When #Bluesky says "Protocols, not platforms", they intent two things:
Grabbing people's attention by telling them what they want to hear.
Presenting the AT Protocol as an alternative to #ActivityPub to capitalize on the current hype around #Mastodon, the #Fediverse and decentralized social media in general.
The two protocols are not equal solutions for the same problem and, in fact, AT is not even a (communication) protocol to begin with.
Rather than starting a #Mastodon vs. #BlueSky war, is there any way to build a bridge between #ActivityPub and #ATProtocol? If both are open standards with roughly comparable functionality, I wouldn’t have thought this should be too difficult, with some limitations. Thoughts? #noxp
If #BlueSky is FOSS, can't we just start running our own servers? If #ATprotocol works, can't somebody design a bridge? Is anybody working on this? What's broken about what they're doing that's preventing this? As a noob who thinks he understands open source and decentralization generally, I don't get how/why they're limiting invites.
@mmasnick this is super helpful. But one question I still can't quite grasp.
In the #ATprotocol server space, do the server admins themselves have the ability to ban users or remove content regardless of and prior to the users choosing their own moderation filters?
I also can't tell if the BlueSky/ATprotocol space allows for admins also have something analogous to #Fediblocking users and content from other ATprotocol servers.
As I’m observing conversations and efforts to build tomorrow’s technology related to #activitypub and #atprotocol, I have concerns regarding how many non-white folx from marginalized and vulnerable communities will show up and for those who do, will they be welcomed?
Because tech development has a long history of the most privileged making decisions that maintain the status quo while inflicting harm on “others”
I decided to split out the #Lexicon parsing into a separate library. It needs some clean-up, better error handling, and tests, but it works. @activitypubblueskybridge
My $10 take on this #Mastodon vs #Bluesky dustup is: Maybe algorithms are a net good?
I've always scrolled chronologically—maniac that I am—so this isn't the issue I'm getting at. Looking at the unintended consequences of how design influences practice: I think I waste more time on Mastodon than I did on #Twitter because I'm seeing the same couple dozen popular (unarguably relevant) posts reboosted x10000.
And this seems to be a consequence of "People are the algorithm" design.
@csgordon The tricky thing about traditional engagement algorithms is that it typically involves giving the algorithm writer as much of your data as possible. On most of Big Social that includes things like "likes", comments, and shares but also lots of "soft data." Like on TikTok, for example, where the amount of time you watch a video is taken into consideration of how much you may have liked that video.
From everything I've been able to tell, #BlueSky intends there to be an algorithmic layer where all sorts of algorithms can access data on any of the instances. Some of these algorithms may be nice and fun, and others may be intentionally rage-baiting people. And I don't think they have controls for users to say "hey don't include my stuff in the rage-bait algorithm but do include it in the cute puppies algorithm."
From my perspective, the #ATProtocol gives way too much of users' data to algorithm writers. It does not feel like much of an improvement on Big Social.
@BrianJohnson@codesmith The term that pops out to me when I read about #ATProtocol is some notion of a "marketplace for algorithms." Despite the homo economicus logic-jargon, if they can eventually enable users to make choices on how or which algorithms effect their feeds, they'll have design advantage.
@jalcine Honestly, I'm not even going to get mad about this. For me, the fediverse already won. I found some really cool people here, and I don't really care too much to connect with anyone who stayed behind on Twitter just to shitpost.
@jalcine I have zero interest in another profit-driven (1) platform with questionable terms of service (2) and lenient attitude towards moderation (3), and one that uses the term "decentralized" very generously, considering its single instance is invite-only.
My hot take on #Bluesky is that it is never going to be federated in practice.
Federation inherently brings considerable complexity - and already well-followed #Twitter users moving over are just seeking a wholesale anti-#Musk alternative.
By and large, they do not care about the benefits of federation (which is understandable)... and so there will be no demand for implementations of the #ATProtocol outside of the default, proprietary instance.
@jerry The AT protocol promises some improvements over ActivityPub that looks interesting and great in theory, like account and posts migration being built into it. The news of their approach to moderation so far has me worried though.
Until it's available for everyone to start hosting their own servers, tinker and stress test it I remain skeptical. I'm very curious to see how it turns out when (or if...) it's released in the wild though.
If #bluesky and the #atprotocol fail to fulfill their promises, I predict there will be another. And another after that one. All funded and pushed by rich bastards with lots of connections. And we little people who have pushed the #fediverse for years will all be expected to give them our attention and not be critical. All while this confusion slows adoption of #ActivityPub by journalists and governments and developers and the decentralization moment passes yet again.
Now if someone could create an interconnect of some sort between #ActivityPub and the #ATProtocol, at least then we won't have yet more fragmentation of communities :(