The footnote in Laurens' article accurately describes Relays as "an optional part of the network" that "can best be seen as an extension of an AppView. " Cory by contrast says "a home for users" is "(a "relay" in Bluesky-ese)." No, Relays only relay data that's stored elsewhere (in PDS'), and a user doesn't log into a Relay (you log into an AppView, and don't even know the relay exists) so there is no way in which it's a "home"As to why the Bluesky relays aren't as expensive to run as the Bluesky AppView, relays just relay the data (although not the media files, which get retrieved directly from PDS's); AppViews do a lot more processing, and are responsible for providing moderation. Over the summer, when there were about 1 million active Bluesky users, they estimated that a network-wide Relay was $150 / month. Since then, things have grown significantly, but they're also introducing a "non-archival" Relay; a few weeks ago, when there were roughly 9 million active usrs, I saw somebody say that he was spending $30/month to run one.For bonus points, am I wrong or is there currently only one BlueSky 'server' which has all the content, images etc?Blulesky runs the PDSs which currently store all the content and images. The software's open-source, and some people host their own (which is apparently fairly easy to do). There are other implementations as well,; Blacksky's, written in rust, is interesting although still fairly early in the development process.And Bluesky is by far the most popular AppView as well; there are a handful of others (frontpage.fyi for a reddit-like link aggregator, White Wind for blogging, a karaoke app whose name I forget) but they're tiny in comparison. So Bluesky certainly dominates the ATmosphere (the ecosystem of plaforms using the AT protocol) -- and it's not clear whether that'll change or not. It's also true that most independent AppViews today use the Bluesky Relay, so it is a single point of failure for the whole ecosystem; providing another one, as the AI companies running Free our Feeds claim they'll do, is potentially useful. Diversity is good! Opinions differ on how much impact that will actually have on the concentration of power in the ATmosphere, but I can certainly see why they decided to focus on that in terms of fundraising - it doesn't exist yet, some people think it's really important, and "everybody knows it's expensive" (even though it isn't). But Cory's explanations are badly off the mark.@negative12dollarbill @laurenshof
