Would you like to see full default interoperability between #ATproto and #ActivityPub without a bridge? (what's preventing this...)
@wjmaggos That doesn't make any sense; it's like saying that you want Chinese and Spanish to be mutually intelligible. You could try to combine the two languages into one mixed thing, but that resulting language would be neither Chinese nor Spanish.
ok. so would the closest to this be getting apps to do the bridging?
of course I used to ask both threads and bluesky to do this, until my Instagram account got suspended. they never told me why but I assume they considered me spam. bluesky marks me as spam sometimes.
@evan @wjmaggos you're making a category error, Evan.
People who aren't us don't give even the tiniest shits about which protocols they're using. 5G phones happily talk to 2G phones over SS7, and the only thing anyone knows about "5G" is that it's "fast" – and they don't even know what that means.
You just said approx. "that doesn't make any sense; it's like saying that you want NG-RAN and SS7 to be mutually intelligible" and people will care about as much as someone who literally says that.
@evan @wjmaggos (i.e. not at all)
Mastodon and BlueSky are different, culturally, and so are the Fediverse and Mastodon and any combination of separate servers that you'd like to imagine. It makes plenty sense to make BlueSky and Mastodon interoperable. Doing so is, frankly, easy, and much easier than the very common approach of translating between Spanish and Chinese (which is what we normally do to communicate instead of inventing a Spanish/Chinese pidgin, which is also a thing humans do!).
@evan @wjmaggos where the real "human-level" challenge is is that e.g. Mastodon and BlueSky are different, culturally, and so are "the Fediverse" and "Truth Social" and any combination of separate servers that we can imagine, or even two random servers *within* the Mastodon-running-Fediverse. Mastodon and Bluesky are way more similar than Pixelfed, and it makes much less sense to federate Pixelfed and Mastodon than it does Mastodon and Bluesky, because they serve different social purposes.
but that assumes there won't be any appreciable experiential diff. at this point, I think there will be. I don't expect anything like fully independent news or government servers on AT.
see @mondoweiss trying to use AT and consider pressure from Israel on most relays. their ability to get a critical story to go viral would be crushed.
arguing over protocols should be focused on why we care about decentralization. which model will be most likely to get us the future we want.
@evan @wjmaggos it's not my message. My message to technologists was "the key piece of interop for humans is identity" and I was told "webfinger isn't part of activitypub" and "domain names are how we do identity on the internet" (despite the idea that paying $25/year to get to manage DNS isn't an end-user-achievable thing in the world we live in) and after years of this I gave up. It makes me sad that Bluesky went with domain names because consistent format really matters, but 🤷♂️
@evan @wjmaggos the actual thing here is that it's *our* message. I don't care about my message, I feel pretty confident that in the very long term, my understanding is what we'll end up doing, and I can't make horses drink. ❤️
(All of this is actually in my Atmosphere talk: https://www.youtube.com/watch?v=d-H1nzWHLoI BUT ALSO I have an idea I'm fleshing out rn that might help explain this)
so don't we have to propose visions and see which ones make the most sense, are most popular, are compatible?
what's special about social media is the boost. virality. not community or the public square. making it decentralized without algos or ads puts the people fully in control. collectively we can determine what info, ideas and art gets the most attention. I think that's the revolution @rabble talks about. I call it #DemocracyOfReach.
which protocol makes that most likely?
@evan @wjmaggos so was I. 😅 I don't do standards stuff anymore because I can't make "technologist horses" drink the "social purpose of technology" water.
(to be fair, I feel SO HAPPY at my job because my coworkers are great at communicating with the right people (most of whom aren't "standards-level" technologists) about these things)
@blaine @wjmaggos The first question asked by the poll is "Would you like to see full default interoperability between #ATproto and #ActivityPub without a bridge?"
The second question is "what's preventing this".
@evan @wjmaggos you're reading it as a technologist – I agree with you that making atproto and activitypub literally interoperable is ... weird (although, I have argued many times that I think the two protocols will converge in functionality and scope, and also separately that lens-based translation of data structures is basically the best thing and would smooth right over the lexicon/AS differences if we had good tooling).
@evan @wjmaggos ... but that's not what the question is asking. It's intent, and the #atproto and #activitypub in the question aren't referring to specifications, they're referring to communities.
@wjmaggos @evan I think it was perfect. This whole conversation has lit a fire under my butt. Partially because any confusion really highlights this challenge we have in this space, where we're trying to navigate social problems with technological solutions, and one of the big barriers is that *that* means getting technologists to see beyond the bits AND getting "everyone else" to care about the technology.
For fun, we can think about three different protocols in the way that they function. ActivityPub, atproto, and nostr.
Nostr would be the most decentralized and most individualist. You don't even pick a single server, you pick on number of different relays which will accept your messages and provide messages to you. It really doesn't matter if in the end which individual relays you pick because in practice it's just a ledger with all the messages that it received, and the protocol itself handles identity through your secret key. If the relay that you were using goes down, your user experience doesn't even notice because there's probably 10 others.
ATproto would be the least decentralized and most collective. It is hypothetically possible to host your own instance, but in practice user management and a lot of other stuff is Central to the main Bluesky organization. Getting banned or getting blocked or whatever, it's not that different from Facebook in that regard. If the main Bluesky service goes down, it will effectively mean the end of bluesky.
ActivityPub would be somewhere in between. You have individual servers that people will pick one or multiple, there is a centralized point where your identity lives, and each server has its own moderation policies and administrator team. If one server goes down, everyone on that server loses access to the fediverse on that server and they also lose their identity from that server, but they can very easily go somewhere else. If mastodon.social goes down, a lot of accounts will become inaccessible but the broader fediverse will be unaffected.
Bridges are obviously possible between the three because we see it, but I tend to think that the three are mutually exclusive and mutually incompatible in their aims and technical details such that integrating any two immediately means giving up some of what that protocol is trying to do.
@sj_zero @evan @wjmaggos I would modify the last bit to "I tend to think that the technical communities designing the three protocols have different aims and integrating any two immediately means giving up some of what the creators of that protocol is trying to do."
I hold the goals of the wider communities using these products WAY over what the goals of the individual technologists designing these protocols. Does the protocol serve the community's goals? Great! If not, 🤷♂️
but I also care what either of them becoming dominant might mean for the larger culture. I don't see the media decentralization we need (obvious under the Trump admin) as likely if AT "wins". but maybe I'm wrong. but unless we have a goal like that and not just decentralization, I don't know why the argument matters except for technologists. it has to something concrete like that. and things like enabling community is fine but that's available in other ways etc.
@wjmaggos purely personally, I don’t see significant value from this.
More people to follow? Don’t care, I don’t follow everyone on mastodon, I’m already missing out gems.
More people follow me? I don’t care, my account isn’t used to drive any economical value.
I can see there being frustrations as protocols fail to work with each other. Or as funded tries to absorb ‘free’.
Or culture shifting towards more hustle due to the nature of ATp being more about money generation.
to me, the point of social media is virality. the boosts. having the best info, ideas and art spread most widely. it's why cities generally improved the human experience. for that, I want everyone involved.
@wjmaggos I don’t think a bridge is necessary for that though. As it’s not that APp is exclusionary from a technical standpoint (although it is socially).
My worry is that a bridge will not result in best info and ideas. Instead it will homogenize ideas around a local peak of ‘best’.
With separation we have two local peaks of best and sometimes one side is better, sometimes the other. But we can learn from each other.
sorry, best is shorthand. there will be multiple very goods. but I want the people who disagree to engage each other, not be on different services. I really wish we had reddit style comment ratings. I don't like that we're doing the separation via politics. our moderation could be better.
Eg, Vancouver sky train developed after seeing the nightmare of highway traffic in LA.
A local decision can be made based on learning from another. In one mega city, that be much harder to do due to increasing layers of bureaucracy
@chris @blaine @wjmaggos it's funny that you say "without bridges" and "however that may happen" in the same breath. If we don't care how it happens, why rule out one particularly powerful technique for interoperability?
BridgyFed works GREAT. It's fucking amazing. Is there room for improvement? Absolutely. But it's a really good service.
@blaine @wjmaggos @evan I'm clinging to “decentralization” as part of my answer to the “What's this Fedimoose thing you keep talking about and why should I care?” because the best explanation I have found is "social networks suck, except for one social network that's been working OK for 40 years, namely email. Why doesn't email suck? Because it's decentralized, nobody owns it."
so just use email. /snark
sorry I'm trying to spark the deep conversation about why we should care about social media decentralization. what does it provide us? what do we imagine it could provide us? is there a difference between social networking and social media?
but I've made my case.
@evan @blaine @wjmaggos One possible problem with that is that even if that happened, this would mean that when one random person follows you from Bluesky, all your future posts would become visible in the open on a public website and the global firehose and to the 30M+ Bluesky users. From what I know about Mastodon users, a lot of them would probably not like that… (since a lot of them very loudly objected specifically to that when Snarfed wanted Bridgy to work this way at first).
another question related to the why discussion.
is this a public medium like blogging or is the goal more private, your chosen community and conversations? the former I call social media vs the latter being social networking. old school twitter vs the friends, family and groups on Facebook.
but don't we have other more closed systems that allow us to choose community. the web provides free speech to everyone. revolutionary. the social web adds an attention layer to that where everyone gets a say. another revolution.
we need other stuff also but this potential is something I don't want to squander, esp when I don't think it means being unable to do the other stuff also.
I go back to the Iraq War and knowing elite info control led to that disaster.
@wjmaggos @evan @blaine @rabble the web can provide access control too! the real "social web" is less about attention to me and more about letting people participate across websites. in that regard, a standard for identity is the most important missing piece. (consider hypothetically, we could expand on WWW-Authenticate and Authorization headers to negotiate identity for an individual HTTP request.)
too technical for me. all I know is I can boost a really important story I find and it can go viral and change the narrative about an important issue. even if very wealthy/powerful don't want it too. or at least that would be possible if everybody was here. and even moreso if fedi was a more important part of how we all handled news instead of outlets with owners with merger and political concerns.
@wjmaggos BlueSky is driven by profit, while Mastodon and other ActivityPub spaces (for the most part) are driven by the people.
their business model concerns me, absolutely.
@wjmaggos @chris @blaine one of my big problems with it too. I think @anewsocial are aware of that limitation and are working on some solutions.
@evan @wjmaggos @chris @blaine @anewsocial
Appreciate all the kind words in the thread! That said, I still strongly believe opt-in is a feature, not a bug. The right solution isn't to force people onto networks they don't want to be a part of; it's to clean up the rough edges of unbridged interactions and content so it makes less of a difference.
@evan @wjmaggos @chris @blaine @anewsocial
I understand the instinct of "make these work together", but there are technical and cultural differences that make it more than just two standards talking to each other. A couple of simple examples are "I don't want to be in a public, auditable firehose" or "I had a bad experience there, I don't trust them with my safety."
@quillmatiq @evan @chris @blaine @anewsocial
I understand if you don't want to have the conversation but these are public systems like blogging with your email address available, we just don't think of them like that. it's old school twitter, not even your network of friends and family on facebook. not having algos/ads doesn't mean everything isn't boostable to the world by default. maybe people want to use something designed to be inherently private, but that's not fedi etc. IMHO
@wjmaggos @evan @chris @blaine @anewsocial
I agree with you there! But often it's not about *how* public things are, it's more about *where* they're public. In other words, many users I've spoken to who are either not bridged or are straight-up anti-bridge have chosen that route because they don't (yet) trust the entities on the other side to have that much ownership over their data. There's also a cohort who don't want to have to depend on *us* to bridge them either.
It's a lot more complex!
@quillmatiq @evan @chris @blaine @anewsocial
they're public on the internet! screenshots?
I think the complexity is mostly about having so few people here and thinking it will stay that way. that being here is safety through obscurity. that a fedi server is a community instead of much more a portal to a possibly universal social media network. they built and maintained this place (huge kudos!) to get away, but actually made the best network not to.
they need a fork that's default opt in.
@quillmatiq @evan @chris @blaine @anewsocial
same with the moderation.
preventing people from being able to tag you with a slur etc will lead to growth, but preventing people from being able to state their unpopular opinion to no in particular will isolate us. I don't think people choose these approaches to keep us small, they just don't want to stumble on stuff that will upset them. they want this place cozy.
but we have to do that browsing the web. it's essential to civics.
sorry. /rant
@wjmaggos before I vote, what does this mean technically? The way I understand it is that AP is message passing and ATProto is shared heap, how would they become interoperable without a bridge unless you change the specs of one or both of them?
Minds implemented ActivityPub and nostr support, not much different here I'd suspect.
- replies
- 1
- announces
- 0
- likes
- 1