I have setup bskysocial.world to test that (there's no web interface, just select this domain when logging in or signing up via the app or bsky.app)

Note: This is for testing only, I can't promise it will remain running.

(I am @ruud.bskysocial.world)

Bluesky

Social media as it should be. Find your community among millions of users, unleash your creativity, and have some fun again.

^{Bluesky Social}

So, from up close it seems like people can have their own servers (i checked wurzelmann.at which is currently on the frontpage) but they do not seem to have their own frontend.

This indeed makes it so that for people to actually SEE your content you must federate with one entity and are controlled by them.

Imo this is very bad because it takes the freedom out of federation. Yes, you dont need to login to an app but if they ban you or defederate or delete your post, nobody will see it, right?

Please someone who has tried and gets the technical details shed light on this.

Bro just use twitter at this point.

That's like choosing gangrene on your foot because you can't decide which shoe to wear and tying the laces seems like a pain in the ass.

That's a wild interpretation of what they said.

Dude described a branding problem, not a technical problem.

I am always self conscious about my long winded replies because sometimes even I wouldn’t wanna read a post that long.

No need to be self-conscious - that's a concise account of a complicated issue, which is going to go long. Long posts become an issue when they are rambling and unfocused.

Thank you for all the info! Been wondering about various things.

Zooming back, it feels like both the fedi and bluesky aren't fully resolved yet and might influence each other and a possible further replacement.

For example, I like the PDS/repository idea and think it would help make migrating easier in the fedi. But I would like more granular control over what is shared to relays as opposed to everything.

Thanks for the explanation. Didn't realize Bluesky/AT is more like a fedi-washed version of ActivityPub rather than a real alternative ...

I'm not sure; on the one hand, I think the fact that federation has become a unique selling point in micro-blogging is indicating a positive trend; so even if people join Bluesky its good for the Fediverse. On the other hand, if federated just becomes another buzz word that means nothing at all, while places where the real innovation is happening are drowned out, the window of opportunity could just close.

Or maybe they do it because it would be very confusing for 99% of users to read “sign in with your atproto account” when no one knows what that is. Its a good thing theyre keeping things simple for now instead of getting the mastodon “its too complicated” issue and having non tech users not use it.

Mastodon and the activitypub culture is very much either “its too complicated” or “how do you not understand its so simple maybe you’re just dumb and shouldnt be here” energy and its why 99% of topics here are just tech and politics. Almost no artists use it and the few that use it are on mastodon.art which defederates from so many instances.

It does matter their reasoning and they can change the word when atproto is more understood. No fedi software uses “sign in with your activitypub account” either because confusing users is a bad thing.

I personally like that theyre doing things to appeal to non tech users because i dont want to use a service no one uses. Its why i dont use any activitypub service for my art because theres no artists using activitypub but bluesky has a huge art community even though they still dont have video

To complete the network, relays are required on ATProto and apparently could be expensive to host, so right now, it appears the only relay is hosted by BlueSky the company. Which makes things slightly centralized.

A number I've seen quoted multiple times now is ~$150 per month to host a relay (Source). Which explains why Whitewind, Smokesignal and Frontpage don't host one, they are mostly still small projects by individual talented devs, but imo if that number is true, it really doesn't seem too outlandish that someone might go for it.

Lemmy.world used to cost that much, but I think they downscaled a bit recently, or are at least planning to as the current growth of the userbase has slowed down.

I was actually surprised by that 150 figure when I first read it, as it is much cheaper than what the BlueSky documentation makes it sound.

It is certainly possible to collect that much in monthly donations, but then again... how do you build a loyal base of supporters for running a mostly hidden piece of infrastructure? People always complain about the instance focussed nature of the fediverse, but the ability to build communities around them and get people actually emotionally invested in their home instance is IMHO rather a strength of it. That is also why I am slightly sceptic of easy account migration tools, as it devalues the instance as yourhome base to a certain extend.

Yeah, the instance-focus aspect of the Fediverse is a good thing. You are a good example with slrpnk.net, but dbzer0, all the language/country-based instances (feddit.org, jlai.lu, aussie.zone, lemmy.ca, feddit.uk) also have their own culture and feeling. That's cool to see.

as it devalues the instance as yourhome base to a certain extend.

That's definitely a thing for generalist instances. I don't really think there is much of a lemm.ee culture for instance. Which is also okay, some people just want access to Lemmy without a strong instance identity.

I believe they're planning to change that into something like "Sign in with your ATmosphere account." (hopefully)

I could be wrong, but the term "ATmosphere" has been used widely by non-technical users.

Yeah, so it's not a bad explanation, maybe a bit biased ;)

The key is that the architecture is very different, and there isn't a direct equivalent of instances. There are PDSes, but they do much less than Fedi instances, and they also don't directly talk to (federate with) each other. The data flows from PDSes to relay(s) to AppView(s) and to clients.

There's a small number of self-hosted personal PDSes (blue.mackuba.eu/directory/pdse…), but the system isn't really open yet to running larger ones with open signup. The Bluesky Relay doesn't currently accept more than 10 users on one PDS (with exceptions like Bridgy).

Technically anyone can run a parallel Relay and/or AppView, and hopefully that will happen, but nobody has done it yet so far (Whitewind/Frontpage are kind of different services on the same protocol).

Last year, I moved conversations into Collections identified by the context element, and we only accepted Add/Remove activities for collections from Collection->attributedTo. All submissions were made to them and only to them.

We previously sent replies only to the thread originator - a concept for restricted access conversations I came up with about the same time the Diaspora folks did. We kept it for a number of years even though Mastodon didn't support it, but moving conversation threads into a simple collection management operation not only makes sense, it bloody works brilliantly and co-exists with microblogging. It also co-exists with FEP-7888.

>In a constrained conversation, the target->id and the context are identical.

So target.attributedTo and context.attributedTo would be identical too

The math here is beyond me, but this statement from the paper seems contradictory:

The obtained equation is covariant in space–time and invariant with respect to any Planck scale. Therefore, the constants of the universe can be reduced to only two quantities: Planck length and Planck time.

Planck time is derived from the speed of light and the gravitational constant. So wouldn't there be at least four universal constants?

Pop sci

The direct article: sciencedirect.com/science/arti… (Jan 2025)

Reddit chatter about it: reddit.com/r/Physics/comments/…

Might be LLM bunk. If you're consuming science news, then first: I recommend PBS Spacetime and second: if a quantum gravity was actually formulated, you'd hear about it there first. It might actually be exciting enough to make CNN.

More to the point though: this sorta thing is too good to be true. Plenty of things are, and are still real. But even still they bear a second glance. This one doesn't pass the sniff test.

It's utter bullshit from the very start. First, it isn't true that the Ricci curvature can be written as they do in eqn (1). Second, in eqn (2) the Einstein tensor (middle term) cannot be replaced by the Ricci tensor (right-hand term), unless the Ricci scalar ("R") is zero, which only happens when there's no energy. They nonchalantly do that replacement without even a hint of explanation.

Elsevier and ScienceDirect should feel ashamed. They can go f**k themselves.