A software developer and Linux nerd, living in Germany. I’m usually a chill dude but my online persona doesn’t always reflect my true personality. Take what I say with a grain of salt, I usually try to be nice and give good advice, though.

I’m into Free Software, selfhosting, microcontrollers and electronics, freedom, privacy and the usual stuff. And a few select other random things, too.

  • 4 Posts
  • 741 Comments
Joined 8 months ago
cake
Cake day: June 25th, 2024

help-circle


  • I’d say this is unlikely to work out. It mainly combines the downsides of two approaches. The centralization will make it less free and diverse and gives power to few people, while the decentralization adds unnecesary complexity. Since at that point it’s mainly one large instance, but that has to send out loads of network traffic to very few people at other places to keep them in the loop. At that point, why not make it 100% centralized? That’d make programming and maintainance way easier.


  • Uh, idk. There’s also a list detailing that: https://join.piefed.social/features/

    It’s a different software, connecting you with the same communities and people. Just has slightly different features, a bit more control here and there, a few perks, has a different design philosophy and is written in an entirely different programming language, which affects participation, maintainability, resource usage… You can see how it looks for example on https://piefed.social/ I always struggle to describe the detailed differences, because there are a lot of them and it has a lot to do with what’s important to you and what you’re used to. It’s a bit like describing how a banana tastes IMO. You better have a look yourself.





  • Hmmh, I don’t think it’s even elitism in this case. Feels to me like something else. But I’m the wrong person to ask, since I do not share that opinion.

    I think your proposal with default subscriptions (or whatever it is exactly) is a solid idea, though. In fact, I’ve heard some people scroll through the “All” feed here on Lemmy and subsequently block the things they’re not interested in. I’d say that’s about the same direction. And I mean why not? We also have sorting by popular, and things are popular for a reason. So we might as well subscribe new users to the 10 most popular communities.

    It’s a bit more complicated than just that, we’d have to take some care not to entirely destroy diversity and pour some cement over the whole thing, or we end up with a small echo chamber of just lemmy.world and AskLemmy and NoStupidQuestions… But I guess there might be some solution in beween the extremes. And things might change due to the size of the platform. An “All” feed might still be useful at our current size, but might prove to become infeasable once we grow.



  • What I’ve seen many times is people stating the opinion that we don’t need to grow. We’re not some big commercial platform and we don’t need to satisfy some investors. Growth will come naturally. Or it won’t.

    My opinion is, judging by the numbers… We aren’t growing for quite some time now, so Lemmy will most likely stay what it is. I’d love if it were a super attractive place, though. And everybody would like to join.

    Sane defaults are always a good idea. I’m a bit split on the “minimal effort” though. Minimal effort is letting some algorithms dictate what to consume, simple truths, and not bothering with complicated stuff like freedom or privacy.

    And what I often see is people trying to solve such problems solely by technical means. And I think that’s not even half of it. We mainly need a nice and welcoming atmosphere, nice and interesting people, good content…


  • I share your opinion. They seem to have clarified a few things, though. Their license states what kind of reuse is allowed. You need to read it thoroughly. For example you can study the code or adapt it for personal hobby projects, if it’s non-commercial and you add the required statements… But I think it’s completely unappealing to use GrayJay or contribute to the project. It’s not Free Software, so you don’t get much in return. They tell you you should send pull requests, but as far as I can see there is no way of logging in to their GitLab. So you somehow need to hunt down their GitHub mirror, and file something there, in the hopes someone is going to read it amongst the hundreds and hundreds of open bugreports… And their phrasing and use of the term “open source” is just annoying and bound to confuse people. I’m not sure what Louis Rossman is doing these days, but when they launched it, he was making videos with lots of outright false claims about the licensing. A lot of that hasn’t been ideal. I’ve sent them some comments back in 2023. But they never replied directly. I believe they took notice of the discussion and promised to step up their game concerning their community. But I don’t think they’re doing a particularly good job. And I suspect they lack a deeper understanding of what Free Software is, what it’s about and good at, how to foster a community that’s not just alike what you get on Youtube as a creator.

    But I’m not mad at them. As long as they keep Louis’ promise of not prosecuting any individual for getting confused by their mixed signals. They seem to be mildly successful with whoever their target audience is. Guess I’m just not a part of that. But I have NewPipe/Tubular, my browser with the proper Ad-blocking in place, so I can live a comfortable life without GrayJay.


  • Don’t they? I’m aware of the events that took place when they released it. But seems they’ve solved licensing by now. There is a License.md in that repo since a few months. FUTO seems to even have written blog posts about their licensing, detailing why they do it. In short: They like to call it open source, while it’s not. It’s source available. Seems from their posts, they mainly want to exclude commercial use, but I’m not sure about their legalese, and the actual license text restricts how people can share and modify it. But the licensing is there by now. It’s just not an open source project. But I agree, they still like to confuse users and twist the meaning of words.





  • hendrik@palaver.p3x.detoSelfhosted@lemmy.worldStalwart mail server
    link
    fedilink
    English
    arrow-up
    4
    ·
    edit-2
    3 days ago

    I’ve tested it (on NixOS). But just for two weeks. I’d say it’s pretty impressive. Certainly works. It was just missing some important (to me) feature (forwarding mail to external mailboxes). But they’ve added it since, so I would like to try again. It doesn’t seem to have all the bells and whistles (and I didn’t have a look at the program code) but the basic features of a mailserver seem to be solid. I can’t really comment on the sustainability of the project, quality of the documentation… I mean if your setup includes niche edge-cases, custom tweaking and hooking into other software, maybe stick with the popular choice. But if you just want a regular, more or less simple mailserver, I’d say go for it.



  • I’m still a bit split on this. And whether the complexity and reliability is good enough for the use case… I mean if you don’t need N-out-of-M, but it’s just two people: cut a password or key in half. Same if it’s N-out-of-N people, you just need to make some puzzle pieces and hand them out, we don’t really need encryption and fancy maths for that. But I guess encrypting something would work, too. Just use a program or algorithm that’s likely still around when it’s going to be used. And you can always add a sheet of paper or PDF with instructions. Maybe save the executable file to to decrypt it somewhere if the solution requires software.


  • Nice. Thanks. Seems I’ve missed some Harry Potter themed stuff. That gave me an idea… Take (or write) an Arduino library (or SSS implemeted in plain C, instead of Go), flash it on a microcontroller like an ESP32 and you have some actual, physical horcroxes. I’d have to think about the form factor, and whether they need displays, or act as a USB thumb drive… But they could light up once you get like 3 of them in bluetooth proximity and reveal the secret. Other than that I think it needed to be part of some well-maintained password vault app. Or be a web service, so people don’t need to worry to get some old computer code running.

    Edit: Seems the Bitcoin people have had a thought at something like this: https://github.com/satoshilabs/slips/blob/master/slip-0039.md


  • Sure. I believe that could be done with minimal effort. Either by a smarthome solution, a script on a wifi router, a script in the autostart of the laptop someone uses every day, or like tasker on a phone. But you need to get it right. Or it’ll fire once you’re on a 14 day trip through Europe (and absent from your house and computer), phones can be lost or replaced… You might move… And you kind of want to make sure it’s robust enough so it actually works once needed, and that might be decades from now…