Sven Slootweg (soft-deprecated) @joepie91@pixie.town

@ruario @aral @Vivaldi @jon I'd expect that the instance admin community will probably be happy to help out and provide feedback/advice on tricky cases, as long as your intentions are to safeguard the community :)

I'd also recommend following @Curator, who frequently posts about *why* the (quite good) moderation policies on .art are what they are, and the considerations that go into them.

Another option, of course, would be to hire someone with instance moderation experience from the community - either to advise on tricky cases, or to do the moderation themselves.

@ruario @aral @Vivaldi @jon Sounds good :)

@ruario @aral @Vivaldi @jon Regarding limits and wanting to introduce new people to the fediverse: I understand where you're coming from, but I think there are better ways of approaching that. Some ideas that immediately come to mind (but that'll probably need refining):

1. Integrating something along the lines of joinmastodon instead, to help people find their place in the fediverse using existing instances (with of course a clear note that they're volunteer-run, to set the right expectations).

2. Setting a user limit, but making it a MAU limit and operating a waitlist. People have a tendency of moving to different, more specific instances over time, which would free up spots on the Vivaldi instance in the process. That way the size can be limited while still allowing a (manageable) steady influx into the network.

3. Some combination of those? "Hey, you've been using fedi for a while, did you know that you can also find a community that's perfect for you while still keeping in touch with the people you already know?" as a subtle prompt-to-move-instances.

There are probably other approaches I haven't thought of yet. Maybe other instance admins watching the #FediAdmin or #FediAdmins tag also have ideas?

Regarding ownership changes: that can take a lot of forms, and not all of them are voluntary. Think bankrupcy proceedings, or the majority shareholder changing their mind for whatever reason (see eg. the recent PolyMC fiasco, which wasn't commercial but otherwise similar).

I'd even argue that the situation with Gargron and mastodon.social is an example of how this can go very wrong. That also wasn't foreseen in the very beginning AFAIK, and now it's causing huge issues for the network as a whole (and they're getting defederated).

Unexpected bad things can always happen, so IMO it is always best to make sure that your worst-case plans are proportional to the potential harm that could be done to someone or something else :) And being a major instance on the network is a large responsibility.

Regarding moderation: that is a promising policy! Certainly a lot better than I've typically seen from corporate instances. 👍

@ruario @aral @Vivaldi Hi, I appreciate that you're around to talk about the details of the implementation :) I'm hoping that this will turn out better than previous attempts by others at corporate instances!

I do have two crucial concerns that I'd like to address, however:

1. You really should commit to an instance size limit of some sort. Huge instances are a problem - not just operationally for you, but also because even if you have good intentions *right now*, a change in eg. leadership may suddenly make it a liability for the network. This remains true even in the face of explosive network growth.

Large systems are tools of power, therefore prone to eventual misuse, and should ideally not be constructed at all.

2. You'll have to think carefully about moderation - don't make the mistake that Librem One did, for example, where they argued that they couldn't moderate anything non-obvious because "they're paying customers" (and got promptly defederated as a result).

Also note that moderation norms on here are generally stricter here than on typical for-profit platforms; this place was built by marginalized folks, and so there is a culture around community safety. It's a good idea to replicate that.

Bundling it with a 'browser account' is also pretty risky here, because it means a ban has more impact (it might affect the functioning of one's browser?), and so you may be more reluctant to issue it. Kind of like the "my Quest doesn't work anymore because I got banned on Facebook for my name" problem. Decoupling it in some way may be better for everybody involved.

--

I'd say that both of these points (instance size and moderation) need concrete commitments immediately from the start; if there are no such commitments, it is highly likely that you will be defederated by a large part of the existing network very quickly, because it will pose a threat to other network participants. Folks have bad experiences with corporate instances. It's best to avoid that situation :)

I'd be happy to provide some further pointers on how to approach this, if wanted - though especially for the point of moderation, keep in mind that while I do have a lot of moderation experience elsewhere, I don't run a fedi instance. It may be best to talk to other actual instance admins for that.

holy shit

I just realized that musk is 'leading' entirely by aphorisms

calling 'bloat' (without understanding what stuff is for), the 80/20 rule (firing 80% of contractors, killing 80% of microservices), and doing a 'scream test' (for turning off those microservices)