Dear Fediverse,

Soon, some venture capitalist will put $5 million into a startup that offers free fediverse accounts with unlimited storage, etc.

Unless every other instance instantly blocks such a service from Day 1, you can kiss your fediverse goodbye and say hello to a new Gmail. Once network effects kick in, they will make the rules.

It _can_ happen here but you have an opportunity to kill it before it does.

#centralisation #fediverse

@aral Hmm, potentially a forgone conclusion, though I expect it to go like this:
1. Someone writes a very nice mastodon-like app but they license it liberally (not AGPL)
2. Someone makes a custom instance with some "cool features" which are not being upstreamed or shared
3. Lots of cool instances with cool extra features
4. Some of these instances become more commercial-centric
5. When the VCs come, already too many political connections between instance admins to make blocking possible.

@aral The "ideal solution" would probably be a software license like AGPL but which also disallows any inter-connection with instances containing proprietary code. This locks it open, but at the cost of being non-free-software (freedom zero) & annoying lots of FSF lawyers.

In the mean time, we can at least distrust any "cool new fediverse app with MIT license"

@aral Another solution is a mastodon inter-connection rules:
Rule 1: Block any instance which uses proprietary code
Rule 2: Also block any instance which doesn't also follow Rule 1 or Rule 2.

@cjd @aral I think here we start to get into failure modes, because the fediverse was never supposed to be about following one-size-fits-all rules apart from adhering to the protocol.

I would much prefer that people run Free Software, but if someone wants to make an instance that doesn't then so long as it talks the same protocol it should still interoperate and it shouldn't be up to me to dictate what software other people can run.
@xj9 @cjd @aral @Gargron

Well I'm not sure about that, but there does seem to be a tendency where instances will keep growing in numbers of users until the size is far beyond anything moderatable. There isn't much incentive for admins to set an upper bound or remove dormant accounts.

Once human scale moderation ends then the trouble begins.

It also looks like some instances have local only posts and that the consequent lack of transparency to the outside world can create an environment in which the admin can start getting away with dubious stuff - particularly on larger instances.

For a while I was thinking that the federated model was an organizational sweet spot, but especially in light of recent instance drama I'm returning to thinking that p2p or single user instances are the optimum type of social architecture.

@bob @xj9 @cjd @aral @Gargron I have trouble reconciling how single-user instances which AIUI means no moderation at all is the solution to weak moderation by large instances since you can already block individuals inside an instance.

@qwazix @Gargron @aral @cjd @xj9 Much of the contention appears to arise from instance blocking by admins, not individual blocking decisions made by users. If everyone were fully in control of their own user experience then individual preferences can be expressed without impinging upon anyone else.

Also when instances are small then the preferences of the admin and those of the users are likely to be similar. It's only when instances start to get large that power dynamics begin to have an effect.
Sign in to participate in the conversation

Server run by the main developers of the project 🐘 It is not focused on any particular niche interest - everyone is welcome as long as you follow our code of conduct!