In my opinion, there are two big things holding Lemmy back right now:
-
Lemmy needs DIDs.
No, not dissociative identity disorder, Decentralized Identities.
The problem is that signing up on one instance locks you to that instance. If the instance goes down, so does all of your data, history, settings, etc. Sure, you can create multiple accounts, but then it’s up to you to create secure, unique passwords for each and manage syncing between them. Nobody will do this for more than two instances.
Without this, people will be less willing to sign up for instances that they perceive “might not make it”, and flock for the biggest ones, thus removing the benefits of federation.
This is especially bad for moderators. Currently, external communities that exist locally on defederated instances cannot be moderated by the home-instance accounts. This isn’t a problem of moderation tooling, but it can be (mostly*) solved by having a single identity that can be used on any instance.
*Banning the account could create the same issue.
-
Communities need to federate too.
Just as instances can share their posts in one page, communities should be able to federate with other, similar communities. This would help to solve the problem of fragmentation and better unify the instances.
Obviously there are plenty of bugs and QoL features that could dramatically improve the usage of Lemmy, but these two things are critical to unification across decentralized services.
What do you think?
EDIT: There’s been a lot (much more than I expected) of good discussion here, so thank you all for providing your opinions.
It was pointed out that there are github issues #1 and #2 addressing these points already, so I wanted to put that in the main post.
Lemmy needs two things to be successful:
- users
- users
and it’s already getting more and more of each of those.
It won’t get more users if it continues to be difficult to use.
I think these are “nice to have” features rather than absolutely essential, but:
For 1. I could deal with just being able to download my list of subscriptions and upload it to another server. That’s the only bit that’s really slow to copy over by hand.
For 2. I think the main thing that really would benefit is the ability to search all active communities on all servers. The way it is now is alright if there are only half a dozen really active instances whose communities I might be interested in, but it doesn’t scale if there are hundreds of servers to check out. Probably the more important of the two IMHO in the long run.
Being able to download your own data would be a start
Yeah, search all federated communities.
It would not be hard to get a list of all communities on each federated instance. Update it a few times a day, even once a day.
But this is the hardest thing for me, searching is a challenge
Lateposting, but DIDs would also solve a problem I have on kbin. I have an account on a certain kbin server because I wanted to pick a smallish server. A server small enough that it isn’t a big hub and thus helps out with federation, but big enough that it’s probably not just one person’s personal server that will only ever run during the 10 minute windows where they personally want to check kbin, which probably won’t overlap with my own.
Sometimes that server goes down. So I also have an account on the biggest kbin server of them all, kbin.social, so I can still use the site and interact with it when my home server is down.
Posting this on Beehaw because I didn’t subscribe to Technology until after this post was made, and I currently have no way to force this post to show up for me on kbin.
