WEBVTT 00:00.000 --> 00:12.960 Absolutely. Hey everyone. Our last talk of the night is going to be with the fantastic 00:12.960 --> 00:20.520 John O'Neill. Who's going to be telling us about how ghost has federated and about networks 00:20.520 --> 00:25.160 generalism in, oh, like I would really choke on that one. The work during all this 00:25.160 --> 00:29.280 on it, works in general. So, good point. 00:34.280 --> 00:36.280 Can we hear me? 00:36.280 --> 00:42.120 Okay, good. Fantastic. Thank you being here. Thank you for staying here. I'm well aware that I'm the 00:42.120 --> 00:48.960 one thing standing between you and food. I'm potentially a cold bear. So, right before we get 00:48.960 --> 00:52.080 into it, I'm going to tell you what. I'm going to get into it. And then if you want to make a 00:52.080 --> 00:56.560 evaluated decision about whether you want to stick with this, then that can be up to you. I'm going 00:56.560 --> 01:01.680 to tell you about retrofitting activity pub to an existing product. So, perhaps a little different 01:01.680 --> 01:05.280 to many of the activity pub native projects you've been hearing about for the rest of the day. 01:06.160 --> 01:10.960 How we've been thinking about the Fed of Us at Ghost and our approach to it. I'm going to share one 01:10.960 --> 01:17.120 strongly held belief with you as well as one key lesson that we've learned along the way. And those 01:17.200 --> 01:23.200 are of course cliffhangers that you'll have to stick around like and subscribe and find out. 01:23.200 --> 01:29.440 There will also be occasionally AI-generated images of pugs. We like to refer to ourselves as 01:29.440 --> 01:35.680 the activity pugs, like to be public to any pugs. If you do not like AI-generated images and you're 01:35.680 --> 01:39.840 morally opposed to them because you wish computer progress would stop when everything would just stay 01:39.840 --> 01:45.200 the same. You are in the wrong place. The Java Track moved to room 218 and you should have 01:45.680 --> 01:51.440 it over on that direction. Otherwise, I will be bringing you AI-generated images of dogs that you 01:51.440 --> 01:58.400 will enjoy and with that we will get started. So, by way of introduction, my name is John O'Neill 01:58.400 --> 02:05.440 and I'm the co-founder and CEO of Ghost.org. Just for my benefit, how many of you have heard of 02:05.440 --> 02:11.680 Ghost in any way? Great, fantastic. And then because it didn't come up the rest of the day, 02:11.680 --> 02:15.920 I'm curious. I'm assuming we're in a dev room. I'm assuming a lot of you are developers. 02:15.920 --> 02:23.760 How many people are not developers? Fantastic. And how many of you have a Fediverse handle 02:23.760 --> 02:30.080 are actively using activity pub? Amazing. And how many of you are working on an activity pub 02:30.080 --> 02:36.240 enabled app? You're a developing one for yourselves. Great. Wow. Okay, really good mix. So, 02:36.240 --> 02:38.880 there's going to be something for all those groups. There's going to be something in here for you. 02:39.840 --> 02:45.280 I am not a developer. I'm a very much a designer, a product person. I do develop on the front end, 02:45.280 --> 02:49.440 but for all the heavy lifting, my co-founder and CEO Hannah is really the brains of the 02:49.440 --> 02:54.880 operation on that front. So, for those amongst you who do not have not come across Ghost before, 02:54.880 --> 03:00.480 it is a content management system. It's completely open source. What makes it different or unique 03:00.480 --> 03:06.720 is that it has built-in support for user signups, memberships, email newsletters and subscriptions. 03:06.720 --> 03:11.840 So, you could have gated access to paid content. You could deliver your blog posts as newsletters 03:11.840 --> 03:16.320 and you can really build an audience around your content to allow you to turn that into a business. 03:17.120 --> 03:24.960 We've been around for about 13 years. So, not a new product by any means. And over that time, 03:24.960 --> 03:30.160 from the beginning, we developed a business model because we wanted an open source to be sustainable. 03:30.160 --> 03:35.600 So, we started with a Kickstarter campaign that gave us some initial get-going money. 03:35.600 --> 03:39.600 We use that to hire some good developers that help us build open source software. 03:39.600 --> 03:44.160 Hopefully, that's popular enough to attract users who want to use it. And for those who 03:44.160 --> 03:48.800 have them who don't want to worry about servers, don't want to deal with upgrades or patches, 03:49.840 --> 03:55.200 we offer a paid hosting service. That hosting service funds the original team who continues 03:55.200 --> 03:58.800 to hire more great developers and that's how we keep this whole operation going. 03:58.800 --> 04:04.640 Call it sustainable open source. What's unique about Ghost in particular is that it's incorporated 04:04.640 --> 04:08.640 as a not-for-profit organization, but getting grant funding seemed really hard, 04:08.640 --> 04:12.400 so coming on with the business model seemed a bit easier and so far that has worked. 04:12.400 --> 04:17.440 So, through just the hosting side of the business alone, we are able to self fund 04:17.440 --> 04:22.000 about $7.5 million a year to award continuing to build open source software, 04:22.000 --> 04:27.760 which is released completely free under an MIT license for everyone. And that model has been 04:27.760 --> 04:32.320 very, very successful for us and continues to allow the kind of indefinite funding of the project. 04:33.040 --> 04:40.880 So, who's using it? What is Ghost useful? Well, there's a lot of news organizations, 04:40.880 --> 04:47.200 perhaps unsurprisingly, who care about publishing online doing so in a modern way. So, we have a 04:47.200 --> 04:53.280 bunch of them, both local national and handful of international. We have most startup and tech 04:53.280 --> 04:59.280 blogs that you've ever heard of. A lot of them are up here, a handful that are not Apple, 04:59.280 --> 05:07.840 OpenAI and several extremely popular and surprisingly creative porn sites. And beyond that, 05:07.840 --> 05:12.080 you have the creator economy, which is obviously booming people on the internet who are making 05:12.080 --> 05:16.240 content, people want to subscribe to, that are popular enough that it has become a full-time 05:16.240 --> 05:22.640 source of income, it is their business. So, that can range from podcasts, to YouTubers, to artists, 05:22.640 --> 05:29.200 to everyone in between, a large and growing number of them. So, what these demographics and group 05:29.200 --> 05:33.920 are people all having common is that they publish professionally for a specific audience. They are 05:33.920 --> 05:38.080 not having a hobby blog, they are trying to put something out there and they're trying to build 05:38.080 --> 05:44.160 and grow an audience around it so that they can turn it into a business or make that business 05:44.160 --> 05:50.000 that they are already running sustainable and something that they can keep doing. So, if you're 05:50.000 --> 05:55.520 trying to have an audience and grow an audience, what do you do? Well, you probably deal with one 05:55.520 --> 06:01.680 of the eight horsemen of the apocalypse in the form of a social network. And social networks have 06:01.680 --> 06:08.640 been very effective over the last decade and a bit for one thing primarily. They've been 06:08.640 --> 06:13.840 incredibly effective for marketing, which is to say putting your content in front of a new audience. 06:14.560 --> 06:19.680 What people think they're also effective for is distribution, which is getting your content 06:19.680 --> 06:24.560 to your existing followers or subscribers. But as we now know, they are not particularly 06:24.640 --> 06:30.160 effective for that because no matter how often you like, subscribe, ring the bell, turn around 06:30.160 --> 06:35.360 three times and click your fingers really quick, often the people you're following's content 06:35.360 --> 06:40.160 doesn't actually sharpen your feet. So, social networks have been good for some things, 06:40.160 --> 06:44.720 but not as good for others. And the smart publishers in the room, the smart creators, the smart 06:44.720 --> 06:50.720 new media professionals, have figured this out and so they use social networks for their marketing 06:50.720 --> 06:56.640 prowess to try and drive traffic back to a property that they own, usually in this case, 06:56.640 --> 07:01.200 an independent website on the internet. And that's where ghost comes in. That's what we do quite 07:01.200 --> 07:07.200 well. So you can build a website, you can capture using sign-up forms, email addresses, 07:07.200 --> 07:12.400 and then you can use those email addresses to deliver your content by email, which no one can 07:12.400 --> 07:17.920 put an algorithm in the way of, so that 100% of your audience, spam filters aside, 07:18.800 --> 07:24.720 actually do see what they've signed up to receive. So you solve your marketing and you solve 07:24.720 --> 07:32.960 your distribution, fantastic, right? Kind of, but what's missing is you still have to use 07:32.960 --> 07:38.000 the social networks. If you don't use the social networks, you just have your independent website 07:38.000 --> 07:43.200 and you have your email addresses, it works, your distribution is kind of perfect, but your 07:43.200 --> 07:49.520 marketing is non-existent. It's quite hard to grow because email inherently does not bring in 07:49.520 --> 07:54.560 new people. You might get the odd forward or something, but it doesn't really have any mechanisms 07:54.560 --> 08:00.080 by which email spreads. The email chains have you seen this funny thing. There's not been 08:00.080 --> 08:08.160 around since the late 90s, early 2000s, RIP. So we wanted to have some kind of network component 08:08.160 --> 08:14.240 in Ghost for many years because all of our rowdy, well-funded competitors have them. I'm going 08:14.240 --> 08:21.600 back to Tumblr at one point. It was medium. More recently, it's Substack. There's a very obvious 08:21.600 --> 08:26.800 sense that networks add value to a publishing products, and that's not just because it helps you 08:26.800 --> 08:31.760 get exposure to a new audience. If that's what you're trying to grow one, but it also gives you 08:31.760 --> 08:35.520 a direct connection to your peers to other people who are doing the same thing that you do, other 08:35.520 --> 08:40.960 people who you respect and admire, or perhaps just want to be associated with, you know, 08:40.960 --> 08:46.000 collaborations on YouTube. I like one of the big things and meeting people who are doing the same 08:46.000 --> 08:52.320 thing as you, is very motivating. When you have an independent website, independent blog on the 08:52.320 --> 08:59.600 web, a lot of the time it's not lonely. You put stuff out there and you wonder, anyone reading 08:59.600 --> 09:04.800 it? It's the great downside of especially kind of self-hosted stuff where there's just not 09:04.800 --> 09:10.880 that fundamental connection to other people compared to one of these networks that are connected. 09:11.440 --> 09:17.120 And so the great sadness is while you are running an independent website and publishing it out 09:17.120 --> 09:23.120 into the open web, you have the sense that you're alone and yet you are surrounded by a far larger 09:23.120 --> 09:29.600 number of people and sites and instances than any of these networks, you just don't know it. 09:30.240 --> 09:35.120 And of course, that's the great promise of activity pub to enrapp and enshrine 09:35.120 --> 09:41.520 our socialist ideals and envelop this capitalist society and something that overwhelms it and fundamentally 09:41.520 --> 09:51.200 brings it down. I don't think that was quite just a strong reaction. And that's what we're here 09:51.200 --> 09:56.480 to do, right? Like Wikipedia surrounding the fucking cyclopedia Britannica and just going, 09:57.440 --> 10:04.800 this is better. That's what we're all hoping for and that is the promise and the ideal 10:05.920 --> 10:12.160 of the Fediverse. That's what we want to do. That's what's so exciting about this is the potential 10:12.160 --> 10:19.760 scale and the potential change in the behavior of how we all use the web. But in thinking about 10:19.760 --> 10:25.440 how we're going to enact this, how we're going to make this come true, we spend quite a long time 10:26.160 --> 10:30.160 trying to figure out what's so special about networks, why on networks so different, 10:30.160 --> 10:34.320 because networks aren't the first form of distribution, they're just the latest kind of mainstream 10:34.320 --> 10:40.880 form of distribution of marketing. So what makes them different to what came before? We've had RSS, 10:40.880 --> 10:46.880 we've had email, these are not new things. But both of those technologies very much the broadcast 10:46.960 --> 10:53.680 model. You have a clear delineation between the publisher and the reader. Sky News tells you 10:53.680 --> 11:00.560 what's important, you the reader get to enjoy what's important. We're informed, you need to be informed. 11:02.560 --> 11:08.560 Social network fundamentally changes that. Everyone is the publisher, everyone is the reader, 11:08.560 --> 11:13.280 every reader is a publisher, every publisher is a reader, it all gets mixed up in this wonderful 11:13.280 --> 11:18.400 weird into mingles thing where the most important thing of the day is not what one organization 11:18.400 --> 11:23.840 decided to put above the fold on the homepage and print off on a piece of dead tree. It is 11:23.840 --> 11:29.760 an emergent property of whatever everyone happens to be talking about or everyone has collectively 11:29.760 --> 11:36.880 decided is important. It's like going in nature from a structured direction, decided by one 11:36.880 --> 11:45.760 goose to no one's really sure what's happening but my god is it beautiful. And this is really 11:45.760 --> 11:51.920 important because the dynamics of the interactions change and the biggest change is that we go 11:51.920 --> 11:57.360 from one way communication to two air communication. Two air communication between every node 11:57.360 --> 12:03.120 is what makes it in effect a network and stories spread when people discuss them. 12:03.920 --> 12:12.960 So my strongly held belief is that you can't be a part of the social web without being social. 12:13.920 --> 12:18.960 And that's really important because a lot of existing products that I've seen think about 12:18.960 --> 12:23.840 at activity pub support or maybe just fed of a support in general via any of the kind of open 12:23.840 --> 12:31.440 protocols. They have this idea that, oh, all we have to do is just add activity pub and then we just 12:31.440 --> 12:37.520 keep on fucking broadcasting out. And it doesn't make any sense because you bend up being this guy. 12:38.960 --> 12:42.720 When social networking started, we all said it's like a cocktail party. Remember to whatever you 12:42.720 --> 12:46.000 trying to explain to your friends, it's like a party, everyone just shows up in a really good time. 12:46.000 --> 12:52.800 If you are only broadcasting without listening, you're this guy at the party. If you're the 12:52.800 --> 12:58.560 only thing your platform does is send content out but when people reply you're going, I'm not 12:58.560 --> 13:05.760 interested. That's not a social network. So I think to do social networking software or to be a 13:05.760 --> 13:10.320 part of the social web, you have to be social and that means you have to read as well as right. 13:11.600 --> 13:16.960 So that's what we've been working on. Gurs is a publishing platform. We've had long-form 13:16.960 --> 13:23.440 content publishing for the web and for email built in for a long time. And when we started thinking about 13:23.520 --> 13:28.640 activity pub, the first thing we thought about was not how do we ram the content of our 13:28.640 --> 13:34.080 publishers down the throats of the Fedabers to make sure that they get more views. How can we connect 13:34.080 --> 13:38.560 ghosts to the Fedabers that our publishers can see what's already there and can start to become 13:38.560 --> 13:43.840 integrated, become part of that community so that we can have those discussions. We can have those 13:43.840 --> 13:48.960 social connections start to make sense because I think that's how things actually grow. 13:49.040 --> 13:56.320 So we've got what is kind of an amalgamation of an email inbox and an RSS reader and a read it later 13:56.320 --> 14:02.720 app that gives you a window into the Fedabers for long-form content rather than the more popular 14:02.720 --> 14:06.960 apps today which are typically short-form or some of the more creative ones we've seen there and 14:06.960 --> 14:14.720 3D objects and social bookmarking that are incredibly cool. So you can open full length 14:14.800 --> 14:18.880 articles directly from the Fedabers and browse through them with creative kind of 14:18.880 --> 14:22.880 most kind of kind of kind of kind of like experience on this where you have some customization 14:22.880 --> 14:28.160 when you're reading options percentage of completion reading time all those types of things. 14:28.160 --> 14:31.920 But once you get to the end of an article, unlike an RSS reader or an email, 14:33.040 --> 14:38.800 there are the replies, there's the networked comments that you can see coming in from anywhere, 14:38.800 --> 14:42.000 you can engage with, you can respond to, you can like, you can interact with. 14:42.960 --> 14:46.880 And if you see a comment you like for someone you can pull up their profile and see what 14:46.880 --> 14:50.400 this question's all about, which is the other thing that's kind of interesting compared to the 14:50.400 --> 14:55.760 old blog comments, you know, if someone left an interesting comment and spy Dave, who's Dave? 14:56.640 --> 15:01.200 No idea, I can't find out. But when it's all an activity pub, when it's all in the Fedabers 15:01.200 --> 15:06.480 you can, you can go and see who, in this case, New York citizens bases and what he's all about, 15:06.480 --> 15:11.280 what he's been posting, which is really exciting. And one of the things that I didn't expect 15:11.280 --> 15:18.320 when we started working on this is how different the feeling is of hitting publish in a federated 15:18.320 --> 15:24.560 independent website compared to an old one. When you suddenly hit publish on your own website 15:24.560 --> 15:29.520 and then shit starts happening, people start saying nice things, more people start following, 15:29.520 --> 15:33.360 people start liking it, someone has said it looks like it's going to be a great year, 15:33.360 --> 15:37.760 congratulations on your progress. I feel like writing another post now, I've read that. 15:38.000 --> 15:44.240 That wasn't a thing that used to happen before. It makes it has taken what was previously a 15:44.240 --> 15:50.720 product that I logs into to use for myself into something that feels alive when I log into it, 15:50.720 --> 15:55.120 because it's not just me that's making stuff happen within the app that is ghost. 15:55.920 --> 16:01.520 There's other people who are influencing my experience when I use it. And that sounds maybe 16:01.600 --> 16:07.600 obvious or somewhat slight, but it's a very different feeling to what I guess I would now describe 16:07.600 --> 16:13.920 as a traditional CMS historically was. So this sadness goes away. 16:16.320 --> 16:19.920 Another thing we've added, we kind of tried to put everything into this, 16:21.760 --> 16:26.240 things like work and we did this inbox view, and we figured that works really good for 16:26.240 --> 16:30.800 long-form content, works nicely. But as soon as you get the short-form content, sometimes 16:32.240 --> 16:39.520 people will share a message and just an image, or as evidence has in fact very creatively broken 16:39.520 --> 16:47.280 ghosts on many times, just word all like emojis. Those didn't look so good in the inbox format. 16:47.280 --> 16:51.920 So we split off articles into this inbox view and then everything else, 16:51.920 --> 16:56.720 Federalist Related into a feed view, which is a more traditional newsfeed. So that we could support 16:56.720 --> 17:02.800 the content type that ghost is most suited to, which kind of long-form open content, but still 17:02.800 --> 17:07.200 allow you to see this more casual feed of what else is going on in the Federalist if you want to. 17:08.080 --> 17:13.680 And that led us to obviously being able to engage with those pieces of content like and 17:13.680 --> 17:17.360 reply to them. Those things are just notes. So then we thought, okay, well, why don't we just 17:17.360 --> 17:24.960 add support for adding notes? So now we then ghost, which never did this before, you can publish 17:24.960 --> 17:31.120 short-form content as well as long-form content. And that is again, a paradigm shift, which we're 17:31.120 --> 17:35.120 just starting to explore, which is kind of exciting. The next question is, could you get that back 17:35.120 --> 17:43.440 onto your personal website? Good question. So one of the things in talking to publishers that we 17:43.440 --> 17:48.640 have about activity problem general, we notice they struggle with and I've observed people just 17:48.640 --> 17:55.440 more broadly struggle with is that use a name change, paradigm shift from what people are used 17:55.440 --> 18:00.880 to in social networks to activity pubs a whole. It was quite difficult to wrap their heads around 18:00.880 --> 18:04.800 what do I do? What is this extra extension on the end of my thing? Why is it sometimes different? 18:04.800 --> 18:10.960 What are all these TLDs? We've had a somewhat easier time talking to them because our publishers 18:11.040 --> 18:16.240 are already from some nights because they understand fundamentally that email is the only thing that 18:16.240 --> 18:23.680 has been bigger than Facebook. And in activity pub is one of those, right? When we tell the story 18:23.680 --> 18:27.840 in this way is people, activity pubs like email and emails, the only thing that has been 18:27.840 --> 18:33.680 bigger than Facebook, then they start to get the wheels turning and that's something that I think 18:33.680 --> 18:39.280 we are still in the very, very early days of people getting. This is the first couple of years, 18:39.280 --> 18:46.720 it's getting easier and easier from here. So we solve marketing, we've solved distribution. 18:46.720 --> 18:52.000 The dream of course is to take away the need for the external marketing. What if your website 18:52.000 --> 18:57.200 was the marketing? Well, if your website did the marketing for you as well as the distribution 18:57.200 --> 19:03.760 and in doing so allowed you actual independence and not having to depend on giant third-party 19:03.760 --> 19:08.240 corporations who might rug pull you any second for the privilege of being able to reach your 19:08.320 --> 19:12.320 own audience or indeed a new audience. That's the dream, that's what we're working towards 19:12.320 --> 19:17.360 and that's what we're trying to achieve. And the key lesson that I want to leave you with for any 19:17.360 --> 19:24.080 of you who are working on activity pub projects or apps of your own is one that I've learned repeatedly 19:24.080 --> 19:30.560 over the years with Ghost in general and again with activity pub and something that I just see 19:30.560 --> 19:36.400 is being a missing lesson within the world. I have to remind myself over and over again and that is 19:36.480 --> 19:42.640 that just being open is not enough. I think for all of us in this room who understand open source 19:42.640 --> 19:49.200 and who are deeply invested in open source, the benefits of being open are so inherently obvious, 19:49.200 --> 19:54.400 it is confusing us to why anyone would choose anything else. It is utterly perplexing if you 19:54.400 --> 19:58.560 have two projects and you say this one is closed and has a billion dollars in funding and this one 19:58.560 --> 20:04.400 is open source, which one would you rather? Of course, the open source one, this one comes with strings 20:04.480 --> 20:08.560 attached. This one will probably be out of business in few years. This one is going to harvest 20:08.560 --> 20:14.960 your data. We all know this, right? But that's because we're so deep in it. And when it comes 20:14.960 --> 20:22.080 to choosing software a lot of times it ends up feeling a little bit like we're so wrapped up. 20:26.720 --> 20:32.720 Like we're collectively so wrapped up in the ideologies of open source and the ideologies 20:32.720 --> 20:40.240 of why our thing is technically better, that we miss out that consumers are choosing the 20:40.240 --> 20:47.840 products that are simply the easiest one to understand. And that's something I wish collectively 20:47.840 --> 20:53.680 we would all be slightly more aware of. And it depends a little bit on your goals. If your goal 20:53.680 --> 20:58.960 is just to do open source because you love open source, that's fine, no problem. But you my goal 20:59.040 --> 21:03.200 and if your goal is similar to my goal is that I want that to be more open source in the world. 21:03.200 --> 21:08.320 I believe the world would be fundamentally more useful if we collaborated more than competed 21:09.040 --> 21:13.840 and if we had more shared infrastructure to depend upon. And if that's your goal, 21:15.280 --> 21:19.520 you've got to look up from the cell phone. You've got to see what's going on because the most 21:19.520 --> 21:26.160 likely person to win this fight is the one who wants to win it the most. And that requires 21:26.160 --> 21:32.480 we the collaborators to be a little bit more competitive. It requires that we get outside our 21:32.480 --> 21:39.680 comfort zone and not just focus on the facts, but what we're doing is open, but try and make it 21:39.680 --> 21:44.880 better. Because Patagonia does amazing things with nature, but if they're fucking range 21:44.880 --> 21:48.480 jackets and they're like water through, it doesn't matter what they're doing for the ecosystem, 21:48.480 --> 21:53.360 no one's going to buy the jackets, right? So if the product doesn't grate, the rest doesn't matter 21:53.440 --> 21:58.720 and if you want to compete within the activity pub space, if you want to enact change to have 21:58.720 --> 22:03.200 this technology be more widely adopted, my message that I would give to any of you working on an 22:03.200 --> 22:10.160 activity pub app is to try and make the product great first and bring all of the open idea deals 22:10.160 --> 22:18.720 along with it. That's my talk. If you would like to hear more, we have a public beta of activity pub 22:18.800 --> 22:23.760 ghost coming next month or a few weeks. General availability will be the summer and you can follow 22:23.760 --> 22:29.280 the newsletter which has behind the scenes developer log updates on this handle or this domain. 22:29.280 --> 22:33.360 If you want to see it on the web or indeed you can subscribe to it via email because 22:33.360 --> 22:38.720 ghost lets you do all of these things. And always remember on the internet nobody knows your adock. 22:38.720 --> 22:42.720 Thank you. 22:42.720 --> 22:51.280 Well the biggest question we've always got is like okay so how many people are on mastone or on 22:51.280 --> 22:58.000 the Fediverse? There was a mastone both in the way. And how many people are on X and L's where? 22:58.000 --> 23:04.560 And I know we've tried so many different arguments. I think the email ones that we've not tried 23:05.520 --> 23:12.640 but do you do that actually work? So yeah. Really? Yeah. We have to have the words for sure. 23:12.640 --> 23:18.400 I mean you can also fed it to me as useful. That shows monthly active users across the whole 23:18.400 --> 23:25.680 Fediverse. It's climbing steadily quite quickly. But I think the the story that goes along with 23:25.680 --> 23:31.280 with this fact or this email narrative is you have to remember that open protocols don't 23:31.360 --> 23:35.760 go away because they're not dependent on one company. Ghost could disappear tomorrow. 23:35.760 --> 23:40.000 Activity pub would not. Massaging could disappear tomorrow. Activity pub would not. 23:40.000 --> 23:46.320 If and when any of the closed source companies go under so does their protocol they disappear 23:46.320 --> 23:52.560 immediately. So all companies have a wall around their protocol that when the company goes under 23:52.560 --> 23:58.080 the protocol goes under. But it's like a compounding thing over time. Open protocols. The 23:58.160 --> 24:02.000 user numbers typically just don't go down. They continue to go up because there's nothing to 24:02.000 --> 24:09.680 make them go away. They do not depend on a single company to succeed. I feel like I'm your new 24:09.680 --> 24:28.880 marketing department somehow. There you go. Evan. Evan. Let's get you an insurance policy. 24:31.760 --> 24:38.160 So you were mentioning before that it doesn't need to be on the open. We also have to make the 24:38.160 --> 24:45.600 coolest stuff. But right now what I see a lot in the Fediverse is just we are copying and 24:45.600 --> 24:51.680 cloning all these ideas of companies, of millions of euros like peer-to-peer, like masterdom, 24:51.680 --> 25:00.400 like pixel fat. And I think it's brilliant. But how do we get like I really like the design of ghost? 25:00.400 --> 25:06.320 Like what is the budget of your team? How you get funding? How do you brainstorm? I don't know 25:06.320 --> 25:14.960 this is brilliant. Thank you. Thank you. Yeah, I mean so it's been two developers and one designer 25:14.960 --> 25:22.320 working on it for about nine months to get to this point. We're expanding team next week to 25:22.320 --> 25:27.360 it's going to be five engineers and three designers. I think at that point to work towards launch. 25:28.400 --> 25:34.320 Funding is also self-funded. All from the blogging platform existing are selling hosting for it 25:34.320 --> 25:42.080 and making revenue from that for the nonprofit organization. So everything DIY. But how 25:42.080 --> 25:46.960 good your design is just depends on how much you value design in a big way. So how much are you 25:46.960 --> 25:52.800 willing to slow down? Because the design is not good enough. How much are you willing to not ship 25:52.800 --> 25:57.440 to keep that quality high? How much are you willing to spend on designers if you are in the context 25:57.520 --> 26:05.440 of a company? Ghost is more design biased because I'm a designer. Before Ghost, I worked 26:05.440 --> 26:10.800 on the WordPress core team and it designed there. So it's kind of a history of publishing plus 26:10.800 --> 26:18.240 design is really very much in my veins. But it's hard man. I don't know. 26:27.520 --> 26:39.360 All right. I should mention that John says that Ghost's implementation is DIY. But he is using 26:39.360 --> 26:45.200 Fedify, which is an open source library and he is actually contributing back financially. Which 26:45.200 --> 26:49.600 means that John is one of the few activity public developers contributing back to open source. 26:49.600 --> 27:02.400 Yeah. The library were used under the hoods to power a activity public. Functionals 27:02.400 --> 27:08.000 it's called Fedify. It's developed by Hong Min He who's an independent Korean developer who's been 27:08.000 --> 27:12.000 working on Fedify for I think a year and a half now. And he was just about to start working on it 27:12.000 --> 27:15.200 and go back to full time employment. And we said no.