That error screenshot was from lemmy.one, which has about 1,100 monthly active users, which is a 5th the users of lemmy.ml, and less than 5% the users of lemmy.world. I’m not in denial that high usage causes problems if the deployed instance doesn’t have adequate resources and configuration for it. And I also acknowledge that spreading the users across federated instances will improve the situation. But federation has to be reliable. There are active bug reports about problems with how the platform is handling federation, and how activity can be lost
- https://github.com/LemmyNet/lemmy/issues/3435
- https://github.com/LemmyNet/lemmy/issues/3101
- https://github.com/LemmyNet/lemmy/issues/3062
- https://github.com/LemmyNet/lemmy/issues/2142
- https://github.com/LemmyNet/lemmy/issues/3121
So the people who are correctly pointing out that these instances are being DDoS’ed by huge user influxes need to also acknowledge that there are still architectural issues with the federation between instances.
While these issues aren’t too big a deal for some communities, they are for others. A community following a sports league and posting threads to discuss matches live and in real time is pretty much only useful with all the users being on the same instance right now. But putting all the users on one instance is making the instance unusable. Combine those factors and people leaving Reddit will either go back to their Reddit sub or find another platform.
No me gusta