I don’t have any good links. There has been an AMA with the developers earlier this year (on lemmy.ml) and I asked some questions regarding the development process and the direction of the project (and finances). I don’t have the link available, since I switched instances in the meantime.
And I followed the Github issues for some time. Usually you get an idea about how developers handle things by looking at their interaction with the community. There are some requests from last summer which I think would be worth looking into, but the devs say they don’t have the manpower. Same applies to several UI bugs. They’re still pretty much untouched as development focuses on the backend. And I’ve heard from people that the maintainers aren’t always happy with contributions. Which I think isn’t great because if you have an open source project along with a community, and then people try to engage but get disappointed because their day worth of coding is wasted and the PR denied… That isn’t going to foster a healthy community. I’m not sure if they’re working towards a different vision of the project, or manpower is that scarce. I mean they have 2 or 3 people working fulltime on Lemmy and they get paid a salary. I think we discussed that in the AMA. They definitely don’t get rich and pay isn’t what a big company would pay.
And there was something with the instance admins that needs improvement. I’m not sure what that was. Either image moderation or resource usage. (Or both.) Because admins need to abide by the law and pay attention to what’s uploaded (ideally without messing with the database manually) and I think they did some database performance improvements in the last few releases. I’m not sure. Rust should be fairly efficient. With databases you need to pay attention to what you’re doing. But I don’t know all the used frameworks. I’m just speculating here.
So… I really don’t have a link. I just occasionally read what people post here and want. And I sometimes follow the development.
And the fediverse as a whole… I don’t think ActivityPub is very efficient. The polling and simple design is compelling, but it’s not very performant. And it has some issues with caching etc. Also people want extensions and functionality that is well-defined and interoperable. But AP is just a well-defined core. And I think not even voting is part of that standard and just something people kind of agreed on. Which is problematic. Lots of normal stuff in Mastodon, Lemmy, etc isn’t really part of the standard. And I’m not sure if they release a new revision at some point. I think the current revision is a bit older as of now. I think we need that because the whole Fediverse is about interoperability.
Yea, generally fediverse projects have moved slowly. Mastodon has basically the same problems AFAICT despite being bigger and wealthier.
In a way I’ve been wondering if the fediverse is a kind of trial run for modern open source culture and what it can do other than packages for specific languages without the sponsorship of big corps. So far, it’s been really cool and super easy to take for granted. But also revealing I think in how hard it is to get things going when people need to make a living and everyone expects the internet to be free.
The performance implications of the protocol have been an elephant in the room for sure. I’ve never seen anyone do any sort of analysis.
And its quality as a foundation for a federated ecosystem seem definitely questionable. Especially as its main champion, Evan, seems super defensive about it and the idea of upgrading it. Sadly, it seems they’re an older tech person and see the protocol as their life’s work. So any proposal for starting again just runs into resistance. That there’s a weird cult around the protocol doesn’t help either.
As for the lemmy devs being antagonistic to contributions. I’ve heard that too but am suspicious. I’ve certainly seen 3rd party contributions go through. The whole image deletion episode was bad though. They even admitted to it to some extent.
I used to think they were way too “cranky” … but I’ve actually come around to the idea they push about not being too demanding of open source devs. It’s a serious issue with burnout being real and sustainability being vital for the fediverse. I was an early firefish user and saw that whole team, instance and project implode from the inside.
Would the devs be better at community management with proper salaries and community support and donations? Prob not! But some form of community manager could probably go far (even now if anyone is up for it) which seems viable if the support base got bigger.
I’m personally hoping to try to contribute by sometime this year, so I guess I’ll see how it goes and let you know if you like!
With the other group based platforms coming along (piefed and nodebb and maybe sub links) I’m hoping it becomes a richer part of the fediverse.
Thanks for the comprehensive comments. I didn’t even know the whole backstory behind ActivityPub. I always just took it for granted. I think we agree that there are quite some issues that are technical in nature. I think we’d need so solve a few of those before we can think about growing the Fediverse substantially. It’s certainly not easy. But adding drama also doesn’t help. I’m quite pisitive about the Fediverse. At least in the mid to long term. Maybe a bit competition and new projects will help add a few new ideas.
Always happy to chat about the fediverse … even when I’m critical it’s out of optimism for its future and promise.
On the point of competition … what’s kinda interesting is that for some people “fediverse” is now a broad term that encompasses BlueSky with their ATP protocol and Nostr and farcaster with their more crypto based protocols. So there’s certainly some strange mix of things at the moment.
I don’t have any good links. There has been an AMA with the developers earlier this year (on lemmy.ml) and I asked some questions regarding the development process and the direction of the project (and finances). I don’t have the link available, since I switched instances in the meantime.
And I followed the Github issues for some time. Usually you get an idea about how developers handle things by looking at their interaction with the community. There are some requests from last summer which I think would be worth looking into, but the devs say they don’t have the manpower. Same applies to several UI bugs. They’re still pretty much untouched as development focuses on the backend. And I’ve heard from people that the maintainers aren’t always happy with contributions. Which I think isn’t great because if you have an open source project along with a community, and then people try to engage but get disappointed because their day worth of coding is wasted and the PR denied… That isn’t going to foster a healthy community. I’m not sure if they’re working towards a different vision of the project, or manpower is that scarce. I mean they have 2 or 3 people working fulltime on Lemmy and they get paid a salary. I think we discussed that in the AMA. They definitely don’t get rich and pay isn’t what a big company would pay.
And there was something with the instance admins that needs improvement. I’m not sure what that was. Either image moderation or resource usage. (Or both.) Because admins need to abide by the law and pay attention to what’s uploaded (ideally without messing with the database manually) and I think they did some database performance improvements in the last few releases. I’m not sure. Rust should be fairly efficient. With databases you need to pay attention to what you’re doing. But I don’t know all the used frameworks. I’m just speculating here.
So… I really don’t have a link. I just occasionally read what people post here and want. And I sometimes follow the development.
And the fediverse as a whole… I don’t think ActivityPub is very efficient. The polling and simple design is compelling, but it’s not very performant. And it has some issues with caching etc. Also people want extensions and functionality that is well-defined and interoperable. But AP is just a well-defined core. And I think not even voting is part of that standard and just something people kind of agreed on. Which is problematic. Lots of normal stuff in Mastodon, Lemmy, etc isn’t really part of the standard. And I’m not sure if they release a new revision at some point. I think the current revision is a bit older as of now. I think we need that because the whole Fediverse is about interoperability.
Yea, generally fediverse projects have moved slowly. Mastodon has basically the same problems AFAICT despite being bigger and wealthier.
In a way I’ve been wondering if the fediverse is a kind of trial run for modern open source culture and what it can do other than packages for specific languages without the sponsorship of big corps. So far, it’s been really cool and super easy to take for granted. But also revealing I think in how hard it is to get things going when people need to make a living and everyone expects the internet to be free.
The performance implications of the protocol have been an elephant in the room for sure. I’ve never seen anyone do any sort of analysis.
And its quality as a foundation for a federated ecosystem seem definitely questionable. Especially as its main champion, Evan, seems super defensive about it and the idea of upgrading it. Sadly, it seems they’re an older tech person and see the protocol as their life’s work. So any proposal for starting again just runs into resistance. That there’s a weird cult around the protocol doesn’t help either.
As for the lemmy devs being antagonistic to contributions. I’ve heard that too but am suspicious. I’ve certainly seen 3rd party contributions go through. The whole image deletion episode was bad though. They even admitted to it to some extent.
I used to think they were way too “cranky” … but I’ve actually come around to the idea they push about not being too demanding of open source devs. It’s a serious issue with burnout being real and sustainability being vital for the fediverse. I was an early firefish user and saw that whole team, instance and project implode from the inside.
Would the devs be better at community management with proper salaries and community support and donations? Prob not! But some form of community manager could probably go far (even now if anyone is up for it) which seems viable if the support base got bigger.
I’m personally hoping to try to contribute by sometime this year, so I guess I’ll see how it goes and let you know if you like!
With the other group based platforms coming along (piefed and nodebb and maybe sub links) I’m hoping it becomes a richer part of the fediverse.
Thanks for the comprehensive comments. I didn’t even know the whole backstory behind ActivityPub. I always just took it for granted. I think we agree that there are quite some issues that are technical in nature. I think we’d need so solve a few of those before we can think about growing the Fediverse substantially. It’s certainly not easy. But adding drama also doesn’t help. I’m quite pisitive about the Fediverse. At least in the mid to long term. Maybe a bit competition and new projects will help add a few new ideas.
Always happy to chat about the fediverse … even when I’m critical it’s out of optimism for its future and promise.
On the point of competition … what’s kinda interesting is that for some people “fediverse” is now a broad term that encompasses BlueSky with their ATP protocol and Nostr and farcaster with their more crypto based protocols. So there’s certainly some strange mix of things at the moment.