I think this decentralization and federation is what web3 is all about, without all the corporations calling everything to do with monkey pixel art that costs a million dollars “web3”
I think this decentralization and federation is what web3 is all about, without all the corporations calling everything to do with monkey pixel art that costs a million dollars “web3”
That’s fine, as long as no instance gets like 99% of the userbase and we are back to square one
That’s my concern - that some company buys up some of the most popular instances, then “encourages” their members to concentrate on just one instance, builds up the number of communities on it until it becomes totally dominant and then cuts out all the other instances.
Not that the others couldn’t just continue, obviously, but if they’re starved of users they’ll be starved of content too and the gravitational pull of the big one(s) might drive the small ones into obscurity or closure.
BRB, got a business idea…
That’s exactly what Apple, Google, and Facebook did to XMPP. They all started with a federated open protocol messaging system. At first iChat users could talk to GChat, and Messenger users, as well as users of thousands of other servers. After they built their network they closed off federation under the guise of “feature development”.
To this day, iMessage still uses an XMPP based backend. But green text is for Apple users only!
Something poetic about Facebook dropping xmpp and now coming back to mastodon
I can totally see the big 5 trying to the same things with activitypub
Might seem naive, but I actually have a hard time imagining this. There’s just not a lot to make one instance more desirable than another, which seems like a bad thing, but I don’t think it is. I decided against signing up on lemmy.ml because it was laggy, so I went with a smaller instance- all the same content, but without the lag. If a lot of content gets created on one instance, there’s no pressure to pile in, because you can view, comment and interact from a different, smaller instance.
There are ways for large players to hijack things if we let them. There’s always the embrace, extend, extinguish method where a company starts adding proprietary features to the protocol and then cutting support to the competitors once they hit a critical mass.
Let’s say they add some proprietary features. That’s basically the difference between kbin and lemmy - they both support enough of the basic feature set required that anything they add on top of it is just “nice to have”, not something which would prevent a lemmy user from switching to kbin if every lemmy instance gets shut down.
Just the fact that lemmy/kbin exists and that we are on it suggests that the scenario is unlikely. Still, the idea would be that Meta would make their own ActivityPub based service. They make it super easy for facebook/Instagram users to join. Then eventually they roll out some feature that needs MetaPub, their new open source (if you agree to their strict license) version of the api. Now if you want to interact with people using those features, you need to go to a Meta approved instance. Eventually they disable the old ActivityPub system and cut ties with standard Lemmy/kbin instances. Probably in the name of security or user experience.
I hear you but I’d like to think the federated nature is too complex and spread out for one entity to take it all. That might be wishful thinking though.