The good part is that you can be your own site admin, or you can choose one who reflects your ideals.

Moderation is great, the trouble is when a single moderation standard is used across a monopolized market.

I am my own site admin. The third largest instance's admin blocked my site (because I disagree with him about censorship), preventing 17,000 people from being able to follow me, or send or receive DMs from me. I also can't follow or read any of those users from my own server, despite their messages/profiles being public and available to the whole web.

This seems strictly better than what we have, though, where Twitter or Facebook decide the same. At least you can move instances, or start your own.

EDIT: also to note, email does kind of work this way with administrator applied spam lists. I would not want to use an email service that couldn't or wouldn't filter spam.

But email does work that way.

On my mailserver, I make the rules. I decide how my spamfilters, greylisting and blacklisting is tuned. I decide who is allowed to have an account. And therefore, I decide what is allowed to be sent and received.

That's a feature many users want. Plenty of users never want to see things they disagree with or take objection to, and want to have someone else enforce that pattern.

As long as it's optional - and in the fediverse design it is - then it doesn't seem entirely unreasonable. Whether or not those instances actually survive and have any users or just immediately collapse into internal bickering is beyond the strict scope.

> Plenty of users never want to see things they disagree with or take objection to, and want to have someone else enforce that pattern.

Am I unusual? When there is hate speech directed at me, I don't really care if I see it or not.

But I do want to prevent other people from seeing that hate speech (I don't want the hate to grow).

It's supremely rude to decide that you can be permitted to read something, but other people can't.

> Your site admin shouldn't be the one deciding who you can DM, or which people can follow your feed.

That's a necessary affordance, given the technology. A mail server could indeed refuse to forward some of your emails, and this could even make some sense e.g. as part of spam prevention policy.

There is certainly an argument for preventing unsolicited messages being received from unknown users, but rarely do people value a spam filter which prevents them from sending messages to people. (I suppose there are some corporate filters which try to prevent accidental sending of sensitive information to unauthorised recipients, but that's not the "feature" we are talking about here).

It is perfectly reasonable for a user to want to DM someone on an instance that has different moderation policies from their own, and it is equally reasonable to want to receive replies to those DMs.

If the specifications or implementations don't allow that, then I suppose it has to be justified by saying that the DMs could be used for sharing copyright infringing material (or worse), and admins don't want to run the legal risk of hosting that on their servers. Legally, though, that doesn't seem any different to operating a mail server, which don't typically have Content ID matching systems on them. Perhaps the implementation of end-to-end encrypted DMs would assuage some of these concerns a little.

It doesn't quite have to work like that, necessarily. Your instance could provide higher-quality tools for users to moderate their own feeds. Or there could be different degrees of opt-in to the site-wide moderation. Or the moderation could be community driven.

I think there is actually a lot of room for experimentation around how we interact online, and moderation is one of the most important areas. We need to think harder, rather than retreating to one of the two default positions.