Lately I’ve been thinking a lot about how discussion communities should fit into this. Relatedly, I’ve also been wondering how he fit Twitter into this model. (The model doesn’t have to fit everything–the goal is to be useful, not 100% correct–but still).
one approach to decentralized discussion would be to treat discussion comments/posts just like regular blog posts: host them on your own domain/some hosting service, and then layer reply notifications and such on top.
I think this is unnecessary. I think it’s better to separate posts into primarily “broadcast” posts, which fall under my “publishing” category, and discussion posts. I think it makes sense for discussion posts to be tied to a particular location/community/piece of software.
that’s vague; so to give an example: I think it’s fine to have a community on a particular service like Slack, Discord, Discourse, etc, and require that people in the community use that service. I don’t think there needs to be some kind of discussion protocol so that e.g. you could have a single community where some people members are using slack and some are using discord; with the slack people never even needing to create a discord account and vice-versa.
But still, there are some ways in which it would be nice to have discussion communities interoperate with the other services in my model (reading, publishing, aggregation). so my question is: what are those ways?
(actually, figuring that out for all the services–figuring out which functions should happen within services, and which should happen via multiple services working together–is pretty much the whole point of this thread. but lately I’ve been particularly thinking about it in regard to discussion services)
and again, the other big question is, does Twitter fit into this model? I’ve been mainly thinking of discussion communities as relatively small, focused communities, like slack, discord and so on. under this model, do you treat “Twitter” as a single, giant discussion community? or do you just leave it out of the model? again, not every facet of online communication needs to fit in this model; rather the idea is simply “going forward, the model is a useful tool for designing new services”.
originally I was inclined towards the latter, under the thinking that you just don’t need a single big community with everyone in it. there might not be anything of value lost if discussion primarily happened in a network of these smaller communities. it would be harder for things to go viral, but a lot of people say that would be a good thing. for a message to spread across communities, you’d have to manually copy and paste it instead of just hitting retweet.
and actually, that still seems like not a bad approach. but at the same time, I think Twitter is valuable for making new connections between people who otherwise might not have been in the same community. maybe those kinds of connections would still happen in my new model, via different methods. 
. but in any case, I guess it doesn’t hurt to treat Twitter like a single community even though it’s huge. if the unbundled model does gain adoption, then either small communities really are sufficient and Twitter can be left to fade away, or Twitter does have unique value and it’ll continue to be used.
OK, that was all kind of abstract, so next I think it’d be best new to list out the actual concrete ways in which these discussion communities would interop with reading, publishing, and aggregation services.
(actually maybe for new I’ll focus just on the interop between publishing and reading services–more on that later)