[Edit: I see the problem, even with a self-hosted instance of 1, when you comment on posts in other instances that data is no longer held on your server, so you don’t own it and can’t control it directly, is that right?]
Not quite. It’s more like Bluesky works, but also not quite.
First, a note on the idea of “your” data. The law gives people rights over certain data. For example, copyright gives people rights over certain content, which translates to rights over data encoding that content. You may think of a movie as being yours because you have the file on your device. The copyright holder still considers it their data and will therefore demand control over your device through DRM.
Rights over data always means rights over what other people do with their computers and devices. Unfortunately, Fediverse users are not very tech-savvy. They demand more rights and regulations and then condemn Big Tech for the predictable consequences. They pull on one end of the string and blame dark powers when the other end moves.
The European GDPR also creates rights over certain data. You have GDPR rights over all data that is directly or indirectly related to you. For example, if I write about the current French President, then Emmanuel Macron has GDPR rights over that data, even if I don’t mention him by name. Of course, his rights will be limited by freedom of information. Also, these rights are rarely recognized outside of Europe.
What legal rights you have over data depends on your location. Copyright is internationally recognized, but its precise reach depends on location; eg the US has Fair Use. Even at a specific location, those rights depend on context, with a lot of gray area. This cannot be implemented technically.
With Lemmy it’s like this: When a user on an instance subscribes to a community, all (recent-ish) posts and comments in that community are downloaded to that instance. Users on that instance are served from their own instance.
Generally, a Fediverse instance keeps a copy of whatever data its local users might need. If your instance was the only source for some data, then every user in the whole world needing it would have to access your server every time they want it. Every user whether registered or unregistered would hit your server every time they reload. If a server buckles under the strain, you just get missing data. It just wouldn’t scale.
Bluesky has Personal Data Servers (PDSs) for that role. Those are the definitive store of some user’s data. This can be self-hosted easily. The data from all users is aggregated by a “relay”, If a PDS is like a personal web server, then a relay is like a search engine. That’s the one that you can’t self-host; takes big time capital expenditure.
I don’t think the Fediverse has a solution for this. Imagine Mastodon or Lemmy with 100M+ users. How do you find stuff? Well, making a crawler and search engine for the Fediverse would be simple. But that would also take major capital expenditure.
The Bluesky relay combines all activity into the “firehose”. Anyone can write apps that get data from the firehose and present them to users. When Bluesky blocked Mississippi, that meant that the official Bluesky App did that. Other Apps still work in that state.
Final bit: When you self-host, you need to be your own legal department. When you use a service, you are shielded to some degree. Eg when you infringe copyright, a social media service will usually just take it down. If you infringe copyright on your web server, or even via torrent, you may get a pretty hefty bill.
Fedi-users cheer when Meta gets sued or settled with a huge fine. Well, good luck running your own Facebook server. Fedi-users mostly aren’t very tech-savvy but when it gets to law, they are positively delusional.
Not quite. It’s more like Bluesky works, but also not quite.
First, a note on the idea of “your” data. The law gives people rights over certain data. For example, copyright gives people rights over certain content, which translates to rights over data encoding that content. You may think of a movie as being yours because you have the file on your device. The copyright holder still considers it their data and will therefore demand control over your device through DRM.
Rights over data always means rights over what other people do with their computers and devices. Unfortunately, Fediverse users are not very tech-savvy. They demand more rights and regulations and then condemn Big Tech for the predictable consequences. They pull on one end of the string and blame dark powers when the other end moves.
The European GDPR also creates rights over certain data. You have GDPR rights over all data that is directly or indirectly related to you. For example, if I write about the current French President, then Emmanuel Macron has GDPR rights over that data, even if I don’t mention him by name. Of course, his rights will be limited by freedom of information. Also, these rights are rarely recognized outside of Europe.
What legal rights you have over data depends on your location. Copyright is internationally recognized, but its precise reach depends on location; eg the US has Fair Use. Even at a specific location, those rights depend on context, with a lot of gray area. This cannot be implemented technically.
With Lemmy it’s like this: When a user on an instance subscribes to a community, all (recent-ish) posts and comments in that community are downloaded to that instance. Users on that instance are served from their own instance.
Generally, a Fediverse instance keeps a copy of whatever data its local users might need. If your instance was the only source for some data, then every user in the whole world needing it would have to access your server every time they want it. Every user whether registered or unregistered would hit your server every time they reload. If a server buckles under the strain, you just get missing data. It just wouldn’t scale.
Bluesky has Personal Data Servers (PDSs) for that role. Those are the definitive store of some user’s data. This can be self-hosted easily. The data from all users is aggregated by a “relay”, If a PDS is like a personal web server, then a relay is like a search engine. That’s the one that you can’t self-host; takes big time capital expenditure.
I don’t think the Fediverse has a solution for this. Imagine Mastodon or Lemmy with 100M+ users. How do you find stuff? Well, making a crawler and search engine for the Fediverse would be simple. But that would also take major capital expenditure.
The Bluesky relay combines all activity into the “firehose”. Anyone can write apps that get data from the firehose and present them to users. When Bluesky blocked Mississippi, that meant that the official Bluesky App did that. Other Apps still work in that state.
Final bit: When you self-host, you need to be your own legal department. When you use a service, you are shielded to some degree. Eg when you infringe copyright, a social media service will usually just take it down. If you infringe copyright on your web server, or even via torrent, you may get a pretty hefty bill.
Fedi-users cheer when Meta gets sued or settled with a huge fine. Well, good luck running your own Facebook server. Fedi-users mostly aren’t very tech-savvy but when it gets to law, they are positively delusional.