It’s all explained in the link I shared, you do like any other website but instead of using multiple servers owned by a single company it’s multiple servers owned by random people and devs can create a front-end to access the data found on those servers.
No, it’s not, each instance hosts its own content, and users are assigned to an instance.
What I’m talking about is making the end user experience the same as any centralized website, the hosts are just hosting parts of the total database (randomly assigned, with backups in mind so everything is hosted on multiple servers, with the option not to host content flagged as NSFW) and people create front-ends to access and interact with that database. Users aren’t assigned to a specific server, their credentials are just part of the database.
As I mentioned in my other comment, think like any other website but instead of relying on AWS to host the data on a bunch of servers all over the world, it’s people like you and me and just like Reddit before the API scandal, you let devs create apps to push and pull data.
From a user perspective that removes the admins from the equation entirely (people weren’t complaining about the sys admins on Reddit, they couldn’t care less about the servers), users are their own boss and filter their experience as they see fit, mods still exist but no one has the power to suddenly device you just can’t interact with tens of thousands of users all of the sudden just because you signed up from the wrong place (in their mind). If an admin decides to stop hosting, the data they had on their server is backed up on other servers and things are rebalanced between servers to make sure there’s still a backup of everything.
I doubt that such a model would be viable in practice. Again, server owners must be able to delete content under certain circumstances. How would you deal with illegal content? How would you deal with communities that share CSAM for example, or are used by criminal or terrorist organizations, black market arms dealing etc? I doubt that mods of such communities will take care of it.
Also, I don’t think that the issue of defederation is as big of a deal as you make it to be.
My instance is a pretty big one and is defederated from other major ones because of the admins on those other instances. That’s tens of thousands of users I (and any other user on my instance) can’t interact with because their admin said so.
They would need to filter the content being pushed to their server just like current admins have to do even if they elect not to host NSFW content. There’s tools to automate the process. If it’s a non issue with the way things work now then it’s a non issue with what I’m proposing.
Sounds you would still have admins, just with extra steps and extra work.
There’s tools to automate the process.
Oh, yeah, because automated tools to recognize illegal content work so well. (/s. For real, there are reasons why computer science experts, privacy experts and civil rights experts are pretty much united against such tools. If you want to know more, look up the resistance against EU proposals for Clients Side Scanning and such.)
Your ideas would not work in reality without open up.platforms for all the things we do not want.
My instance is a pretty big one and is defederated from other major ones
Which ones? I looked through some major instances’ lists of blocked instances and sh.itjust.works isn’t on any.
Enlighten me. Where do you want to host stuff?
It’s all explained in the link I shared, you do like any other website but instead of using multiple servers owned by a single company it’s multiple servers owned by random people and devs can create a front-end to access the data found on those servers.
But that’s literally how the whole fediverse works right now.
No, it’s not, each instance hosts its own content, and users are assigned to an instance.
What I’m talking about is making the end user experience the same as any centralized website, the hosts are just hosting parts of the total database (randomly assigned, with backups in mind so everything is hosted on multiple servers, with the option not to host content flagged as NSFW) and people create front-ends to access and interact with that database. Users aren’t assigned to a specific server, their credentials are just part of the database.
As I mentioned in my other comment, think like any other website but instead of relying on AWS to host the data on a bunch of servers all over the world, it’s people like you and me and just like Reddit before the API scandal, you let devs create apps to push and pull data.
From a user perspective that removes the admins from the equation entirely (people weren’t complaining about the sys admins on Reddit, they couldn’t care less about the servers), users are their own boss and filter their experience as they see fit, mods still exist but no one has the power to suddenly device you just can’t interact with tens of thousands of users all of the sudden just because you signed up from the wrong place (in their mind). If an admin decides to stop hosting, the data they had on their server is backed up on other servers and things are rebalanced between servers to make sure there’s still a backup of everything.
I doubt that such a model would be viable in practice. Again, server owners must be able to delete content under certain circumstances. How would you deal with illegal content? How would you deal with communities that share CSAM for example, or are used by criminal or terrorist organizations, black market arms dealing etc? I doubt that mods of such communities will take care of it.
Also, I don’t think that the issue of defederation is as big of a deal as you make it to be.
My instance is a pretty big one and is defederated from other major ones because of the admins on those other instances. That’s tens of thousands of users I (and any other user on my instance) can’t interact with because their admin said so.
They would need to filter the content being pushed to their server just like current admins have to do even if they elect not to host NSFW content. There’s tools to automate the process. If it’s a non issue with the way things work now then it’s a non issue with what I’m proposing.
Sounds you would still have admins, just with extra steps and extra work.
Oh, yeah, because automated tools to recognize illegal content work so well. (/s. For real, there are reasons why computer science experts, privacy experts and civil rights experts are pretty much united against such tools. If you want to know more, look up the resistance against EU proposals for Clients Side Scanning and such.)
Your ideas would not work in reality without open up.platforms for all the things we do not want.
Which ones? I looked through some major instances’ lists of blocked instances and sh.itjust.works isn’t on any.