starting out with an unpopular opinion: of all the centralized social media platforms, Facebook was always my favorite.

Why? it is the most full featured. Has threads, reactions, groups, “Pages”, polls, and it even has granular privacy controls (for hiding content from other users, not to be confused with Facebook’s privacy violations and commercial data use).

This makes me wonder, could we have a Facebook-like experience using Lemmy as a backend? similar to how lemmy has a phpBB experience using lemmyBB.

Lemmy already has threads, and communities can represent groups. Pages and user pages can be simulated with communities.

We would be missing polls and reactions, which I can live with. I am not at all mad that we would be missing content algorithms either.

Although we can’t make it identical to Facebook, I think it will get reasonably close and exemplify most of the good parts.

I am thinking to take this project on, but wondering if people have thoughts, if this already exists, or if people would even want to use this.

    • matcha_addict@lemy.lolOP
      link
      fedilink
      English
      arrow-up
      2
      ·
      10 days ago

      You’re right that lemmy never intended to be a Facebook mock-up (and I hope my post did not seem to allude to that), but I was wondering if it could be made to be so by changing only the client but keeping the rest intact.

      My limited knowledge makes me think that lemmy is easier to host, consuming less resources in general, hence why I want to consider it first. But I could be wrong.

      • wizardbeard@lemmy.dbzer0.com
        link
        fedilink
        English
        arrow-up
        5
        ·
        10 days ago

        Lemmy is easier to host specifically for the reasons that it wouldn’t make a good Facebook alternative. Hosting what is mainly text content and links is pretty lightweight.

        • matcha_addict@lemy.lolOP
          link
          fedilink
          English
          arrow-up
          1
          ·
          10 days ago

          What I am proposing would use the same exact lemmy backend. How would it magically use more resources with the same backend?

          What you said is incorrect, the difference in features offered is not the only reason for the difference in resource utilization. Lemmy meme communities have plenty of images going around.

          • wizardbeard@lemmy.dbzer0.com
            link
            fedilink
            English
            arrow-up
            3
            ·
            10 days ago

            Ok, so now you’re looking at Lemmy plus Pictrs (I think that’s the image hosting software most instances use on the backend, but I might have the name wrong).

            If you also want video support loke Facebook, that’s an additional load on the server for whatever you use. Probably piped?

            Maybe all of that together is still lighter weight than Friendica. I’d be surprised but it’s possible.

            So far we both seem to be going off of gut feelings and anecdotes, so if you (or anyone) has harder numbers that would be neat.


            Then there’s the workload of the modifications to the code, especially what would be needed to manage potentially nuanced visibility settings for each piece of uploaded content. That would mean that you wouldn’t only need to mess with Lemmy, but how each of the other services work as well.

            And then how do you manage federation while ensuring the visibility settings are kept?

            By the time you would have a proof of concept that had the basic features down, it would be significantly diverged from Lemmy anyway.

            You mention LemmyBB in your other response. I wasn’t aware of it before, but an “old school” forum is significantly more similar in function to Reddit/Lemmy than Facebook is. It’s just Reddit/Lemmy without nested comment replies, so everything is a top level comment (or displayed as one) in chronological order. Plus upvote/downvotes wouldn’t need to be displayed.


            None of this makes what you’re suggesting impossible, but you’ve got many additional layers of complexity right out of the gate compared to just using what already exists for the job at a resource efficiency hit.

            Would it truly not just be simpler to work on improving Friendica’s efficiency rather than rebuilding the wheel using layer upon layer of leaky abstractions and approximations?