No reactions beside: monopolies are bad for innovation and why we cannot have nice things. You might hear some people say "but these big companies innovate". They were mostly done innovating two decades ago, now they just snuff out innovation and acquisition is one of their main tools.
well if you are waiting for the monopolies to be broken don't wait they will not be broken monopolies are here to stay, capitalistism for the rich and socialism also for the rich they best thing you can do is be rich yourself
I'll post this as a top level comment, because I think it is crucial: A few people are saying that it is expensive to run a relay but others have done it for as low as $34/month [1]. So unless somebody presents other proof that it is expensive I would say that those posters are either wrong and trying to mislead us on purpose.
Edit: actually the article links to somebody doing it for $18/month.
I've also done a full network replica (all the data indexed in a postgres) on a raspberry pi (with an 8tb nvme attached via a hat). Its really not expensive to do . And if I wanted to drop data older than say 3 months, it would be even cheaper still.
Case in point! This is an often mentioned statement on which the argument that "atproto is no decentralized" largely hinges. There are honest atproto digs out there but that is not one.
> There are honest atproto digs out there but that is not one.
Which one? The expense, or not being decentralized? The latter remains valid because the majority of the userbase chooses ("only" by default ofc) to coordinate through a single operator. Network effects mean that you either play by their rules or you aren't allowed in the garden.
It's good to learn that a full mirror is so cheap but I think the criticism still holds to the extent that it's a high enough price that unless something changes it will continue to discourage the network from ever becoming truly federated. Compare to activitypub where you can stand up a fully self sufficient node on more or less anything that's capable of networking. The obvious downside being that the network is more fragmented and often less reliable overall (ex nodes are regularly flaky or go missing entirely, no single unified view of the network, etc etc all the perfectly valid complaints about AP).
I think AP, AT, and nostr all get certain things right but all have major downsides baked into their designs. Note that I don't mean this comment to be negative, merely to respond to your remark that the dig in question is somehow invalid.
> We run a full AppView + PDS + Relay for ~$1,772/mo. PDS is cheap (~$0.03/user/mo, 4 vCPUs, 32GB RAM). AppView is the expensive part; indexes the entire network (16TB DB), not just your users. Storage scales linearly with network activity. PDS scales linearly with your account count.
The thing that makes ATproto nice and decentralized is that PDSs are decoupled from the application itself. Anyone can run a separate AppView, and just a handful of AppViews is enough to give people meaningful choice on ATproto, with the benefit of being much more approachable than ActivityPub for people who are not technically inclined.
People live in different conditions, have different income and (even more crucial) different expenses.
Not to mention subjective need. I may be interested in a relay, but I don't really need one. So for me 18$ will be too much. 5$? Maybe. (This is about as much as I'm paying for Fastmail right now, for example).
I'll go halfers with you, any other takers? I feel like sharing infrastructure via small online co-ops can take the bite out of the cost. So much cheaper then the cost of being the product via meta/goog etc.
Just pull a tarball from a signed URL, install deps, and run from systemd. Rolls out in 30 seconds, remarkably stable. Initial bootstrap of deps/paths is maybe 5 minutes.
They became much like woodworking or power tools. Accessible to anyone who wants them, but still requires an investment to learn and use. While the majority still buys their stuff from retail.
Or rents a printer for one-off designs. Unless you 3d print on the regular it's easier to pay someone to print one-off designs. You get a printer that gets regularly used and services and a knowledgeable operator. Not at all dissimilar to fancy commercial sign printers. In a past life working at $large-uni we really did try to make those damn things self-service but it was so much easier for the staff to be the print queue.
It turns out they're really great at building toys, cosplay gear and little plastic parts for things, but in general not that useful in most people's daily lives. Kind of like Ai.
> At every layer, the answer is "anyone can run their own." At every layer, almost nobody does.
But people do and it is reportedly fairly easy so the majority of people are on Bluesky's layers while all is well. But also I don't understand why any of this is a reason to be "wary", it's a great place to be with some unique technical properties - it is way more "open" than any other platform of similar scale.
At this point I despair at anyone who doesn’t understand that the problem isn’t the specific architecture, it’s social media as a scaled up, algorithmically driven concept. Stick so many people on one social graph that can’t possibly be effectively moderated by humans and it will turn into the same pit every time.
Atproto identity is going in the right direction but I hope they go in that direction harder. For example plc.directory (maps DID to public keys I think?) is heavily centralizing force.
