Interesting little dilemma I'm facing with the testing of conversation containers. The parts I'm having the most difficulty with are 1) groups and 2) moderated comments.

Ironically, conversation containers make both of these things a piece of cake. A constrained conversation by definition relays all content to its audience, which is exactly what a group does. A constrained conversation also by definition is a moderated conversation because a comment isn't part of the conversation until it has been added to it by the owner.

So why is there a problem?

Mastodon.

We altered everything we did in bizarre ways a few years ago just  to make our consent-driven communications compatible with the elephant in the room. And even though this new architecture does away with all of the ActivityPub hacks and performs the same operations in completely legal ActivityPub objects without requiring any extensions, Mastodon doesn't recognise the entire concept of groups and moderated content. So we'll either need to come up with hacks that make our stuff compatible with their primitive data model, or just say "fugg it" and do it the right way and "fugg them" if they don't want to play in a more capable vision of the ActivityPub universe.

I'm leaning towards the latter. We're creating a better, friendlier, and safer fediverse. This train has been building momentum for 14 years and has no brakes.  

Please don't stand on the tracks.