thisismissem@hachyderm.io I admire your passion and I'm sorry that your proposal was rejected for "procedural reasons," especially because I got the impression that the reason wasn't just procedural, but above all ideological.

I admire your passion—I was saying—and this leads me to ignore the unpleasant accusations you've leveled at me. They're unpleasant and false, because there's no misinformation in my words: the fact that some developers are working hard to create cross-pollination between Bluesky and the Fediverse doesn't mean this has any impact on Bluesky PBC's strategies.

In fact, it's always worth remembering that just as the board of directors is on the top floor, making decisions based on the prospects of return on the backers' investment; on the first floor is the kindergarten where the most visionary developers generate ideas.

The point is that these ideas aren't functional to a business model! Creating an app that collects data is; creating a LinkedIn instance is; Creating a PSD linked to a brand is; inventing monetization tools is. Conversely, building bridges and cross-fertilizing is a fruitful intellectual exercise, and it also builds a legacy for humanity, but it doesn't serve the business. And if it does, it's only in the long term. Too long...

In fact, Bluesky doesn't answer to the community, but to its backers. And backers will never understand why it's necessary to invest in cross-fertilizing with Activitypub, nor why this cross-fertilization could lead to a decrease in ROI or an increase in that "R"...

Moreover, as reported by Forbes, various data points to a significant year-over-year decline in Bluesky's average number of daily active users on mobile devices; consequently, according to Social Media Today, activity and engagement on the platform have decreased, and despite the increase in overall user numbers, post volume has declined.

Today, Bluesky's backers are not happy. And hearing about Activitypub will make them even less happy and more nervous, unless they read a nice, "strictly confidential" slide on why "cross-pollinating with Activitypub" could make Bluesky the sole centralized social media player in the next three years: that would make them much happier.

As for developers who enjoy protocols, I admire them and support them if I can; but at the same time, I urge them to be careful not to waste their time on a lost cause. ZOT and Nostr are so inspiring...

> Disclaimer: I apologize for my English, but I don't know this language and I write only with the help of an automatic translator.

I feel as though I need to jump in here to de-escalate, as things are starting to get out of hand.

What I can safely assume is that if we are not in the room, we cannot make any predictions on what will or won't happen. It's easy to mis-trust large organizations and assume the worse intentions, but that is biased and unfair.

Do you disagree? Probably, but I've seen the exact same malintent levied at a smaller scale against the Mastodon developers, simply because they are the largest implementor of the fediverse and are therefore bad by definition.

Something I hear often is the argument that one should judge a company not by what it says, but by what it does → don't blindly trust Facebook just because it says it has your best interests at heart, look at what it does with moderation tooling (or lack thereof), data privacy, etc.

So we need to apply that same lens to BlueSky. thisismissem@hachyderm.io is saying that developers there have been more open and forthcoming to some of her ideas — ideas important to her, and to the fediverse as a whole, and that she has received a rather lacklustre response from the AP side. This is an important observation we should take into consideration.

My approach to these issues is probably unfiltered, and I'm sorry if this makes my statements seem too categorical.

I should also point out that when I talk about Bluesky being tied to huge funding (and therefore adequate returns), I'm not expressing a moral judgment. I want to be completely non-partisan on the matter.

I simply wanted to provide my best explanation for the disruptions we're witnessing as a result of attempts at cross-pollination between the Blueskysphere and the Fedisphere.

Let me try to explain myself better:

• considering the vertical nature of Bluesky PBC
• considering the horizontal nature of the Fediverse
  I don't rule out the possibility of cross-pollination, but I do rule out the possibility of it being guided by the same principles.

The development of the Fediverse is, in fact, driven by the community of developers who work only on the application layer and know that (almost) none of them has the power to decide on the "protocol." And it isn't based on a single business model. On the other hand, those who have decision-making power over the protocol know that any change would have a huge impact on an extremely diverse ecosystem. It's not easy to decide what to change because it's not easy to understand what impact such a change could have.
Bluesky's development, on the other hand, revolves around a single entity that holds decision-making power over the protocol, running the server and developing the app and APIs that dominate that ecosystem. And it's based on a business model that was already defined well before the protocol was created, with a protocol that was also developed with a business model in mind!

The developers of the Fediverse were therefore Darwinianly selected by circumstances and today appear to be a bit more hacker-like, a bit more experimental, more adept at circumventing limitations, and (this isn't always a good thing...) more oriented toward community-driven financial support (and self-driven, because luckily for them, they all have IT jobs in a company). Moreover, not everyone is highly knowledgeable about the Activitypub protocol. And some of them are real "gourmets" of controversy...
Bluesky developers, on the other hand, seem decidedly more "secular" to me; they also have to deal with a more rigid protocol (definitely more protocol-based than the Activitypub protocol), strong centralized decision-making power, and objectively have more limited room for manoeuvre. Furthermore, these developers' livelihood model isn't clear to me (I mean, beyond their IT jobs at some company: do they all work for Bluesky PBC?).
I don't know... they seem like two worlds that aren't easily compatible, even from a social perspective...

If this is true, then it shouldn't be surprising if the attitude of Bluesky stakeholders (the real ones, those sitting at the top) is positive only when a change could benefit their business model.
Conversely, the responses from Fediverse stakeholders (i.e., those dozen or so de facto influencers who, with a nod, can determine the public's favor or hostility toward an initiative) might seem more disappointing.

I reiterate that even if I were right, this attitude wouldn't stop the new ideas germinating between the Fediverse and Bluesky developers.
It must be said, however, that since Bluesky was launched, I haven't seen any particular innovations. Recently, however, I've seen several new ideas emerge in the Fediverse, and these ideas, despite the rapidly declining user base, have led to very promising developments in the federated ecosystem over the past two years.

All of this, however, would explain the communication difficulties between the two worlds, linked to the fact that the Bluesky leadership is too high-flying and the Fediverse stakeholders are too free-wheeling.

I hope I've explained myself better, despite the language barrier.

julian said in I also want to see #activitypub get some of the primitives that #atproto has such as decentralized identifiers (except for real), personal data stores, content addresses, etc.:
> This is an important observation we should take into consideration.

My theory, however, would explain this reaction...

See you soon and have a good evening.

Bluesky's role in the ATmosphere is somewhat analogous to Mastodon's in the Fediverse back in 2022: it's the biggest and best-known app, and gets almost all the press, but isn't the main source of innovation.

So when I look at the interesting stuff happening in the ATmosphere, it's Blacksky (creating intercommunal software with a strong mutual aid focus), Germ (E2EE messaging integrated with ATProto accounts), Aendra's XBlock labeler, Anisota (a reallllly different UI for microblogging), microcosm (refactoring the underlying architecture), the feed that Graze put together for the New York mayor's election ... etc etc. They're all experimental, adept at circumventing limitations, community-focused, self-driven. None of them are paid by Bluesky.

But just as people outside fedi tend to reduce it to Mastodon, people outside the ATmosphere tend to reduce it to Bluesky. And just as that leads people outside fedi to overlook all the interesting and promising things here (and irritates and marginalized everybody where who's working on stuff other than Mastodon), people in fedi wind up overlooking all the interesting and promising things happening in the ATmosphere (and irritate and marginalize everybody there who's working on stuff other than Bluesky). And that specifically includes some of the high-profile fedi influencers who have been especially hostile to the ATmosphere.

As for Bluesky PBC, no of course I don't trust them. They're a venture-funded startup, of course they're going to put their own interests first; they've shown they don't prioritize safety or community; etc etc etc. Then again, back in 2022 Mastodon gGmbH's was completely controlled by a BDFL who had a track record of putting Mastodon's instances first and not prioritizing safety and community. But just as Mastodon gGmbH's issues weren't the only thing going on in fedi as a whole back then, I don't see Bluesky PBC's issues as the only thing going on in the ATmosphere these days.

(Of course, Mastodon's recent organizational changes are very encouraging, and hopefully they'll start prioritizing safety and community and working to advance fedi's interests as a whole. Still, that hasn't happened yet, and they're still by far the largest platform and user base ... so let's not go patting ourselves on our collective back too hard on this front yet!)

@informapirata

Back to @captaincalliope.blue's original pointsin this thread:

I want a new flagship #atproto app that isn't a Twitter clone on the surface.

Indeed -- Rudy Fraser talked a lot about this in his ATmosphereConf presentation. And the same is true here in fedi, where Mastodon is still dominant. I think of this in a couple dimensions

• getting beyond microblogging, but still skeumorphic to well-known centralized systems (Pixelfed and Lemmy here; Flashes, Skylight, and Streamplace in the ATmosphere). If done well, that's valiuable in terms of getting people to the ecosystem (from an activism perspective, the lack of a skeumorphic approximation to Facebook groups is a huge barrier) but my guess this is still likely to have somewhat limited impact. When was the last time when a better skeumorphic app ever really caught on and displaced an incumbent>

• less-skeumorphic software is where it really gets exciting. There's a lot of momentum here -- Bonfire, Bandwagon, Piefed are three good examples here in fedi, the stuff Blacksky is working in the ATmosphere -- but it's very hard to predict what will and won't catch on.

"I also want to see #activitypub get some of the primitives that #atproto has such as decentralized identifiers (except for real), personal data stores, content addresses, etc.

I want to see both protocols cross-pollinate with each other's strengths. And perhaps share infrastructure like identities."

Yeah, totally agree. I think t here's been some cross-pollination in both directions (Blacksky's local-only posts were directly inspired by Hometown, Mastodon's FASP and Fedisovery are somewhat influenced by AT's Relay-based architecture) but there's certainly a lot of room for improvement.

curious how you imagine adding did-style identities + personal data stores to activitypub without breaking existing servers/clients – do you see it as an extension layer (like webfinger 2.0) or a more radical protocol rev?

Your analysis is very accurate and objectively correct.. However, there is an analogy that is not entirely relevant.

thenexusofprivacy@infosec.exchange said in I also want to see #activitypub get some of the primitives that #atproto has such as decentralized identifiers (except for real), personal data stores, content addresses, etc.:
> Bluesky's role in the ATmosphere is somewhat analogous to Mastodon's in the Fediverse back in 2022: it's the biggest and best-known app, and gets almost all the press, but isn't the main source of innovation.

In fact, 2022 doesn't correspond so precisely with the year 2025 for Bluesky.
Mastodon was six years old in 2022, and in those six years it gained its leadership by competing with equally structured or even much more structured alternative platforms like diaspora*, Friendica, Misskey, and, more recently, Lemmy, who proposed a different interaction model.
Throughout that period, Mastodon was never considered a leader by its "competitors," and often not even a point of reference.
This led to an evolution of the Fediverse ecosystem that was much more influenced by natural selection, and thanks to this natural selection and also thanks to the many dead branches, there is so much diversity.

I'm fairly certain that this diversity is the fundamental asset of the Fediverse, and that this asset will hardly ever be possessed by the Bluesky ecosystem.

Agreed that Mastodon 2022 / Bluesky 2025 is far from a precise correspondence! But while fedi's ecosystem is very diverse in some ways (and I agree that's a fundamental asset), it's much less diverse in others. Similarly the ATmosphere ecosystem is very diverse in some ways.but much less diverse in others. So my guess is that most fedi projects have an analogue in the ATmosphere -- not as far along because it's newer, and we'll see how many continue and how many wind up as dead ends.

Of course there are plenty of exceptions. Off the top of my head, I can't think of ATmosphere analogies of Bonfire, Frequency, or the friendica/Hubzilla/(streams)/Forte family. Then again there are also ATmosphere projects where I can't think of any fedi analogy: Blacksky, Gerrm, XBlock labeler, Anisota, Semble (a sense-making app). And some of the ways the ecosystems aren't diverse mirror each other; for example both are very Europe/US/Canada-centric.

@informapirata

thenexusofprivacy@infosec.exchange said:
> less-skeumorphic software is where it really gets exciting.

Another great but just-launched example of this is Sidetrail, which is for taking a journey through some content: https://sidetrail.app/

It uses AT Protocol, but isn't microblogging, isn't photo sharing, isn't forums or link aggregation, isn't blogging, etc. It's something entirely new.

informapirata if you ask any ActivityPub software author about Mastodon's role in ActivityPub and standardisation, then you'll often see that ActivityPub software authors end up having to build for compatibility with Mastodon. There are certain quirks of Mastodon's specific implementation of ActivityPub that cause certain problems (I won't enumerate them here), but a very common trend right now is to see protocol implementation by lowest common denominator (i.e., compatibility with the largest application). Take for instance a new Music scrobbling service developer I was chatting with a few weeks ago, I asked him if he was going to use the Listen activity to federate out listens. His answer? No, I'm going to use Note so that people on Mastodon can follow what I'm listening to.

Another example of this, Pixelfed and Loops. Take a guess what object type they both federate out. Is it Image or Video? Nope. It's not. It's Note, once again for compatibility with Mastodon. See: https://blog.joinloops.org/loops-joins-the-fediverse/ (heading "Smart Content Representation")

Another fun fact: We all know that we use the @username@server.example syntax for mentioning people on the Fediverse. Take a guess where that is defined? I'll wait. Hint: It's not ActivityPub. That was Mastodon. Mastodon chose twitter style handles and used Webfinger to implement them. Webfinger is not actually part of the activitypub or social web specifications, it's completely separate. We have since had to write a document explaining all the quirky weird things needed to implement webfinger with ActivityPub, but it's inherently not part of ActivityPub: https://swicg.github.io/activitypub-webfinger/

(I also learned that ActivityPub does not actually have a real concept of "usernames" recently in: https://hachyderm.io/[thisismissem](/user/thisismissem%40activitypub.space)/115401128778866273 )

So is ActivityPub and Fediverse really not just controlled by the whims of Mastodon and Mastodon's legacy as the largest service in the Fediverse?

Also, technically the Decentralized Identifiers work is compatible with Actors in ActivityStreams 2.0, and that was a deliberate choice. Both use the alsoKnownAs property, and they use the same or an earlier version of the same security context in JSON-LD, and that was, I'm told, a deliberate design decision.

@oxpsi I don't think either require a radical protocol rev. ActivityPods already uses Solid as the personal data store, and as @thisismissem pointed out broad adoption of ActivityPub C2S (perhaps with some improvements) would move things in a very PDS-like direction. Similarly for identity Forte and Mitra

For identity there are various implementations within ActivityPub that separate identity from the hosting instance. And as @thisismissem says elsewhere in this thread

"Also, technically the Decentralized Identifiers work is compatible with Actors in ActivityStreams 2.0, and that was a deliberate choice. Both use the alsoKnownAs property, and they use the same or an earlier version of the same security context in JSON-LD, and that was, I'm told, a deliberate design decision."

@thenexusofprivacy @oxpsi I've also quietly been working on a research project to explore exactly this topic, and figure out ways to be compatible whilst different.

thisismissem Let's say I'm very influenced by the fact that I started my instance administration journey with a Friendica server, later joined by a Lemmy server.

The Friendica and Lemmy developers have never been subservient to Mastodon.

And while the Friendica developers have always claimed a specific mission, which is to always seek maximum compatibility and connectivity with other systems, the Lemmy developers have always asserted their independence from Mastodon's solutions.

For the rest, you're right about the de facto dictatorship that Mastodon often exerts, especially over "younger" developers, but it's not a real form of control.

Furthermore, things are changing faster than they seem, thanks to the evolution of two platforms: WordPress on one hand and the one we're talking about now on the other.

Right, but to my knowledge both Lemmy and Friendica both use Webfinger (not sure when Friendica implemented this whether before / after Mastodon), but still, there's a lot of stuff in the fediverse that is down to Mastodon's implementation, rather than anything else.

My only bug bear with Lemmy is that communities and users are not distinct actors. Why not setup a community.server.example subdomain and server community actors from that instead of the main domain? Then you could reliably address communities from all software.