Yeah, I’ve seen that proposal. It’s not all that much more compelling to me. If mods were willing to be grouped together, why would they not also be willing to merge their communities entirely and retire one of them? What compelling capability does three but kind of one communities give you that one community with three mods doesn’t?

And more importantly, are those capabilities more valuable than the simplicity of people not having to understand these compound communities. Federated designs are generally significantly more complicated to use than “normal” designs. A sprinkle of federation in a system design is a very powerful hedge against individuals coopting the ecosystem, and that make the extra complexity worthwhile (to me, there are smart folks who say that federation is DOA because it’s inherently too complex). But almost all people with experience designing federated systems agree that a heaping shovelful of federated system design makes a system an unusable mess of conceptual complexity that no one will bother to learn. It seems pretty clear to me that compound communities fall into this heaping shovelful category, and do no “pull their weight” in complexity. Reasonable people can disagree, but I both would not use such a feature unless I had to… and I think the lemmyverse would be better off without it.

Instead, it’s my belief that devs should focus on improving community discovery so people naturally find the most active communities, low-traffic ones die off, and we eventually reach a mature state where… like reddit… even though community duplication is allowed… no one cares about it because the better run communities naturally grow and are easy to find. Improving community discovery has more powerful beneficial side-effects in other ways as well, and if done properly could reduce complexity rather than increase it as this proposal does. But to each their own, maybe other people see something I don’t.