The Threadiverse is growing again, this time offering an effort to build a new Lemmy replacement from scratch, with API compatibility.

Seems like an interesting effort. A developer is building an alternative Java-based backend to Lemmy’s Rust-based one, with the goal of building in a handful of different features. The dev is looking at using this compatibility to migrate their instance over to the new platform, while allowing the community to use their apps of choice.

  • Margot Robbie@lemmy.worldEnglish
    01·
    10 months ago

    Having a frontend rewrite seemed more critical than trying reimplementing the backend in a different language.

    Remember, Lemmy had 4 years of development to iron out bugs, and this is essentially promising to make something in months that has a fully compatible backend to support all the third party apps, while adding features on top of what Lemmy has, and with a better front end with better mod tools to boot, with a complete rewrite of everything.

    The scope of this project has planned for is already unviable. Suppose that Sublinks does reach feature parity to the current version of Lemmy, congratulations, the backend or mod tools is not something a regular user is going to notice or care about at all, all they will know is that suddenly, there are weird bugs that wasn’t there before, and that causes frustration.

    And this project is going to get more developer traction because… Java?

    I’d like to be proven wrong, but I’m very sceptical about the success of Sublinks, because it look like a project that was started out of tech arrogance to prove a point than out of a real need, I don’t work in tech, but the general trajectory of these kind of projects is that “enthusiasm from frustration” can only take you so far before the annoyance of dealing with mundane problems piles up, and the project fizzles out and ends with a whimper.

      • wesley@yall.theatl.socialEnglish
        1·
        10 months ago

        Just peeking at the source code and it all looks like pretty standard stuff. Looks just like apps I’ve worked on at several jobs.

        Is it sexy? No. But a lot of people have experience with this and could help develop.

        Only time will tell if this project pays off though

      • Margot Robbie@lemmy.worldEnglish
        01·
        10 months ago

        I’m pretty sure Nutomic was a Java dev before starting work on Lemmy and learning Rust from scratch. That by itself should already speak volumes.

        One-Up projects like this rarely ever turn out well, that’s from my own experiences. Even though this isn’t a popular view, I still think I’m right on this one, we can circle back in say, 6 months, to see if my predictions are right.

        • originalucifer@moist.catsweat.com
          0·
          10 months ago

          it it common to announce a ‘major rewrite’ without having it complete?

          i mean, at the moment, theres little to discern it from lemmy at the moment… why make a big public proclamation about it before you even touch the front end?

        • nutomic@lemmy.mlEnglish
          01·
          10 months ago

          I’m pretty sure Nutomic was a Java dev before starting work on Lemmy and learning Rust from scratch.

          That is true, I used to be an Android developer and then learned Rust by writing code for Lemmy. Are you by any chance my new stalker?

          And if we’re comparing the languages, the fact alone that there are no Nullpointerexceptions makes Rust infinitely better than Java for me. I also agree that this sort of copycat project will soon be forgotten. For example have you ever heard of Rustodon?

        • Dessalines@lemmy.mlEnglish
          01·
          10 months ago

          I also was a professional java dev, and also had to use spring boot in most corporate environments.

          I don’t wanna knock anyone’s re-write, because I know how difficult it is to dissuade someone when they’re excited about a project. But to me, starting a new project or doing a rewrite, is the best opportunity to learn a newer, better language. We taught ourself Rust by coding lemmy, and I recently learned Kotlin / jetpack compose because I wanted to learn android development. Learning new languages is not an issue for most programmers; we have to learn new frameworks and languages every year or so if we want to keep up.

          This is potentially hundreds of hours of wasted time that could be spent on other things. Even if someone absolutely hates Rust and doesn’t want to contribute to the massive amount of open issues on Lemmy, there are still a lot of front-ends that could use more contributors.

    • spaduf@slrpnk.netEnglish
      1·
      10 months ago

      Lemmy had 4 years of development to iron out bugs

      Lemmy had 4 years to accrue technical debt and make foot-guns first-class features. A rewrite is probably exactly what it needs.

        • spaduf@slrpnk.netEnglish
          1·
          10 months ago

          I have and if I’m honest I’m probably a little bit too harsh. I think the bigger issue is honestly the priorities of the dev team. There’s good reason that this project is focussing on moderation tooling.

            • spaduf@slrpnk.netEnglish
              1·
              10 months ago

              Some things that seem hard to argue with:

              • A mod panel with things like ‘add moderator’ (maybe this could be attached to the new moderator view?)
              • Targeted reports (choose who receives it; admin/moderator)
              • Moderation actions on jerboa
              • Moderator edits. There’s a fine line here and I can understand why you wouldn’t want total edit capabilities but it’d be nice to at least be able to do things like mark as nsfw and add content warnings. This sort of feature should also probably target megathreads
              • Private communities (I know local only communities are in the works but there’s a whole mess of other criteria that would be useful)

              My own personal wishlist:

              • Karma requirements
              • First class wikis
              • Hashtags (I actually think a super simple stopgap solution here is to just have them link to the appropriate search page)
              • Flairs

              There’s some other stuff that I have seen PRs for and I do understand y’all are working hard. I appreciate the work you’ve done so far and the communities you’ve helped build. The Internet is undoubtedly a better place for it.

        • 0ops@lemm.eeEnglish
          1·
          10 months ago

          Magically no, but sometimes a clean slate is easier than a refactor. I’m speaking generally though, I’ve never looked at Lemmy’s code, and I’m not even who you originally replied to.