Fun story for everyone with the AT Protocol trademark: the Mastodon trademark for many many years was owned just by Eugen personally, not by the Mastodon gGmbH, which is part of the reason they had to recently pay Eugen €1 million euros.
Said something interesting to a friend today when they asked about different open social web protocols. It might be a little bit controversial, but:
I think that ActivityPub is decentralized in principle, whereas AT Protocol is decentralized in practice. 🧵
Oh dear lord, I knew it was bad, but didn't know it was this bad: hackers.pub/@fedify/2026...
(I've mostly just contributed to activitypub projects that already existed, so didn't usually need to experience these specific pains)
W Social, never short on amusing failures: blog.elenarossini.com/w-social-a-b...
I'm pretty sure LinkedIn is just lying at this point, unless there's something I don't know?
It's kinda nice to see all my AT Protocol friends signing up on a server on the fediverse that I did a significant amount of work on (I used to be on the Hachyderm Infrastructure team)
It was really interesting playing with a beta of this over the weekend, and for research & information gathering on ATP it's a game changer
I think what I really want on articles from @standard.site on @bsky.app is a "add to reading list" with an app that emails me once a week with what I've saved to read. I don't want subscriptions. Maybe an integration with like standard reader? Or similar?
Whereas AT Protocol is decentralized in practice because your data is stored on your PDS, and you can always move your data, even without the PDS's cooperation, because the data is all signed by an identity that the PDS does not control or at least, can have alternative control.
In today's Fediverse, there are multiple servers running the exact same sort of software, and they share messages between each other. Which is great, but the problem is that your account and all your data is anchored to a specific server.
Min reviewers should probably also include agentic reviewers, but often code review is not just about code structure & style in just that code but how it relates to other code & projects and product roadmaps
Going to start trying to share little tech things I learn from time to time with this blog, so no long form, just short little snippets.
This "leaf computing" talk is super interesting! www.jeremiahlee.com/posts/leaf-c...
by @jeremiahlee.bsky.social (I think)
I just thought this was kind of interesting as a comparison between the two protocols. It is maybe a spicy take.
Personally, I would rather my data be portable and that I can use whatever apps I want, compared to having my account tied to a specific server.
Mark Zuckerberg once again being the leader in creepy non-consensual shit.
This doesn't make sense until you look at tokens as a financial instrument and open router as essentially an exchange or stock market for AI essentially (maybe more like a brokerage firm), in which case it fits with Stripe trying the same thing with crypto without explicitly becoming a C2C exchange
We should be tackling the harms instead of restricting people's access and freedom. Many teens are able to use social media today to find community where they otherwise couldn't, and that's important.
Depriving the next generation of social participation until they're adults isn't the way to go.
As far as applications go, for AT Protocol, we can of course host multiple copies of the applications, but we don't necessarily need to, it's a choice to. Applications can interact directly with the PDS, or they can work through a relay or some other shared infrastructure.
Wrote a little something about timelines and expectations. It's definitely a rambling this one.
I've just seen for the first time what @bad-example.com looks like in person, and somehow they look exactly like what I would expect from their work.
The thing about operating within a federated network is that you've gotta communicate with others, not just pretend you've been doing so.
(As Elena notes: this sort of rate limiting issue isn't unique to AT Protocol, but can be present on all federated networks)
Had something rare happen the other day: I showed someone a package I'd previously written and said it needed some updating to work with the latest framework version, and like two hours later they sent me a pull request & it's looking good
Addendum 2: "but what about did:plc??", well, firstly, it's a Swiss Association with multiple members (and everything's being moved there), secondly, there's did:web and it's fully supported (though lacks some of the identity proof features — hopefully we start supporting did:webvh in the future)