First of all, let me say that while the admins can do what they want, I think it’s pretty unfortunate that beehaw is defederating from .world and shitjustworks. I don’t want to see the fediverse fracture, I want it to encourage conversation with others.
But, beehaw has done what they’ve done. And I must say it’s annoying to be able to see their posts and comments and not engage with them. Plus, I could see their large, still visible communities taking away from ours without people realizing their own posts and comments aren’t working.
So will .world be defederating from beehaw? I don’t have a horse in the race, I’m jw
It’s not really a limitation of the software, it’s all up to each instance to decided how they handle things. Instance admins can code in a banner as you describe and they can make defederation transparent such as how Beehaw lists all linked and blocked instances. Every instance can do this, they just have to choose to implement it.
Beehaw didn’t come up with
/instances. That’s builtin to all Lemmy instance. It’s part of the Lemmy source code. Designed, written, and tested by the Lemmy devs. It’s part of their design for transparency.Instance operators aren’t software engineers. I have no doubt some are, and I have no doubt most are not. I also doubt that most are familiar with rust, typescript, postgres in addition to the Lemmy API along with ActivityPub. I’m sure the average admin can “code in” a banner involving some or all of that just like the average person can rebuild an engine, install custom shocks, and weld together a roll cage. But let’s say they do “code in” a banner that only displays itself depending on federation status. Now they have a separate patch set that has to be maintained and tested with each update to Lemmy. And published too because Lemmy is licensed under the AGPL.
So yes, this very much is a limitation of the software. The user should be informed, “you can’t comment or interact with this other instance.” Was this an oversight of the devs? I would say no. As a software engineer there are some things you don’t even think of as problems until they pop up as problems. It’s part of the engineering process and evolution of a code base.
And my intention from this long winded comment isn’t to rip your head off. Even if I come off like that sometimes 🙃. It’s me speaking from my knowledge and experience as an engineer.
To be fair, setting up an instance with your own patch is how improvements are made in the first place. Only a day after I set up my own instance, I forked the repo and started tinkering. Same case with lemm.ee. I’m a software engineer by profession but this is the first time I’ve touched Rust and inferno (I use typescript but with React/NextJS, inferno is new to me and pretty interesting). What’s funny to me is that lemmynsfw.com was the first highly populated instance I saw which is actually running their own patch. But hey, porn has always had a heavy hand in driving tech. Half of the best features of YT were ripped right out of PornHub.
Mind you, while lemmy-ui will probably be easy to pick up for most frontend devs working with a modern tech stack, I’m sure the main Lemmy repo will drive a lot of devs away just because it’s in Rust. But I can see why it’s like that. Lemmy is barely a blip in terms of CPU and RAM. The big factor for most instances is really just going to be storage space.
That’s true. This is one of those things where they would probably accept a pull request. Assuming it’s just a ui fix.
Someone already opened an issue on it.