, open to everyone without needing to ask for permission and in which we can add information to be able to add new features”.

Logo Podcast Index
Podcast Index (source: github)

When I talk about features, it can be very simple stuff, it can be very complicated stuff, it can be transcription, chaptering, links, speakers’ names, geolocation, well it can be anything you can imagine but what’s important is that Podcast 2.0 is open to everyone and anyone can say “Me, I’d like that feature. I wish we could add that. »

What’s really amazing is that when we started developing Castopod, it was just before the release of Podcast 2.0 and Podcast Index. So we had a project to develop a podcast platform that would be linked to ActivityPub, knowing that it was not easy because we were going to create our hosting platform, people were going to be able to connect to it via ActivityPub, we are thinking of Mastodon etc. which had already existed since then, and had already been popular for two years. But for the rest, for the others, for podcast listening apps, we had no legitimacy, no seniority, a voice that doesn’t carry weight. There was little chance that people would buy in. And Podcast 2.0 was a great springboard for us, a kind of megaphone, because one of the first things we did, we said it would be nice if we could add in the RSS feed the possibility to link to the conversation, especially if it takes place on the Fediverse, of ActivityPub, so that listening apps know where this conversation is taking place. And it was integrated and today we have listening apps, from people we don’t know at all, but who have implemented this feature that we had requested.

Photo Adam Curry
Adam Curry (source: wikipedia)

Walid : When you started, how did you get in touch with Adam Curry? Which platform do you use? How did it actually go to insert yourself into this nascent ecosystem? I would like you to explain that a little bit.

Yassine : Actually, what happened was that the chance we had was that we arrived, we started to develop Castpod when Podcast 2.0 and Podcast Index arrived. So there was really an alignment of the planets that happened at the right time. We started developing Castpod in May 2020, well back in 2020. It was really…

The presentation of Podcast 2.0

Walid : Wait, I’m going to ask a question about this because it’s a transition, given this temporality, was one of Castopod’s vocations also to be a kind of state of the art of what was going to happen in Podcasting 2.0?

Benjamin : In hindsight, yes. I don’t know if we realized it at the time. We assumed that Podcast 2.0 was great. We couldn’t have asked for anything better, we would have wanted to do it, but we didn’t have the legitimacy of an Adam Curry, clearly. And what’s really great is that in less than six months, we ended up with about forty platforms, hosts, listening apps, that were following Podcast 2.0. In fact, before Podcast 2.0, there was also a chicken-and-egg problem. If I added a feature as a host and no listening app integrated it, well it’s useless. And on the other hand, if as a listening application I integrate a feature but no one provides the information on the hosting side, it’s useless. And all of a sudden, everyone looked at each other a bit like an earthenware dog and no one made the first move, no one agreed. Obviously, there are a lot of features that were in the air, people were asking for them. I’m thinking of Giovanni from Podcloud who had already written technical specifications saying we should do this, we should do that. It can’t work on its own. The stroke of genius of Podcast 2.0 is to have a voice that is audible to everyone, that makes everyone agree and that says well, here we are, we’re all going to do this.

Walid : He’s independent of everyone else after all?

Benjamin : Actually, there are a lot of things. There’s the fact that Adam Curry was already famous, he had legitimacy as a podcaster, he had developed the first podcast index in the 2000s, there’s a lot of stuff. And then the fact that he comes in and says “Apple-style governance where everyone asks a person and that person just says no to every request, it’s not very encouraging.” That is to say, at some point people end up getting bored, of course. So the idea was also to say “we’re going to do a community and participatory project” and so to answer your question, how do we ask for the addition of a feature in Podcast 2.0? In fact, it’s done on GitHub, it’s done in an ultra-classic way as you would ask for a new feature on any project. And then, there are some kind of sprints, I don’t know if the word is the most appropriate, but in any case, it looks like a sprint that takes place every month, every two months, every six months, in fact, it depends, where you validate. There are back and forths, there are people who say, “Ah, but it would be better to add this thing. This feature has to be at the podcast level, at the episode level, at the level of both, it has to be a character string, an integer.” It can be very technical, it can be very functional. Sometimes there are a lot of people who respond, who feel involved, sometimes no one cares, it really depends on what you have to offer. There are no-brainers, typically, having transcription in the podcast to allow for better accessibility, better SEO, better discoverability. It seems obvious to me that it has to be done and so it was done very, very quickly. And today, it’s only Podcast 2.0 platforms that offer that. Unbelievable, that is, today you’re listening to a podcast on Apple Podcast, you’re sure, there won’t be any subtitles, there’s going to be no indexing of the podcast content. It’s impossible because Apple today doesn’t have a built-in Podcast 2.0 feature, which makes me as a listener say, “Don’t use Apple Podcast. It’s just the most backward platform in the entire podcast ecosystem today.”

Walid : I’m going to give an example because I discovered when I migrated to Castopod, I discovered a feature called the management of stakeholders, the “people”. There’s always the issue of, OK, I’m the host of the podcast, there’s the speakers, but possibly, I’d like to thank other people, actually, or quote other people who put me in touch, who helped share the episode, etc. And in fact, I discovered that you, you implement speaker management and therefore on a speaker, I can put a link to LinkedIn or etc. And that in fact we also find ourselves facing the implementation in client applications where typically, I don’t know me, Podcast Addict will implement this part of Podcasting 2.0 and so you can see, and it’s cool the speakers, but that if I use AntennaPod which is my usual player, he doesn’t see it. As a creator, you have great tools at your disposal, but they are not always available to your user who is going to use. It’s kind of dumb.

Benjamin : Yes, absolutely. So actually, what you have to see is that features are like everything, but it’s not unique to podcasts.

Not everyone uses all the features, not all software implements them. And that’s completely normal. And then even, and this is deliberate, there are Podcasting 2.0 features, we are pretty sure at the time they are released that they will not be unanimous. But that’s by design, which is to say that we can’t predict the success of one feature over another, because at the end of the day, adoption is the listeners who decide, it’s the podcasters who decide, and it’s certainly not the people who are around the table in the podcast 2.0 community when we say “oh well, It would be nice to do that.” After that, the functional spectrum of applications changes from one to the other. That’s also why, as a podcaster, you have a responsibility to tell your listeners to use a certain app instead, because there, you’ll have more information than anywhere else. And there are features to manage donations, funding, money, there are features for… Maybe there’s a lot of stuff.

There are some who will manage the transcript by displaying it in subtitle mode, there are some who will manage it by displaying it in scrollable text mode. After that, everyone has their own needs and prerequisites. That’s what’s great about podcasting, is that there’s no one right way to do it, there are plenty of ways to do things that are more or less adapted to the content and to the listener.

Walid : Yassine, I wanted to ask a question: have you contributed to these features, these new podcasting 2.0 features and if so, on which you have contributed?

Yassine : So yes, especially at the beginning we had a lot of feedback to do on the different specs that were implemented, the different RSS tags that were implemented.
Each time, we have discussions with Benjamin to find out what was best for us, for Castopod, and then by becoming part of the podcast ecosystem.

Benjamin : Actually, I was the one who intervened directly on Podcasting 2.0, at least on the requests. So I was saying, there’s everything that is “social interact”, that is, the possibility as a listener to be able to interact with the podcaster from the listening app. We also worked a lot on the recommendation.
So that’s a work in progress. We regularly give our opinions and in fact all this is available either on GitHub or on Mastodon. In general, what happens is that discussions start on Mastodon, someone says “oh well, I’d like to do that” and then depending on the echo that there is, it’s going to be… Either people say “yes, that’s a really good idea” or “no, I don’t understand” or “that’s a bad idea”.

And depending on the answers, well, it will give rise to a new request on GitHub. And in the end, it gives rise to a specification in the DTD, the namespace of Podcasting 2.0. We didn’t say it, but at the end of the day, it’s nothing more or less than a DTD on a namespace.

Walid : And these specifications, are they supported by an organization like the W3C or is there a standard or is it just a community project?

Benjamin: Today it’s a purely community project that is governed by the rules of the community. There isn’t one person who says yes, no. It is the whole community that decides what can exist and what will exist.

Yassine : So it’s Adam and Dave (Editor’s note: Dave Jones) who are at the origin of the Podcast Index project and it’s Dave who takes responsibility for managing the different sprints etc to include according to what the community says should be prioritized or not, etc. So he’s the one who manages the project schedule.

Benjamin : He has a secretarial role more than a project manager’s role. He’s there because someone has to move the machine forward and then he says, well, here we are, we’re going to close for this sprint, we’re going to close on such and such a date. Behind the scenes, he is not the one who takes responsibility for what goes in or what doesn’t. He is an animator.

Walid : Last question on the subject of Postcasting 2.0. This is an issue that we have already discussed together. We were talking about chaptering and you’d say to me, “Oh yes, chaptering, basically, we implemented it as soon as it came out, we didn’t go back on it too much, etc.” But it’s something we implemented as soon as it came out. So my question is, when do you implement the new features, do you implement all the new features of Podcasting 2.0? How do you make a choice about what you want to implement or not? It’s based on what?

Benjamin : There’s no simple answer to that. It really depends. It’s not that there isn’t a simple answer, it’s that there isn’t a one-size-fits-all answer. It really depends. Of course, a feature that everyone asks us for is likely to be implemented faster than something that no one cares about. And usually, that’s how it goes. It can be a personal choice because we think it’s good or because sometimes there are things that make us love it.
There are some features that we find more interesting than others. I am thinking in particular of geolocation. Personally, I thought it was great to be able, in the 2000s, when we’re all used to using OpenStreetMap or Google Maps, to be able to search for podcasts by geographical area, to say such a podcast is about such and such a place. There are a lot of tourist podcasts for example, and not being able to do a geolocated search on a piece of content, I thought it was completely crazy.
As soon as the spec came out, and it wasn’t very complicated, we implemented that. That is to say, the possibility to say such and such an episode is about such and such a place, or such and such a podcast is about such and such a place, and behind it on Castopod, we put a map where we will have pinpoints (Editor’s note: dots on a map) to be able to visualize which episode is about which geographical area. And that, in my memory, anyway, it’s a trick, I thought it was so cool to be able to do it. And then behind it, by the way, as soon as it was implemented, there was someone who said to me “Oh well, we’re doing a podcast festival in Amsterdam and we have this problem and we’re interested, wouldn’t you like to come and talk about it? And so I went to talk to the festival, the podcast in Amsterdam, about this feature.

So there was a real demand as well. So in general, but it works a little bit for everything, it’s not specific to our organization, to our way of working, we’re talking about open source projects in addition, what is most in demand is the most likely to see the light of day quickly. And anyway, in the end, you have to have a price that Castopod open source if there is a podcast 2.0 feature that is not implemented and there is some, we don’t implement everything today, anyone can add it, the code is open so anyone can add what would be missing and then there are things too or because Castopod evolves a lot from one version on On the other hand, there are changes that are made, whether they are invisible changes or visible changes, whether in the administration part or in the public part, there are many, many changes that take place and that’s the same, there are people who contribute and who will modify this or that. All goodwill is welcome.

Walid : We’ll talk about the community part later. But Yassine?

Yassine : I can add when will we implement the new podcasting 2.0 features?

When we started Castopod, Podcast Index came at the same time, Podcasting 2.0 as well. So the first beacons, they were still very interesting and there was a lot to do. So we added them as soon as they came, we released them on Castopod, we were sometimes even the first to do it.

Yassine Doghri

And then, as Castopod matures, we have to make choices about more stability, make choices of… even features that there are on Podcasting 2.0. Some of them we haven’t implemented directly, and now we’re relying a little more on the community to make choices in terms of priority, as Benjamin said. There you go. The interesting features that we thought were interesting at the beginning, we have implemented them all.

This episode was recorded on January 23, 2024.

To go further


This podcast is published under the double license Art Libre 1.3 or later – CC BY-SA 2.0 or later.

Leave a Comment