cross-posted from: https://hachyderm.io/users/maegul/statuses/111820598712013429
Is decentralised federated social media over engineered?
Can’t get this brain fart out of my head.
What would the simplest, FOSS, alternative look like and would it be worth it?
Quick thoughts:
* FOSS platforms intended to be big single servers, but dedicated to …
* Shared/Single Sign On
* Easy cross posting
* Enabling and building universal Multi-platform clients.
* Unlike email, supporting small serversNo duplication/federation/protocol required, just software.
I still think we need a simple social media protocol that gives me the power to curate my feed rather than hoping my admins don’t defederate with everybody else (followed by hordes of drooling goons telling me to start my own instance).
Well that’s kinda the point of my quick suggestion in the original post.
Instead of committing to federation, how about committing to aggregating clients that allow you to do exactly this. Right now, there’s no app that will work for both lemmy/kbin and mastodon/microblogging. No way to unify the notifications or even combine the feeds or just have a unified interface for the two platforms (that are, let’s face, both just full of text messages and feeds).
By allowing each platform to be distinct but remain open with their APIs and “play nice with each other” while leaning into the value of aggregators as a primary part of the value proposition of the system, users might be better served.
I have a hard time imagining what that looks like, which is just a failure of my ability to think about these technologies. But what I’m talking about is a little different, simply because I don’t think we can go from these diverse systems into something simple and elegantly connected.
I mean something like email but structured differently. Though email still has spam filters and blacklists, and a new social media protocol might still need those (inevitably infringing on my curatorial freedom similarly to defederation).
My point is that I’m still looking for something new, rather than to reform the defediverse.
Edit:
I might be wrong. It might be good to leverage what we started here and reform the tech to give users more freedom, and take pressure from admins.
Also… maybe email is not the example I should follow. Maybe it’s more like torrents. P2P social media.
Well I’m spitballing here, so I wouldn’t worry about not being able to imagine it! I’m struggling too!!
Is there a chance that BlueSky is more like what you’re after?
No, that’s fairly centralized too. I think I want a peer-to-peer social media protocol. Maybe more like torrents than email.
The fundamental problem is that all this data needs to be hosted somewhere. P2P systems have the issue of persistence: either posts only stick around as long as the people who posted them keep their server online, which is then a burden on anyone who wants to be active in the community, or everyone shares the responsibility for hosting, and then what happens if someone posts CP? Is it just mirrored across the entire P2P system, and each person has to individually root out the CP or just be okay with hosting CP?
Torrents work because you have to actively join a torrent. But discoverability is handled from the outside, through trackers. Trackers choose what they want to host.
Tor or really I2P are the closest equivalents, but they work because everything is encrypted going through them. It’s a privacy thing. With social media, everything is public by design.
Persistence could be traded off for decentralization. Just like torrents’ associated data are stored on people’s computers (and the data dies if nobody is seeding) this kind of social media doesn’t have to be permanently stored on a server.
Yea right. Me too I think. It’s out there and has been for a while. Just don’t think it’s ever taken off.
You might find this interesting: https://pfrazee.com/blog/why-not-p2p
I’m checking this out!
Yea I didn’t know before seeing that that one of the BlueSky devs (the author of the blog post) was heavily involved in p2p stuff (eg beaker browser).