When people talk about decentralisation in the atproto ecosystem, the conversation almost always centres on Bluesky. Conversations on decentralisation are messy at the best of times, but the strong focus on Bluesky and microblogging make it even more complicated. Decentralisation describes two different things at once: a technical architecture for how networks are structured, and the actual behaviour of people using those networks. Because Bluesky is so central to atproto, the social dynamics of hosting >99.9% overwhelm any technical dynamics that atproto provides regarding decentralisation.
Blogging platforms like Leaflet and WhiteWind offer another lens for understanding decentralisation and atproto. The blogging and writing platforms built on atproto operate on a much smaller scale, there are multiple distinct players in the space. None of these players dominate the field in the way that Bluesky dominates the microblogging side of atproto. By examining how these different platforms interact with each other, we can see some of the decentralisation dynamics of atproto that are harder to see by looking at Bluesky.
The current state of long-form writing and publishing
The first player on the scene was WhiteWind, which was also the first major alternative project on atproto besides Bluesky, and launched in early 2024. The platform offers a simple markdown-based editor, a homepage that shows popular and most recent articles published using WhiteWind. There are also some integrations with Bluesky, with Bluesky posts that talk about the article visible below it.
The developer for WhiteWind however has stopped further work on the platform for a while now, leaving the space open for other writing and blogging platforms.
Leaflet is a blogging and publishing platform that started development in the summer of 2024. Only in May 2025 did the platform become really integrated with atproto. In the half year since, Leaflet has quickly become the most popular blogging platform on atproto, and it is actively seeing further development. Leaflet is a block-based editor, that does not use markdown. Leaflet is now also starting to move towards the social side of blogging, with its own comment section (that exists outside of Bluesky), a reader feed that shows all recently published Leaflet posts, and a discovery page for finding other Leaflets.
What I find noteworthy about Leaflet is how quickly it is managing to find a user base of early adopters that is not tech people talking about the tech itself (like yours truly). Of the 20 Leaflets that were published in the last 24 hours, 3 were about atproto itself. Instead there are posts about football, politics, short stories, personal diary blogs, or overviews on how someone is putting a film camera on a boat. That many the blogs talk about things other than tech is a very healthy sign for the future of Leaflet.
PiPup is a recently launched platform for long-form writing on atproto, that focuses creating blogs with Markdown. It supports a wide variety of specialised tools, from creating interactive math charts to sharing music with ABC notation and code syntax highlighting. The platform launched with a tool to convert WhiteWind posts to PiPup posts. It also has a reader feed, but where the Leaflet reader feed only shows posts made on Leaflet, PiPup's feed shows posts made with WhiteWind, Leaflet and PiPup.
Another example of a writing and blogging app that's currently in development and that looks very promising is Offprint. Their current work can be seen here. Offprint also explicitly mentions monetisation as part of their platform, advertising Offprint as "Write, monetize, and own your audience without the middleman." Pckt.blog is blogging platform that is currently working on atproto integration. Weaver is another long-form writing platform for atproto that is currently in development, and has gotten a grant by Graze for the work.
On decentralisation and interoperability
Together, the various long-form writing apps create a new market for blogging on atproto, with a very different dynamic between them. Instead of one company that works on the protocol and also hosts virtually all users, there are multiple competitors all building in the same space. The Apps act more as collaborators than as competitors however. It gives some insight in what decentralisation and interoperability on atproto actually looks like when there are multiple different apps all working towards the same objective, without one company being overwhelmingly dominant. Some observations:
Due to how atproto works, you can create posts on any of the blogging platforms without actually using the platform. This is as simple as publishing blog post to your PDS, using their lexicon. This makes it that there is a distinction between using the blogging platform (by logging in to Leaflet for example) and using their lexicon (publishing a post to your PDS that uses the same lexicon). The nature of long-form blog posts make this use case much more likely than with microblogging. For example, people have been building personal websites with posts on the website using the WhiteWind lexicon, without interacting with the WhiteWind platform (and thus not agreeing to the WhiteWind ToS). There is not yet clear language to reflect this: do people do who this use WhiteWind or not?
Interoperability is an active point of concern between the developers of the various platforms, and something they are collaborating on. Because the writing formats of each platform is by definition open, interoperability is always an option, and if the platform developers themselves don't support it, other third-party developers might add it themselves.
Leaflet is the first blogging platform that allows people to follow a specific publication, creating a 'following' relationship that exists outside of Bluesky. I'm curious how the other apps will respond to this? Create their own form of following writers on their platforms? Will we see a move to a shared lexicon for following blogs? Or another option?
Because text is fairly simple format to work with, competitive interoperability is also an option. An example of this is how PiPup has created a tool to import all your WhiteWind posts into PiPup as well. Other tools by external developers to import WhiteWind posts into Leaflet also exist. In a well-developed ecosystem for long-form writing on atproto it seems highly likely that there will be easily accessible tools to port posts in every lexicon to every other lexicon.
Long-form writing and publishing platforms combine multiple different functions: the writing and publishing part, and the social function (commenting on a post on Leaflet), and the discovery and distribution part (the reader feeds on PiPup and Leaflet). This allows for further disaggregation in the future, and I would expect services that are dedicated only to the discovery of blogs to pop up. Such discovery services have an incentive to aggregate from every long-form writing platform. This creates a dynamic where the discovery services want to treat posts by the different platforms as much the same as possible, while each individual platform wants their post features to be different and unique from the other platforms.
Moderation remains one of the most challenging problems of any social platform, and that is no different in the space of blogging. Both Leaflet and PiPup have fairly lax rules around which content can be published on their platform. It is unclear what, if any, moderation systems they currently have in place. The challenge of moderation becomes even more complicated since the use case of publishing to a platform's lexicon without using the platform itself. The blogging platforms are also moving into the space of discovery and providing 'reader' feeds, and how they will moderate blog posts that they show on their platform remains an open question.
The existence of multiple different blogging platforms gives an indication of what the incentives for decentralisation actually are on atproto. When a platform like WhiteWind stops development, that does not result in other people setting up their own version of WhiteWind. Instead, people start building their own software for the same market. There is no real reason to run your own version of Leaflet, but there is an incentive to build another publishing platform that is compatible with Leaflet but also has other features. A similar dynamic is visible on Bluesky: there is little incentive to run a direct copy of the platform, but platforms that are interoperable but can also offer something new (Blacksky, Gander) are able to find their place on the network.
I continue to think that the field of blogging and long-form writing on atproto is the place with the most potential for growth of the atproto network, and there is a ton of exciting experimentation and building happening in the space. I'm curious to see how this will further develop.