More seriously, I'm not seeing how federation solves any social network problem better than subreddits did? In either case, the quality of the discussion will depend on the moderators, who can be good, bad, or evil.
At one time, making sure that users could block other users on the same social network might have seemed like it might enough to prevent abuse and keep social networks out of the moderation business. Users that don't get along can just block each other. Good enough?
Maybe not. You say you know the "worst case" but that might not be imaginative enough? There are apparently people nastier than that, and they're creative, and they gang up.
I suspect that if federation really got going, there would be plenty of abuse. Not sure what form it would take. Chain mail? Phishing? Something new? Who is going to be in a position to fix it? How much drama will this cause?
Even in the above botnet+phishing situation, nothing is stopping instances from cutting federation off completely and then banning suspected users, a move that a growth-oriented network would think many times before instituting.
Though there are email blacklists for known spammers, and it can be hard to administer a new email server. You wouldn't want to block Gmail, though.
Speaking of which, I'm not seeing why every instance operator would decide not to grow. It seems like, if the federated protocol is successful and allows it, new large providers would spring up like Gmail and Hotmail.
> Though there are email blacklists for known spammers, and it can be hard to administer a new email server. You wouldn't want to block Gmail, though.
Most people typically find an instance with people similar to them. Most of my followers/followees are on the same instance as me. While blacklisting an instance can be problematic, it's also probably not that big of a deal for most. Moderators are more willing to take actions like this when each user does not contribute to their bottom line.
> Speaking of which, I'm not seeing why every instance operator would decide not to grow. It seems like, if the federated protocol is successful and allows it, new large providers would spring up like Gmail and Hotmail.
Hotmail and Gmail came about because running your own email is difficult (I used to run my own email). You have to setup SPF records, DKIM records, on top of which many mailers send non-standard mail out, which you have to make sure your rules accept these non-standard pieces of mail. Right now the ActivityPub standard is well defined, and most instances already follow the server-to-server API. Spam is also a huge problem in the email world because of the ease of sending an email. You don't need a mailserver to send mail to an address, but you need to setup an instance and have other users specifically follow you to send spam toots, or Fediverse messages. On top of this is a community interest in creating software (like Pleroma) that is easy to install and administer. All of these combined decrease the friction for lay (for multiple definitions of lay) users to run their own instance, which makes the gulf between a (not-yet-existent) commercial instance and a personal instance a lot smaller. While there will be a space for commercial providers which use ads or subscriptions to pay for their instances, the ease of personal setup makes centralization a lot less of a tendency for the Fediverse than it did for email.
Except that it's not at a per-user level, but at a per-instance level. Sorta like groups.
Instance A could decide that they don't like the moderation policies of instance B, and then not federate with them. This means all users on B are blocked from all users on A.
Rather than individual users having to play whack-a-mole, they can come together.
People can make their own thing under their own geography-bound laws while still connecting with a global network. That's really neat.