search

Minds / engine

mirror of https://gitlab.com/minds/engine
https://minds.com
GNU Affero General Public License v3.0
200 stars 93 forks source link

Plans for federation / roadmap? #11

Closed ghost closed 6 years ago

ghost commented 7 years ago

I was just wondering if there are any plans for implementing federation for Minds, I couldn't find a roadmap so perhaps it'd be a good idea to start writing one, it could include an "off-road"-section to explain why certain popular or obvious suggestions aren't implemented.

Some suggestions for the roadmap (by me, I hereby assign the copyright to Minds Inc. and license it MIT):

Perhaps my roadmap is a bit too fine-grained right now, and would be better as a bunch of issues in the issue tracker; however I'm not super invested in this platform as of right now and I don't want to spam the issue tracker so I refrained from that.

markharding commented 7 years ago

Thanks for this. To be brutally honest, my view is that the current examples of 'federation' have failed from both a consumer and technical point of view. The interfaces and experiences are too confusing for over 99% of users and whilst they claim to be decentralized, when you look into them deeper, they are not at all.

I believe the solution would be to create a torrent style, p2p network with no central points. This would require a huge shift in the way we use browsers and social networks. Projects like Webtorrent and WebRTC are making this more and more possible

ottman commented 7 years ago

I tend to agree with Mark. However I'm not going to say we'd be unwilling to accept an intelligent implementation/contribution of current federation status quo. It depends. But for us it makes more sense to putting our resources to truly P2P. We want this for reasons from privacy to scalability and web freedom.

ghost commented 7 years ago

I also believe a P2P-style network would be ideal, perhaps in combination with "gateway nodes" for users that only access the network via the browser, providing identities for "leech" users under their banner/namespace.

Some combination of ratings, web-of-trust, and meta-ratings (rating graph isomorphism? A recommendation system based on rating agreement, to automatically separate bots from humans?) would be necessary to combat spam/abuse in a semi-automated fashion, and the process would have to be transparant to detect abuse of the system.

Sorry that my post is a bit rambly, I've been having a hard time putting these ideas together in a structured way, I've been trying for at least a year.

On Fri, Jul 28, 2017 at 1:55 PM, Bill Ottman notifications@github.com wrote:

I tend to agree with Mark. However I'm not going to say we'd be unwilling to accept an intelligent implementation/contribution of current federation status quo. It depends. But for us it makes more sense to putting our resources to truly P2P. We want this for reasons from privacy to scalability and web freedom.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/Minds/engine/issues/11#issuecomment-318633283, or mute the thread https://github.com/notifications/unsubscribe-auth/AEePF_LVp1YgS30MLlQ_6IdrDg_NudvBks5sScw_gaJpZM4Omcpt .

markharding commented 6 years ago

I'm closing this issue as its been a while with no activity. Feel free to continue the conversation over in the Open Source Group (https://www.minds.com/groups/profile/365903183068794880).