Building Bluesky as a Platform - Eric Bailey

We met with Bluesky engineer Eric Bailey to hear all about what it has been like to watch its popularity surge over the past year, how it uses the AT Protocol to let users own their own data, and what the future looks like for the platform.

----------------------------------------
Welcome to WeAreDevelopers, the #1 developer community in Europe!
This is your one-stop destination for the latest tech insights, tutorials, and career advice to elevate your developer career.

Stay updated Dev Digest, with our weekly newsletter featuring the most recent tech trends, career guidance, and original content crafted by developers, for developers. Subscribe now

Interested in advancing your career? Browse through our job board featuring over 190,000 jobs. Unlock job opportunities with a free developer profile

Don't miss out on the annual highlight of every developer's calendar - the WeAreDevelopers World Congress. Network with 15,000 peers and learn from over 500 speakers. Secure your spot and save 10% with "wearedevs_yt"

#CareerInTech #Tech #ProgrammingTutorials #DevRel #CodingTools #TechJobs #TechTalks #DeveloperSkills...

Transcript

Welcome to another coffee with developers. And here's Eric, who has his coffee. I got my coffee as well. Now Eric works for a thing that made me happy lately. And the thing is, right now, social media has changed.
There's places where there's lots of followers, where you have lots of people that you follow, and when the mood changes, then you don't know where to go. Right now I feel a bit lost. There was X, there was. Twitter is now X. Not a fun place for me anymore.
Facebook is where I do a lot of nerdy stuff, but not like the normal thing. LinkedIn is for work. Then I got Mastodon, which is impossible to explain to a normal human being that is not a geek. And then there is Blue sky that you work in. And I'm very excited that that happened because when I went on Blue sky and I started looking into it, I looked at threads as well.
It didn't quite grow as fast as I wanted to. Everybody that followed me and I followed in 2005, 2006 on Twitter, then came on Blue Sky. It felt like it's the same creative community getting together there again and starting from scratch. Why do you think that is? Why did people think Blue sky was the place that people flocked to?
What was the thing that you think made people go there? Yeah, well, it's a good question. I mean, I think to some extent there's something that just speaks to people about these like text only platforms. I mean Twitter, when it first came out, we choted a little bit before we started recording. But you know, it started out as a text message thing and it was like this, you know, interfaceless sort of, you know, way of communicating with people or communicating it like one to many kind of communication.
And I think that like made sense to a lot of people back then and it still makes sense to a lot of people now. Text only social platforms like Reddit and things like that are also know, massively popular. And so I think there's like a subset of users that are active on social media that still really like the format of kind of like dense content and the kind of like, you know, the lore building and the myth building that can exist on a platform like this that doesn't necessarily exist on something like Instagram. And so that's, you know, to contrast with Instagram. Instagram is just sort of a different modality entirely or like something like TikTok.
It's just purely kind of consumption as opposed to, you know, creation or like this is like an overused word but like communion with Other people. It's like you're creating like communities, you're chatting with people, things like that. And so I think like, you know, platforms like that still have a lot of cultural sway with people and people are still interested in them. Why Blue sky in particular took off. You know, I think there's some subset of people when they first joined that were interested in the decentralization aspect of it, which we're still proving out and still working towards from like a protocol level.
And so like that's very compelling to a lot of folks from the standpoint of like owning their own data and this network, living on if you know, they had ever got bought or it closed down or something like that. Neither of those things are on Horizon, of course, but. But you know, there's like those sorts of people that they were attracted towards it and then I think, you know, most recently, I think it's pretty clearly just the things that are happening on other social networks that people are, you know, they rightfully are concerned about. And you know, they're being driven here not necessarily because they know like the Twitter like modality, because they are interested in this like new, fresh space which you know, has policies that they agree with and a lot of other like, like minded individuals and things like that. So I think it's like shifted over time.
There's still like this suburren of, you know, this is like the modality that people like. And they're here for decentralization. They're early adopters, they're like thinkers. They're people who wantna like have conversations and things like that. But then there's a lot of folks that are coming on more recently that you are just here because they're like, hey, this is fresh and this feels good.
And you know, at the moment it feels better than other social networks. So it's an interesting dynamic to cater to both types of audiences. I'm sure there's other audiences there as well, but those's the first two that come to mind. I think it also is. I mean, Mastodon is also decentralized, but as I said, it feels geeky.
And I think the interface of like Blue sky brings the same decentralization, but it feels like a simple app to use. And that's what people are looking for these days. They just want to download an app and get started and not like find out which server they're connecting to through which person. That was a winner moment as well, I guess. Yeah, that's a really good point.
Yeah, Mastodon definitely makes it very clear that you're on your own server and it's distinct from other servers. And I think that's like a really powerful thing for small communities. But what Blue sky set out to do was a little bit different, to kind of obfuscate that complexity a away so that people don't have to think about it but you know, work towards like a world where like it can just easily seamlessly become you know, the kind of like islands world of different decentralized nodes, different PDs, hosts of your data that all work together in this like global network. And that's. That made Blue Sky a little bit unique from, from Mastodon.
And you know, thus far I think folks have enjoyed that. So, so one thing I like when Twitter came out was that it was also a platform. I could actually build bots. I can build like I can pull the data, remix it, it get kind of like it was. There's RSS feeds for everything.
There was a simple API for everything. Now you dialed that up to 11 with like blue sky, right? Like what's, what's the. How do I access the platform of Blue sky and what does it give me as a developer? Yeah, totally.
You know, I like to. So I come from like a kind of front end background. Most of my friends in development are front end or like kind of like me just sort of application developers that do some back end and know the way around postgres and things but aren't necessarily doing like infrastructure and stuff. And so when I'm talking to them and trying to pitch them on Blue sky, what I normally describe it as is sort of like a global public database where everyone can host their own data and it's globally available on public APIs. And so you can use that essentially as like a backing store for lots of types of applications.
Specifically good for applications where users are generating their own content, not that it can't be used for other things. And so it's like, I mean how you interact with it is, you know, it's just purely by these APIs. And we've got infrastructure that is like publicly available like our relay which like relays data between all of users PD'like these are their hosts where they actually write their data to it. Realize that out to any subscribers that want it. So this is all public data.
There's kind of like an interesting architecture you have to understand. But more or less for the average application developer you can think about it as like you write to one place using an API that we provide and then you read from this global firehose of data from the relay, which is again, it'just like a public API we provide. And then once you get that data into your application, you can do whatever you want with it. You could put it into postgres, you can put into Mongo whatever you're comfortable with and then run whatever application you want on top of that. And so it's really kind of sky's the limit at the moment.
There's some sharp edges here and there, you know, that we're working on, but it's getting very close to being like a, you know, a very easy thing to get to get set up and running. If you're you remotely familiar with handling websockets and databases and things like that. Do you provide SDKs? Do you provide demo code somewhere where you can get started? Yes, we do, Yep.
There's a whole suite of. So most of our repos are open Source on our GitHub and within that is essentially every reference implementation that you need right now, Like I mentioned, there's a couple things that we still need to kind of create blueprints for, for instance, like repo sync and backfilling people's data. Like if you wanted to bootstrap theue application and get all their data from their PDs into your application, that's a little bit difficult right now. We don't have like a very clear roadmap for that. The code's all there, but it's not maybe clear to dig out.
But yeah, within those repos there's a whole suite of typescript packages, there's a few Go packages as well. And the community has written packages for Python and other languages. And what those packages do is use our lexicon system, which is sort of a JSON schema type thing that defines the interfaces and defines the RPC methods that we support. And we use those to generate these clients. And so these clients essentially just have methods that are say like, well, you like login for instance, or like we have an OAuth that we're rolling out at the moment so you can use OAuth to log in, but then it's like get post or get post thread or get profile or whatever, these things that once you're authenticated, you can just, you know, fetch the user'data that you need and it comes back in like a nicely strictly typed, if you're using TypeScript or whatever, strictly typed format that's, we hope, very easy to use.
And we've also got documentation websites that are getting better and better every day. The app Proto which is like the underlying protocol of Blue sky documentation site is like quite complete actually. Like lots of good guides and things. There's actually a good starter kit on there for building your own full stack application on app Proto so highly recommend people going to check that out. It's actually a really, really nice guide and there's a reot to accomp company that you can actually pull down and run yourself and like you know, test this out.
So yeah, we, we're working more and more on these kind of like reference impitations and toy apps and things like that to help people get up and running in minutes as opposed to you know, hours of reading through docs and things like that. How much interest is there from the outside? Like is it just enthusiasts or do you have companies approaching you as well to say like we want to move our content there? Yeah, I mean so it definitely started with individuals and like the community is actually quite strong I think. I mean as we, you mentioned, I think before we started recording there's all these community led projects like Flashes and Pink sky and gosh, what's the blue screen?
Which is a funny name but that's like a TikTok, a TikTok app style app. And so there's people like already building full stack applications and there's lots and lots of starter kits for like little bots and chat bots and feed generators and things like that. Feed gens are very interesting in our network. Anyone can create a feed quite easily in their services that help you do that as well. And from a perspective of companies, we started to get more companies reaching out.
We've talked about doing integrations with other like blogging platforms and things like that. We have close ties to WordPress, that kind of thing but nothing's really like I don't think solidified there exactly. I'm probably the wrong person to ask actually. Probably wanna talk to like our relations folks like Emily, if you know. Emily, Emily from the network.
But certainly we're getting interest and we're seeing more and more organizations move their accounts over. So you know, folks like the New York Times getting their journalists verified domain names that use the.nytimes.com domain. So you know, that's really exciting too. Organizations are, you know, they do seem to embrace it and they're getting like more excited about using it as a platform, you know, to communicate with their, with their users or their fans or, or whatever and we hope to see more people, you know, building like applications and really storing their data on it in the future. And that's, that's one of the things that we're really like driving towards is making those facilities more mature and you know, capable for use cases like that.
Emily has a post on mozilla.org today as well about the future of social media. I always excited about that because I used to work for Mozilla and like basically I'm like oh this is still there. So this is fun to get these kind of support. The domain thing was a good point. Like there is a way to validate yourself or log into the platform using your own domain.
I didn't find that straightforward to do. Is something that. Because there was also a really strange way to have like sub. I don't know, it was a dctor had a blog post about this where he said like the turning things into clickable links, you had to actually count characters within it when you logked in with your own thing. Yeah.
So that seems to be a thing that is quite an uphill battle for people to get started right now. But I guess this is a full request a way to get better, right? Yeah, I mean more or less. We are open source and so we always welcome pull requests to improve our UIs and UX flows and things like that. Yeah, the domain thing is definitely a kind of a sticky point for a lot of folks.
It is a little bit more technical right now. You do have to touch DNS or you at least need to be able to deploy a new file or like a new route to your website. There's two ways to verify you to DNS or via a well known a JSON file that you just upload to your site and those are like fairly easy concepts. But it is still like manual work and most people even if they have a website, don't necessarily control it to that extent or understand DNS. I mean DNS still kind of scares me every time I have to touch it.
But yeah we are. That is something like on the roadmap eventually is to help make that easier. We do have a domain product like a separate little like micro site where you can actually purchase a domain and like automatically hook it up to your Blue sky account. But it's very separate, it's very rough. It was like something I think we released it in maybe June 2023 and it basically hasn't evolved since then.
And so like you know, it's something we just keep thinking like oh we'll get it, we'll get that eventually into the app, that kind of thing. And you know, we just haven't gotten around to it yet. But the domain thing is like an exciting and kind of like unique part of Blue sky that we definitely want to. To push on more because it's, it's a great way for I gu said New York Times or other well known domains to verify their employees and users and things like that. I think you'd get up pages, you just put a cname file there and actually get the domain connected that way.
Or when you got your site blocked by Google for some reason you had to put the meta tag in there with a hash. So that kind of stuff, once you did it once, it probably can be revalidated. So that's one way of doing it. But I think it's an interesting opportunity to actually get like people to comment on your blog from Blue sky, for example, or to get data back into your systems. Right now I do it by.
I mean I used to do it with the Twitter API. Now I do it by scraping Twitter and basically then turning it into a JSON file and putting it out again. But it feels weird to have to hack your own feed. Yeah, in your case you. That's actually.
Right, exactly. Yeah. Actually you bring up a really cool use case that has started to spring up recently, powering comments pages on blog posts and things like that using Blue sky posts. It's not like, you know, it's not something we necessarily like intended when we built these kind of systems, but it works really well for that. And you know, we're you potentially interested in doing more there.
I know everyone probably remembers like Facebook comments or like Discus or something like that. Yeah, folks are using it. You know, it's like they'll post a post with a link to their blog and then use that. All the replies to that post become replies to the blog and it shows up on their blog. And since this is all open data and it's free to access, you know, they don't have to do any of that sort of like wacky scraping stuff.
It's just as easy as just pulling in that data and rendering it. So it's like super cool. Would love to see more of that kind of thing or more deeply integrated into something like, you know, medium or something like that. Could be, could be really cool. Who knows?
Sky's limit, that's not on the roadmap as far as I know, but got out to medium. That would be really cool if that were to happen. Yeah, it will be. I mean the problem we discuss and with Facebook was that it was tracking your users across domains as well. So that's why it actually, most ad blockers actually that don't allow for discuss any longer and Facebook as well.
So that's a really tricky one to get the open source people to do. So how big is the team behind it at the moment? Like how many people are you on Blue sky? Yeah, we're 21, I think at the moment. 21 core contributors.
And then we have, you know, a team of like 100 moderators as well and some other, some support folks and things like that as well. But the core team that's contributing to like the code bases is 21 at the moment, I believe. Wow. So when you do you do. I mean you said you had moderators, but do you do also automated scraping and blocking?
Like do you have Akismet running or something like that? Or how does that work? So if I were to use comments from Blue sky on my blog, is there any way that spam already gets filtered out or do I have to do it myself? Yeah, that's a very good question. We do have a lot of facilities for that internally.
So we mentioned the human moderators first. We have an entire application that's dedicated to just purely doing human moderation that's called ozone. That's also open source. Folks have run their own instances of this as well. So they can moderate sub communities within the platform.
You could subscribe to any of the moderation services that you would like. And then what those moderation services do is they get reports or they look at the firehose of data coming through like our public relay and they apply different rules and things like that. It might be natural language processing or it might be image detection, things like that. And they apply what are called labels. And these labels are just other JSON objects that are lexically defined in our like lexicon systems.
So they're shared types so that our application understands how to react to them. But it decorates all these records with these labels. And so those labels we can then use to either filter content in the server or filter content in the front end or sometimes just hide content entirely if it's something you don't wanna see. Maybe it's like adult content or something like that. And so that's like the core of our moderation platform is basically these labels.
And so then yes, we have automated systems as well that you know, scrape for, for instance, we use. I forget what the platform is, but it's a very well known csam, like child being stuff scraper. So all that stuff is like very behind the scenes. They don't open source that of course cause it's very sensitive info but we send these things off to them. It's you kind of best in class software to ensure that we're not getting any sort of that content on the network.
And that's like one of the first things we run. And there's different like layers of these automated systems. But in the end if we aren't certain what's happening, it falls back to human moderation. And that's where those human moderators really come into play. But you know, we're trying to get better and better and more efficient at processing data so that we can catch everything before it gets on the network.
Cause you know, some things still get through and if that happens then users can report those posts from the app itself. So if they see something that slipped through and maybe it didn't get like adult content tag or label, they can report it to our moderation service. They can actually report it to any moderation service that they subscribe to. And those moderation services then can decide to apply those labels. And so those labels apply no matter where you fetch this data.
So if you're fetching like your replies on your blog and you subscribe to certain moderation services, those labels will be applied, that content will be filtered and you can do additional filtering on the front end to if you really wanna just like prevent anything from slipping through that kind of thing. And that's like what's kind of like really interesting about Blue Skies as this idea of like stackable moderation. You can combine all these different sources of data, you can run your own automation tools and you can apply your own labels to these posts and then that data is now like globally available and so people can can react to it. I mean it always feels like weird that you want to filter things at the same time. I mean I worked for Yahoo, I worked for lots and lots of community places and the self filtering community or like a community note doesn't work, not in the times of bot armies that actually can upvote things that other people don'vote.
So there has to be a level. And of course if you weren't open source then the bad players would be able would find ways to work around it. So there has to be a bit of a secrecy in there but it always feels kind of icky to do it as an open source person. Right? Yeah, well actually I mean I think essentially all of our stuff is open source except for some of those like sensitive APIs.
We're using third party services for you Know the child content and things like that. And which makes it, it does make it actually quite interesting because people, you know, do have a pretty good understanding of how to game the system. And so it has made it kind of a unique challenge for us, you know, to react faster than the spammers. We're getting quite good at it. I know, for instance, yesterday we saw a wave and I think we were taking down 25 accounts a second, which is like, you know, that's pretty good.
We'reacting quite fast to some of these things now. So we're getting better and better. We're definitely aware of the spam problem and the challenges that face us in that regard. But, you know, we'll improve over time for sure. I did a recording for coffee with developers earlier about image recognition and automatic image analysis.
And I love that years ago that was the thing that sand dunes were actually filtered as adult content. Oh, really? It looked like arms and legs kind of thing. Like, you know, like. And that was a big problem.
That was. The system didn't quite understand that in terms of traffic, you saw a few spikes at the beginning. Like, there were lots of headlines that like Blue sky got like 200,000 more users or whatever.
Was it like, was there waves and waves of more people coming in, or is it a continuous growth or is there a decline happening right now? So what's the situation like? Yeah, it's been interesting. When I joined, I think we were at about 180,000 users. And back then a wave of 200,000 users is obviously very exciting.
I think we broke a million sometime in fall of 2023, maybe September, October or something like that. And since then it's gone. And just sort of what I understand is a fairly standard growth curve for social networks like this called like a step function. And so, you know, we hit that 1 million growth, uh, just over a year ago. We released, um, we removed invite codes and we saw growth from there as well, up to, I think, cash.
I think at that point we're like 3 or 4 million, maybe 5 million, something like that. And then, you know, we stayed flat for a long time. And then we see another. We saw a big wave of Brazilian users joining and we jumped up to something like seven or eight. And then in November, around the election, that kind of thing, we, you know, over tripled the network.
And so it goes in these kind of waves of like, you know, goes up and then it flattens out and it goes back up and then it flattens out. And then now we've jumped way up and you know, we've kind of flattened out a little bit, but it's actually in the last few weeks we've, you know, started spiking again. If you look back at growth from Instagram, for instance, their next step function once they got to like the 30 million point was about 80 million. So, you know, we'we're all rooting for that. You'll see if we get there.
But, but yeah, no, I think we're following a pretty standard curve. There are ebbs and flows of interest. You know, when we release new features, we get a few more, you know, we get some more excitement. People kind of log back on. Sometimes there's criticism of us and you, sometimes rightfully so, and things kinda ebb and flow.
But overall I think we've seen essentially just ever upwards. Like at the moment the growth is still sustaining and it's super exciting for all of us. So how does it work in terms of like when you reach say a Taylor Swift tomorrow says, I only do blue from now on. Traffic costs money. Like machine computing costs money.
Like how do you make money? Or do you. Because of the distributed nature, you have to care for a lot less traffic than others do. Or what's the model behind it? Like how secure is it that we will have?
Even beyond Taylor Swift, we will have blue Skye? That's a good question. We often use Taylor Swift as the example as well. So Taylor, if you're listening, let us know, let us know what you're thinking. I call it later.
It's all good. Yeah, yeah, okay. Yeah, lined up anyway. Text her later and just let her know. Yeah, yeah, that will be, that will be an interesting time.
Thankfully we do have a lot of capacity and we've done a lot to make our like per user costs quite low. A while back we moved to entirely on premises infrastructure or not entirely. I mean, we have some, you know, we have redundancy and we have other systems external to that that we use as well. And I'm definitely not the right person to talk to about that. That stuff is super interesting though and there's a lot of people on the team that would love to talk about it if you're ever interested.
But yeah, in terms of like handling that capacity, we should be pretty good. I mean that's like kind of the general consensus internally. Of course. You never know, you know, when things kind of happen. But we've done a lot just in the past few months because we've seen the network, you know, triple, almost quadruple at this point to make all of our systems more efficient.
That's essentially tied up a lot of the backend and infrastructure team for the last few months. Um, and so the, the work I've been doing has essentially been like without a backend team to kind of like, you know, to work with. So it's a lot of just front end tweaks and things like that. But yeah, in terms of like trying to scale and uh, and maintain Blue sky long term monetization is something we're definitely starting to think about. I don't know like how much I can say with our plans there.
But overall like we don't want to become an ad sponsored network because that incentivizes things that we do not want to introduce to our network. For instance, we're really good with links right now. Publications see more link traffic from Blue sky than they do from X, even though we're like orders of magnitude smaller than them still. And one of the reasons that is, well you, it's sort of a theory but you know, when you're an ad supported networks, you wanna keep eyeballs on your platform as much as possible. If you click on a link, that person is being taken off your app.
And so we don't wanna do that sort of thing, we don't wanna like incentivize people to stay. And so what we're trying to explore is ways to enable third party creators to make money either through like peer to peer payments or. There's been some interesting experimentation with feed generators. There's this graze that's social that allows you to create feeds and they're actually introducing a way to sponsor posts and like boost posts within these feeds. And these feeds, they haven't talked about feed generators yet, but they're entirely plug and play.
So anyone can create a feed and it can be published and then you can follow this feed or you can pin it to your home screen and see that content. And so within that feed, if these feed creators want to you know, place an ad for, you know, KFC or something like that, you know, it can show up in that feed and it's like it's an ad, it can show up towards the top. Users on Blue sky would't necessarily know it's an ad because we don't have any sort of UI to show that it's an ad. But that's an example of how like third parties could actually like make money on Blue Sky. And so one of the things we'd like to do is to help facilitate types of things like that to Help our community grow and incentivize people to build on Blue Sky.
So it's sort of a rising tide, raises all boats sort of scenario where you know, we'll get maybe a small cut of that for facilitating maybe payments or something that. But really the majority of the profit will go towards these communities and creators and you know, hopefully grow like that to a point where we can be sustainable. So that's just one of the things we're talking about. You know, there's some other things as well that again it's like, I don't know, like how much I can say externally and I also don't necessarily have a full picture of, you know, our future plans in the moment. We'll talk about that probably the next time we have a team on site so I'll know more and a couple months but.
But yeah, definitely something we're focused on right now is we want to obviously go the distance and stick around long term. So you're distributed then or you're remote or what's the structure of the culture in the company? Yeah, we're all distributed. There's a handful of folks in Seattle and some in sf, but then the rest of us are kind of scattered about the US and we have, we have some folks in Europe as well. Brussels and London and Italy and you all over.
So very good for time zone coverage as well. It's nice to have someone, you know, for on call kind of things, pager duty. Someone's kind of always, always online essentially. And so, and yeah, so we're all distributed. We get together a couple times a year, two, three times a year and get everyone together in the same city and that's always a lot of fun.
I really value those times cause the team is so talented and I don't often get to interface with you know, some of the deeper back end like infrastructure guys and stuff like that or even really I don't get, you know, Jay and Rose, you know, are so busy with, with you know, talking to folks and you know, our advisors and board and things like that that I don't get a lot of FaceTime with them. So it's always really exciting to sit down with them and get to see them every day at these team on site. So yeah, I'm not surprised with the link thinging because one of the main changes, like a lot of things annoy me about X at the moment but one of the changes that annoy me the most is that actually if you have a post on X with a link in it which I normally do. I find something cool, I share it and I do a comment on it and say like this is happening and here is a proof URL where it happens. The algorithm of X actually punishes for you for that now.
So basically they just want content. So you stay within the app and then most of the people do some post and then the first comment will be the link people on LinkedIn as well. Although LinkedIn doesn't do that algorithm yet. But it seems like the growth hackers of the social medias out there are now all thinking like I have to make a random statement without any proof that is actually worth more than something where people can make up their own mind by following a link. So I'm not surprised that this is actually paying dividends in your platform because it'on the other it's actually discouraged because it's the whole like stay within the app, stay addicted, become TikTok basically that'all of them want to become.
And I think the traffic bit is also because video and images are actually also valued higher than text on most of the other platforms now. And that drives me crazy. Like, I mean I go to Google and I search for like, how do I do pancakes? I want to have like a pancake recipe. I don't want a 7 a half minute video of like people, please like and subscribe.
And my pancakes are the best. And yeah, it's a very strange world where we became like everybody's a content creator and everything is becoming like multimedia is better than that. And my favorite are is YouTube thumbnails where everything is like. Or. Yeah, right.
Yeah, exactly, yeah, it's changed so much. It's almost. Yeah, yeah. I mean, no, yeah, you make a really good point. It is like very funny and also like kind of dark in a way that, you know, search engines like Google don't turn up really great results anymore.
It's just these, it's the results that get the most eyeballs. I mean especially for someone like Google, you know, I can't speak too much but you know, like they're driving folks towards, you know, YouTube and things because Google owns YouTube. And so it's like this kind of like strange ecosystem of these like huge platforms that, that you know, have their own best interests in mind. Of course. I mean it just, that's just natural.
But you know, they've catered towards kind of a human condition of, you know, we want easy, snackable kind of content and sometimes that doesn't equate to really high quality kind of stuff and stuff that's actually kind of nutritious and good for us, you know, I don't know, you know, like Blue Skies is a social network in the end as well. But we hope to not fall into some of those trappings. Cause we're all people, all, you know, everyone on the team is a person that's deeply invested in the Internet and has been, you know, extremely online from a young age. And so we've seen this kind of shift and we don't like it, you know, and so we do not want to become that and we want to actually facilitate the opposite, you know, to create healthy interactions. A good content repository, you know, like this is like the public town square, I think that the people need, and we wanna make that like as healthy and as good for humanity as possible.
Without, I mean, not trying to be too lofty here with our goals, but we all are very much on the same page here. There's no one on the team that necessarily set out to build a social network necessarily. We were here to build a protocol that can power the future of social networks and other applications and just the wider Internet in general. It just so happened there was a great opportunity to build a social network. And it'you know, it's proving our hypothesis that, you know, we need like this kind of like public infrastructure, a way to own your data, stackable moderation, you know, choose your own feeds.
All these kind of theories that we had are starting to be proved out. And it's a very exciting time for a platform like this to be growing up just because of the currents that are happening in the rest of the ecosystem. So, yeah, it's gon toa be very interesting. It's been a very interesting experiment. It's going to continue to be a very interesting experiment.
Excellent. Feeds are a good point. I mean, I get most of my information from RSS feed still. I'm using Feedly. I was really annoyed when Google Reader shut down because they're.
I mean, it was very, very outdated and it was very hard to maintain. But I mean, the argument that they could monetize it made no sense because you can easily monetize a feed, but there's other platforms that try to do that and I love that the data is distributed and that I can actually always get a copy of it, no matter what happens. Because Tumblr was a great platform and it started as well, it was lovely. And Reddit, of course, like all the things there. But as soon as it's closed and it's really hard to maintain and it's really hard to actually get data out of it.
I mean, you can get export from Twitter, but it takes two days, and it is a really unstructured data feed, basically. It's like, here's a dump of the database, basically. So I like that a lot. I hope that it will get into people's communication channels. But I like that you said, we don't want to become a social network, because that is also where I.
I almost lost interest already in Twitter in, like, 2010, when everything was about celebrities and sports teams and companies. And you're like, I don't care. Kellogg's doesn't have anything to say to me as'not a person. You know, I mean, you can do it by having people working for companies. I mean, I love, for example, that Sweden had, like, gave the Twitter account to a Swedish citizen every day.
And it could be the most random people posting whatever they wanted. And really? Yeah, I kind of love that. Like, that's kind of like, that's amazing. Yeah.
That's what a government can do when it trust its people. But it's just's. Really. It's really interesting to see that. That your aspirations are differently than just, like, we want to be the next one.
Ello. I don't know what. Yeah, yeah, yeah. That was fun. I remember that.
Everyone rushing to get their usernames and everything. Yeah. I don't know what happened to it. I think it shut down recently, actually. Yeah.
All right. I mean, the fun bit that, like, the fake currency and the fake scarceness of, like, invite codes works at the beginning when you want to grow something like that, it really happens. Like, I remember, like, oh, you got a Twitter in mind. Oh, that's amazing. Yeah.
Yeah. You. Is strange that. Yeah, we saw the same thing. Yeah, exactly.
Like, people were very excited. I mean, there were folks, you know, months after we removed invite codes that were like, I'm still waiting for invite code. I'm so excited to join. Yeah. It's a strange thing, you know, humans are.
Humans are strange birds. Yeah, well, same. Same birds. We had. We had the bird, we had the egg, and now we got an X, but you still got a blue sky and you got your butterfly.
So that's a good thing, too. Yeah. Cool. Well, I thank you very much. This was great.
Anything you want from our community. Like, what should people get started with that? What should feedback do you want? What can we do for you? Well, I mean, I'd love to see people building on more platforms, you know, like, we're not quite at the point where we can like build on something like Vercel or something like a full stack application, but we're very, very close.
And so like understanding like a lot of the pain points for developers that are working with maybe more. I'm saying this from like a front end perspective, you know, again, like I'm not a deep infrastructure gu. I'm not gonna be running like lots of Docker containers or Kubernetes or anything like that. And so like I'm really excited for the day when people can spin up a starter kit or a boilerplate on Vercel or Cloudflare or netlify or something like that and build more or less fully featured applications on our network. And so understanding kind of the pain points there, you know, issues with like Backfill for instance is one that comes top of mind is really, really valuable.
And just knowing like where the gaps and knowledge are. And you know, in the past that's predominantly come I think from external developers working on things. Of course we all, you know, dog food, this stuff too. We all have like our little toy apps and things we're building, but community members perspectives are, you know, far more valuable than ours sometimes because we have this kind of internal bias of like generally knowing how things work and stuff. So getting someone bootstrapped from zero to one and logging with OAuth and ingesting data and handling, you know, user stuff, it's, you know, that's like super, super valuable.
And I think we're like right in the cusp of being like really good for that. And so I'm really excited for, for developers to start building, you know, toy apps, but building things that you know, could go the distance and you know, really scale up at something point. So talk about front end. I'm a huge accessibility person and I'm huge, I'm old. But like what I find amazing is like when you upload an image and you basically it says immediately here, alt text describes images for blind and low vision users.
Thank you so much. Every image should have an alternative text. So that's really cool that it does that. What drives me crazy is though, like why isn't that an entry field? Why do I still have to press that alt to actually start editing it?
Yeah, that tiny little button. Yeah, that could be just the text box that allows me to do it. Right, that's a good point. Yeah. Right?
Yeah. Accessibility is definitely something we need. You know, we've done some, but I'm sure you'found other pitfalls as well. It is something on our minds. I Actually currently have some work in progress code that depends on some other like typescript things.
We have some facilities that are gonna make constructing these types of UIs a lot easier, coming very soon. And one of the things I wanna do after that is really like revamp some accessibility things and the composer in general, lots of plans for improving that. And so that's actually a really good call out. To make the alt text more obvious and easier for people to add. We've also had a suggestion of like, actually, do we have this integrated yet?
I don't know, but like trying to like suggest alt text based on the image, you know, it's a fairly easy feature to get in. It's really. There's like no reason for people not to use alt text other than it's, you know, it's kind of difficult to add and it takes time. But if we can like grease the wheels there and make it easier, then there's really no reason we can't make all this stuff like extremely accessible and pleasant for people to use with a screen reader and stuff like that. So, yeah, definitely, definitely on our minds.
If you have other thoughts, you know, definitely hit me up on the network or anyone that's listening to this as well. We're very much open to these types of suggestions and the GitHub's open as well, so, you know, feel free to file an issue. We need to get caught up on issues a little bit. I'll recognize that too. We're a little bit behind, but.
But we do want to hear about those things and get them on the roadmap, so. Excellent. Well, Eric, thank you so much. This was great just for coffee with developers and we're going to release that in a few days and it's going to be the newsletter, so great stuff and see you in blue sky. Awesome.
Thank you so much, Chris. This has been great, really appreciate it.