I’m running on a personal lemmy instance, and I’ve been able to simply re-subscribe to the communities that I was subscribed to on my previous lemmy.ml account.
But what if I didn’t have that? How would I discover those communities?
On the micro blogging fediverse, I can use relays, follow other peoples boosts, or join gup.pe groups etc for content discovery and to give me federated content in general on which to do content discovery.
What does that look like in the lemmyverse niche of the fediverse? How does a small single person instance find new content? How do they get richer content search options etc? Right now, I’m just using search on lemmy.ml for that, but that’s a work around, not a solution
On bigger instances (like lemmy.ml ) you can click the instances on the bottom, which shows you all the linked (and blocked) lemmy instances. From there you can put them in search and it’s to my understanding they’ll start showing up in ‘all’ post feed. Again another work around, but this is closer to what you want.
I was hoping that groups from instances that my instance knows about would show up in group search, but that doesn’t seem to work either. They only show up in search if someone from my instance has subscribed to them.
I had someone from a science based instance reply to me somewhere, and so that instance now appears in the instances list for my lemmy instance, but the groups there don’t show up from a plain old search. I have to go to that site, find the group, get the URL and then search for it on my site, which is a pain.
And of course, lemmy is still in development, so I can work with that, but I was wondering if there was a future vision as to how it might work?
Yeah you’re right about groups/communities not showing up in search or the communities view. That’s unfortunate that they don’t all federate upon discovery of the instance. I kinda assumed discovery of the instance was discovery of all communities.
Federated communities do show up both on the search page, and the communities page.
https://lemmy.ml/communities/listing_type/All/page/1
But looking from Ada’s instance: https://lemmy.blahaj.zone/communities/listing_type/All/page/1
You don’t see all lemmy.ml communities. Unless this is a case of “federation takes some time” (might be, I’m not very familiar with the inner workings) it seems that the only communities known to lemmy.blahaj.zone are those that have been searched for.
Yep that’s correct, instances don’t automatically connect to all communities. Someone from that instance has to subscribe to it first, for that instance to start getting updates.
Right, but that’s the very problem I’m trying to solve! On a tiny instance, most of the communities I’m interested in on other instances won’t have anyone subscribed from my instance. It will be me being the first to subscribe. So I’m trying to find a way of finding them when no one else follows them yet :)
NRSK uses an admin-run “Fetcher” account to subscribe to various extenal communities and pull interesting content to our home “All” timeline. It’s a combination between browsing through communities on known and/or federated instances, and using join-lemmy.org to find new and unconnected instances. It’s all manually, which is a bummer but better than basically only local content for fresh users.
That’s an interesting idea! Not so relevant for my particular circumstances, but it’s a good idea if we ever do start to get more users in!
It would be pretty easy to write a bot which does this automatically. Problem is that you could easily end up with a lot of spam or other unwanted content.
Pulling all communities would definitively lead to unwanted content.
Since it’s basically a work-around for the local only (incl. already pulled communities) search, more ideal solutions could be:
Having some way to discover unfetched content (even if only from already federated instances) would be a significant addition to Lemmy!
@dessalines@lemmy.ml @ada@lemmy.blahaj.zone
How would you feel about bundling this type of automatic discovery into lemmy, as an optional server setting (default off)?
It’d be one of the periodic maintenance jobs we already have in there.