Discussion
Been brainstorming on a new concept(*) of "social network" based on ego graphs:
Everyone has their own "bubble": a network that they control through who they add/remove as friends (and who these friends add, and so on).
If someone is in your bubble, you're in their bubble.
Everything you post in your bubble, and everything people reply, is visible to everyone in your bubble (and no one else).
You keep multiple bubbles along different themes ("friends", "family" etc)
/🧵
@deivudesu This reminds me of Google+ Circles: https://google.fandom.com/wiki/Google%2B_Circles
Which sorta confirms your "bound to never never go viral, and therefore likely to never take off" deduction :)
Crucially, everything about this is fully e2e encrypted and decentralised: your account is a crypto key, and you are mainly responsible for keeping the list of people you are "friends" with.
3rd party servers act as dumb "mailbox" relays (to make up for the fact that most people's devices aren't always online or can't easily work as p2p), but only store fully encrypted payloads, and can theoretically be replaced by any other system.
2/ 🧵
@deivudesu that sounds like the core of chat mail/Delta chat.
@happyborg Interesting! Wasn't even aware of delta chat… 😅
From what I can tell, there are more than a few similarities… especially in the decentralised e2ee model.
*but* delta chat is designed primarily for symmetrical chat rooms (à la Signal), with no easy handling of "large" audience (eg ~1000), and no particular concept of a social graph (the central shifting ego-graph shaped mentioned above)… Which makes it quite different in the role it can play.
I think?
@deivudesu quite possibly too different. I was struck by the similarity of the chatmail architecture to yours, but don't know if that could be used or adapted to achieve what you want.
Building this around ego-graphs (subgraphs of the real underlying social network, centred around one user) allows for a much simpler (and more decentralised) model than protocols like #ActivityPub, #Matrix etc, that need to reconcile multiple participants having equal rights on a discussion space…
The trade-off, is that it does not scale very well: these ego-graphs must remain at a manageable size (~10-15k people).
… which is a feature, not a bug™
#mastodon #telegram #signal #graphs
3/ 🧵
IMHO, the main reason modern web (/social networks) suck, is power law, and everything that derives from it.
Dopamine-based, clout-mining, social media have turned us all into permanent badly-paid marketing interns: producing "content", rather than actually seeking to share ideas and cool stuff.
For the few that made a job out of it: that's cool, and you should definitely keep running that hustle.
But I just want to actually connect with my fellow humans. Exchange ideas and cool stuff.
4/ 🧵
@deivudesu I've been feeling the pressure with platforms driving engagement and pushing for activity. As a former lead on community software I am missing the ability to have sincere interactions at a micro level with people around me, currently spread over many, privacy-last, networks. Last night I drafted my thoughts on paper, and tried setting up a landing page for something new. I would love to hear what you think of it from your context:
https://kopling.org
@deivudesu very interesting. I have little time to respond rn, but this aligns with the notion in Social experience design #SX of Personal social networking..
https://coding.social/blog/reimagine-social/#personal-social-networking
As such, a social network that *by design* makes it impossible to build a large following, and renders algorithm-gaming/likes-mining pointless, is both:
* the only way to get back to the Magical Web of Yore™ (the connectedness and whimsy, without the Debordian obsession with selling oneself)
* bound to never never "go viral", and therefore likely to never take off…
But hey, it's worth a try…
5/ 🧵
Install bonfire.cafe
Get the full app experience
