@strypey Bluesky is currently to the ATmosphere what StatusNet was to the Identiverse back in the day (only moreso): a venture-funded startup that controls the protocol and runs what's by far the largest server.
-
(1/2)
@thenexusofprivacy
> Bluesky's Relay is currently to the ATmosphere what Google' search engine is to the modern web (only moreso)That's exactly what I was thinking when I made the post.
-
(2/2)
I had to think about this one for a minute.
@thenexusofprivacy
> Bluesky is currently to the ATmosphere what StatusNet was to the IdentiverseThis neologism is new to me. From context, I presume you mean the network of servers running StatusNet (or implementing OStatus)?
The most obvious difference is that when Identi.ca switched to pump.io, the rest of that network could continue to communicate. People running independent BS PDS would have to set up a new relay and appview.
-
@strypey I meant StatusNet the company, including the servers they ran, the software they provided, the proprietary protocol they developed. And yes the Gnu Social network survived the pumpocalypse, if Bluesky does something similar or shuts down I’d expect the ATmosphere to survive as well
-
(1/?)
@thenexusofprivacy
> the proprietary protocol they developedNot being vetted by a standards body is *not* what makes a protocol proprietary. The criteria for that are things like not having a published spec, and making breaking changes without consulting (or at least warning) independent implementers.
-
(2/2)
@thenexusofprivacy
> the Gnu Social network survived the pumpocalypse, if Bluesky ... shuts down I’d expect the ATmosphere to survivePerhaps, but I doubt it, for 2 reasons. First, the key technical difference I described in the OP. The second is a consequence of this; the people most likely to have the skills to set up a new centralised aggregator (relay) are more likely to see the weakness of the BS/AT model -especially after a total crash - and set up fediverse servers instead.
-
Coda: ATmosphere without BS is - technically speaking - just a bad Nostr clone. The only things attracting anyone to it now are the VC-funded rapid prototyping, and the big bump in population making the news. Which as we've seen with Eternal November in the fediverse, only keeps a network growing for so long.
-
@strypey well, if you want to try to convince other AP loyalists that AT’s not proprietary just because specs have been published and they’ve let people know about the relatively infrequent breaking changes, go for it, but my impression is that most people think protocols controlled by a single company are proprietary. But whatever, if the analogy don’t work for you it doesn’t work for you!
Thr impression I get from people who are looking at setting up relays is that there’s no way they’d do it on a fediverse infrastructure so I’m not sure why you think otherwise. AP isn’t good for flat all-public networks and that’s what they’re interested in. In practice it’s probably a moot point because I don’t Bluesky’s going away any time soon, but time will tell!
-
(1/?)
@thenexusofprivacy
> if you want to try to convince other AP loyalists that AT’s not proprietaryIf people want to use terms inappropriately, in ways that blunt their incisiveness, that's their business. I'm more than happy to debate anyone who wants to convince me that AT is a proprietary protocol. They know where to find me.
TL;DR BlueSky is centralised, but whether the protocol is open or proprietary is mostly irrelevant to the reasons for that.
-
(2/2)
@thenexusofprivacy
>AP isn’t good for flat all-public networksI've seen this claim made, along with the corresponding claim that AP is better for private social networking. It seems obviously wrong.
As your own research reveals, the existing fediverse isn't much better for private communication than the DataFarms. While orgs having a fediverse account on their own server is better than having a traditional website, under their own domain name, because any AP app can follow and comment.
-
It sounds like the only two things that we we agree on in this discussion is that Bluesky's whole-network Relay is currently to the ATmosphere what Google's search engine is to the modern web, and that people can use "proprietary" however they want and some meanings are more effective than others.
The actually-existing fediverse is not good at being a large flat network -- missing replies, the challenges of global search (especially on small instances), etc. So if the belief that AT's better for a large flat network seems obviously wrong to you, you're either ignoring these issues, or optimistically assuming there's a solution.
Conversely, parts of the actually-existing fediverse are actually quite good at scoped-visibility (as opposed to all-pulblic) networked communities (as opposed to a flat Twitter-like network). Unfortnantely that's not the model Eugen likes, so Mastodon doesn't support it particularly well. So if you think it's obviously wrong that AP is currently better than AT for this, you're probably confusing Mastodon with the broader fediverse. Of course, people are working on scoped-visibility networked communities in the AT ecosystem as well, so AP's lead might be tenuous, but it's still experimental over there. Time will tell!
