Jack Dorsey skipped ActivityPub, built AtProto, lost Twitter, funded Bluesky, watched it become a company with VCs and a board, said it was "repeating all the mistakes," left, and now funds Nostr.
-
Decentralization isn't supposed to make things easier for the people using it. It's not supposed to be a better social "app." That's not the point. The whole reason for decentralization is to prevent admin abuse. You put up with a little more hassle as a user, and when the admin sells you out to Nazis, you'll be ready to adapt. Then sellouts don't take over the network, and nobody gets their elections rigged in favor of some tyrannical monster, or whatever.
Criticizing Activitypub for having an optional server that has too many people on it is fine, but you can't equate that to a network run by crummy venture capitalists who worked for Twitter, that won't function without permission from one central authority.
CC: @mastodonmigration@mastodon.online @baralheia@dragonchat.org@cy @mastodonmigration @baralheia @thisismissem Nobody at Bluesky worked at Twitter
-
@cy @mastodonmigration @baralheia @thisismissem Nobody at Bluesky worked at Twitter
-
@thisismissem @mastodonmigration @baralheia The problem is money. It's not cheap to run that network. You have personal experience of how hard it is to squeeze money out for important social-media work. Who's going to pay to keep it on the air?
@timbray @mastodonmigration @baralheia but that's a universal problem no matter the protocol. The fediverse struggles so much for funding. AT Protocol projects too struggle for funding. As for Bluesky, I think it's too early to say. There could be things they successfully monetize to help make them sustainable without more investment, only time will tell.
Who pays to keep all the fediverse servers online, develop the projects, etc. i think adding up the cumulative costs across operators would be quite insightful, because many a fediverse server has shutdown due to money.
-
@baralheia @thisismissem @mastodonmigration relays: I think this is more or less complete: https://compare.hose.cam, though I think it's missing these new ones: https://sri.leaflet.pub/3mddrqk5ays27.
I've recently looked at which of them really cover the whole network, I'm working on setting up a website with live stats on that: https://bsky.app/profile/mackuba.eu/post/3mdhbbocmrc26
AppViews: for Bluesky microblogging I think right now there's only Bluesky's and Blacksky's that are live & public.
@mackuba @baralheia @mastodonmigration neat. We just added a proper section on Relays to the new AT Protocol website (see my post on Bluesky about it) and we included an abbreviated list there.
-
@evan @boris @reflex @dansup @quillmatiq will you forgive me cos I asked Gemini
: Destroy as suitable. Under dependency challenges, it says:
- Identity Dependency: did:plc directory Bsky owned
- "Centralized Indexing: users can host their own PDS, but rely on "relays" to discover other users. Currently, the main relay is operated by Bky. Replacing this requires significant compute power."
- "Atproto's adoption depends on it having a "killer app" other than the initial microblogging client"@sheislaurence @evan @boris @reflex @dansup @quillmatiq
the strategy seems pretty clear based on how the protocol works and VC strategies we've seen before. survive and grow via VC until AT is accepted as the open social protocol. till everybody thinks that's the one to build on. have bluesky be to AT what google is to HTTP. an open protocol wasn't enough then either.
I will complain until I see their "unfair" advantage (imo) end and we know how they plan to provide an ROI to their investors.
-
@sheislaurence @evan @boris @reflex @dansup @quillmatiq
the strategy seems pretty clear based on how the protocol works and VC strategies we've seen before. survive and grow via VC until AT is accepted as the open social protocol. till everybody thinks that's the one to build on. have bluesky be to AT what google is to HTTP. an open protocol wasn't enough then either.
I will complain until I see their "unfair" advantage (imo) end and we know how they plan to provide an ROI to their investors.
@wjmaggos @sheislaurence @evan @boris @dansup @quillmatiq That last part especially, they won't even say who their investors are at this point.
-
@wjmaggos @sheislaurence @evan @boris @dansup @quillmatiq That last part especially, they won't even say who their investors are at this point.
@reflex @wjmaggos @sheislaurence @dansup @quillmatiq I don't think that's true. They're on CrunchBase.
-
@reflex @wjmaggos @sheislaurence @dansup @quillmatiq I don't think that's true. They're on CrunchBase.
@evan please remove me from replies, William Maggos is a troll who spreads misinfo & is generally unkind who I have long blocked (yes I understand you’re pushing back against his misinfo)
(These thread canoes with a general tendency to not trim reply mentions in many clients is not great)
-
@reflex @wjmaggos @sheislaurence @dansup @quillmatiq I don't think that's true. They're on CrunchBase.
@evan @wjmaggos @sheislaurence @boris @dansup @quillmatiq Does it have the results of the latest funding round last year because they've been silent about that? People keep asking and getting no answers. I can't see the funding data on CrunchBase, perhaps you can?
X competitor Bluesky is being valued at around $700 million in a new funding round after explosive growth in the wake of Trump's victory
Bluesky is raising new funding led by Bain Capital Ventures that would value the social media company at around $700 million, according to sources.
Business Insider (www.businessinsider.com)
-
@reflex @wjmaggos @sheislaurence @dansup @quillmatiq I don't think that's true. They're on CrunchBase.
@evan @reflex @wjmaggos @sheislaurence @dansup @quillmatiq
Evan, it is not at all clear who owns Bluesky, or even how much money they have raised and from whom.
More about the mystery here...
Mastodon Migration (@mastodonmigration@mastodon.online)
Who owns Bluesky? The curious mystery of the Bluesky Series B funding round. Did it happen? If you look at pitchbook (https://pitchbook.com/profiles/company/484831-81) it seems Bluesky closed a whopping $97 million funding round resulting in a $700M valuation in January 2025. VCpedia lists Greylock, alumni ventures and Skyseed as participants (https://vcpedia.com/rounds/5195). The interesting thing is there are no press releases or other media coverage confirming this financing. Read on... 1/2 #Bluesky #WhoOwnsBluesky
Mastodon (mastodon.online)
-
@evan @reflex @wjmaggos @sheislaurence @dansup @quillmatiq
Evan, it is not at all clear who owns Bluesky, or even how much money they have raised and from whom.
More about the mystery here...
Mastodon Migration (@mastodonmigration@mastodon.online)
Who owns Bluesky? The curious mystery of the Bluesky Series B funding round. Did it happen? If you look at pitchbook (https://pitchbook.com/profiles/company/484831-81) it seems Bluesky closed a whopping $97 million funding round resulting in a $700M valuation in January 2025. VCpedia lists Greylock, alumni ventures and Skyseed as participants (https://vcpedia.com/rounds/5195). The interesting thing is there are no press releases or other media coverage confirming this financing. Read on... 1/2 #Bluesky #WhoOwnsBluesky
Mastodon (mastodon.online)
@mastodonmigration thanks! I had heard there was another round in the works, but I didn't know the details. I appreciate the detective work.
-
@evan @reflex @wjmaggos @sheislaurence @dansup @quillmatiq
Evan, it is not at all clear who owns Bluesky, or even how much money they have raised and from whom.
More about the mystery here...
Mastodon Migration (@mastodonmigration@mastodon.online)
Who owns Bluesky? The curious mystery of the Bluesky Series B funding round. Did it happen? If you look at pitchbook (https://pitchbook.com/profiles/company/484831-81) it seems Bluesky closed a whopping $97 million funding round resulting in a $700M valuation in January 2025. VCpedia lists Greylock, alumni ventures and Skyseed as participants (https://vcpedia.com/rounds/5195). The interesting thing is there are no press releases or other media coverage confirming this financing. Read on... 1/2 #Bluesky #WhoOwnsBluesky
Mastodon (mastodon.online)
@mastodonmigration @evan @reflex @wjmaggos @dansup @quillmatiq it's interesting that the #transparency report #Bluesky posted less than a month ago doesn't mention anything about investors. Having personally worked in the transparency sector, it is the first time I see a company suggest the word doesn't relate to financial transparency 🫣. https://bsky.social/about/blog/01-29-2026-transparency-report-2025
-
@thisismissem @mastodonmigration @baralheia @cwebber
I would argue that neither the AP nor atproto are built directly for patterns of human communication that exist in the real world.
The "everyone messages everyone else and must have access to any message from anyone at any time" pattern embodied by the present Bluesky is not a real human way of communicating and forming communities, it is a figment of tech companies' imaginations and represents a massive amount of over-indexing that relies on, and therefore tends towards, centralized platforms.
The "everyone preferentially messages people in their nearby vicinity but sometimes people further away" view embodied by most present Fediverse software assumes a flat social network in which "non-local" is functionally the same in all cases and does not model human social networks very well.
AP assumes you are building bunch of villages with a flat road network between them. atproto assumes you are building Saudi Arabia's The Line.
@ricci @thisismissem @mastodonmigration @baralheia @cwebber
The atproto model is in the lineage of the web 1.0: everyone has their own website and google indexes them all for everyone to "see" the network. Wherever Google, and the tech industry following its steps, went is the exact direction bluesky is going to go.
I'd really love to know why you think the AP model doesn't map human societies ? Maybe a concentric model of trust, from closer to larger, is something you have in mind ? -
@ricci @thisismissem @mastodonmigration @baralheia @cwebber
The atproto model is in the lineage of the web 1.0: everyone has their own website and google indexes them all for everyone to "see" the network. Wherever Google, and the tech industry following its steps, went is the exact direction bluesky is going to go.
I'd really love to know why you think the AP model doesn't map human societies ? Maybe a concentric model of trust, from closer to larger, is something you have in mind ?@rakoo @ricci AP as implemented places you on a server which is your identity, that server is a specific vertical of a online social presence (microblogging, images, videos, short videos, articles, forums, link aggregator)
The AP C2S model separates to a degree the identity from the application. You do still only have one social graph and inbox/outbox, so it's not ideal, most people have different social groups on different verticals of platforms.
But as long as AP is deployed in the topology and systems it is today, it does not do the "thing" that people do socially.
Mastodon doesn't give you a "community" just because you're on the same server (no local only posting, local feed is too noisy on larger servers), Loops arguably removes all local community thanks to algorithmic feed – I don't think they've a local feed that I've seen in press.
AT Protocol makes getting into social spaces in different verticals easy. Conceptually AP C2S is very similar: you have a place that is your identity + data, and then you join places with that identity (maybe customising the identity or social graph for that vertical application)
-
@rakoo @ricci AP as implemented places you on a server which is your identity, that server is a specific vertical of a online social presence (microblogging, images, videos, short videos, articles, forums, link aggregator)
The AP C2S model separates to a degree the identity from the application. You do still only have one social graph and inbox/outbox, so it's not ideal, most people have different social groups on different verticals of platforms.
But as long as AP is deployed in the topology and systems it is today, it does not do the "thing" that people do socially.
Mastodon doesn't give you a "community" just because you're on the same server (no local only posting, local feed is too noisy on larger servers), Loops arguably removes all local community thanks to algorithmic feed – I don't think they've a local feed that I've seen in press.
AT Protocol makes getting into social spaces in different verticals easy. Conceptually AP C2S is very similar: you have a place that is your identity + data, and then you join places with that identity (maybe customising the identity or social graph for that vertical application)
@rakoo @ricci have a read of Lauren's article: https://connectedplaces.online/where-does-community-live/
Yes, community on AT Protocol is a nascent concept still, but the separation of identity + data from applications makes it possible to experiment and have one social graph or many.
One project doing community spaces on AT Protocol is: https://github.com/collectivesocial/open-social
-
@ricci @thisismissem @mastodonmigration @baralheia @cwebber
The atproto model is in the lineage of the web 1.0: everyone has their own website and google indexes them all for everyone to "see" the network. Wherever Google, and the tech industry following its steps, went is the exact direction bluesky is going to go.
I'd really love to know why you think the AP model doesn't map human societies ? Maybe a concentric model of trust, from closer to larger, is something you have in mind ?@rakoo @baralheia @thisismissem @mastodonmigration @cwebber
Yeah great question! It's that everything past the local level is flat from a network/protocol level - all communities are 'equidistant' at the network layer, which isn't how it works for human communication and society.
So I'm agreeing with your point about circles of trust, but down a layer at the protocol - and I don't think it's an accident that Mastodon and other fedi software have not really gone very far in implementing such things given that - while it's certainly possible - it's not inherent in AP.
But yeah I think AP is far *closer* to how humans actually communicate than atproto
-
@rakoo @ricci have a read of Lauren's article: https://connectedplaces.online/where-does-community-live/
Yes, community on AT Protocol is a nascent concept still, but the separation of identity + data from applications makes it possible to experiment and have one social graph or many.
One project doing community spaces on AT Protocol is: https://github.com/collectivesocial/open-social
@thisismissem @ricci
yes, if we're looking at mastodon and the model it has created that all microblogging apps have copied, then community doesn't really exist in the technical parts but must be artificially built up. The more interesting example is the threadiverse where communities are literal spaces: people congregate towards one or any number, they are independent from your server and from your identity. This, to me, feels closer to how communities start to create: pick an obvious topic, make obvious-y rules about what is on-topic or not to guide what people can talk about, then possibly graduate from there to another form (maybe a specific, closed community with your people). I do think more visibility should be given to the threadiverse rather than microblogging, or even mastodon, because of all the problems you have listed. And the future direction of AP should definitely split the server from the usage and build apps on the client only ! -
@thisismissem @ricci
yes, if we're looking at mastodon and the model it has created that all microblogging apps have copied, then community doesn't really exist in the technical parts but must be artificially built up. The more interesting example is the threadiverse where communities are literal spaces: people congregate towards one or any number, they are independent from your server and from your identity. This, to me, feels closer to how communities start to create: pick an obvious topic, make obvious-y rules about what is on-topic or not to guide what people can talk about, then possibly graduate from there to another form (maybe a specific, closed community with your people). I do think more visibility should be given to the threadiverse rather than microblogging, or even mastodon, because of all the problems you have listed. And the future direction of AP should definitely split the server from the usage and build apps on the client only !@rakoo @ricci @thisismissem this makes the most sense to me. I think "we" on the AP have a hard time with this because we alternate between servers describing themselves as neutral providers a la email or already being community focused (like the Indieweb server I'm on).
PS by the Threadiverse do you mean Threads and some other assortment of apps?
I think the way Laurens described reddit as
-
@rakoo @ricci @thisismissem this makes the most sense to me. I think "we" on the AP have a hard time with this because we alternate between servers describing themselves as neutral providers a la email or already being community focused (like the Indieweb server I'm on).
PS by the Threadiverse do you mean Threads and some other assortment of apps?
I think the way Laurens described reddit as
@McNeely @ricci @thisismissem No, it's the reddit-like: lemmy, mbin, piefed, nodebb and even discourse. Basically forums -
@thisismissem @ricci
yes, if we're looking at mastodon and the model it has created that all microblogging apps have copied, then community doesn't really exist in the technical parts but must be artificially built up. The more interesting example is the threadiverse where communities are literal spaces: people congregate towards one or any number, they are independent from your server and from your identity. This, to me, feels closer to how communities start to create: pick an obvious topic, make obvious-y rules about what is on-topic or not to guide what people can talk about, then possibly graduate from there to another form (maybe a specific, closed community with your people). I do think more visibility should be given to the threadiverse rather than microblogging, or even mastodon, because of all the problems you have listed. And the future direction of AP should definitely split the server from the usage and build apps on the client only !@rakoo@blah.rako.space completely right.
The "community" aspect on microblog UI is shallow at best. Instance names and domains are signalling community, but you're still screaming into a public town square about anything and everything.
Threadiverse absolutely does it better, but the crossover between it and the wider fediverse is minimal at best (I am posting on NodeBB right now.)
I'm going to be talking about this next week at FediMTL!
FediMTL - Digital Sovereignty Conference
A Canadian conference on digital sovereignty and the social web - February 24, 2026
FediMTL (fedimtl.ca)
