Hi everyone, I’m having trouble finding a Lemmy instance that works well for me. The main instance I use is down, and most others are too slow. I’m wondering if there’s a way to choose an instance based on latency and the least blocked users. I found two relevant issues on the awesome-lemmy-instances GitHub page: issue #12 about choosing an instance based on latency and issue #17 about choosing an instance with the least blocked users. However, I’m not sure how to implement these into the main script to generate a readme with a few recommended instances. Does anyone have any advice or tips on how to choose a Lemmy instance based on these criteria? Thanks in advance for your help!

  • Jeena@jemmy.jeena.net
    link
    fedilink
    English
    arrow-up
    17
    ·
    11 months ago

    For those two specifically I’d suggest to host one yourself. You then are 100% in control over blocking users/instances and the latency is super short because you’re the only one on the server. That’s what I’m doing.

    • RxBrad@lemmings.world
      link
      fedilink
      English
      arrow-up
      1
      ·
      11 months ago

      What kind of storage requirements are there?

      It feels like you’d eat up a lot of storage, fast. Especially if you subscribe to anything video/photo-related.

      • Jeena@jemmy.jeena.net
        link
        fedilink
        English
        arrow-up
        3
        ·
        edit-2
        11 months ago

        With Mastodon yes because it caches everything on your server, but with Lemmy no, because it hot-links media from the other server without caching it.

        jeena@Abraham:~/lemmy/volumes$ du -sh *
        8.0K	lemmy-ui
        5.2G	pictrs
        2.8G	postgres
        

        I’m subscribed to around 50 communities for about 2 months.

        • RxBrad@lemmings.world
          link
          fedilink
          English
          arrow-up
          2
          ·
          11 months ago

          Huh. That’s surprisingly light.

          I self-host my own Mastodon server and relay about 50 hashtags (no full servers). Even with media getting flushed every 3 days, it still hovers around 20GB.

          I’m probably gonna do it now, because of course I will.

    • Dandroid@dandroid.app
      link
      fedilink
      English
      arrow-up
      3
      ·
      11 months ago

      That’s what I do. I’m basically immune to any defederation drama, as I can choose who to federate or defederate with myself, and I’m immune to any downtime (other than the time I accidentally nuked my MBR the morning before I left on an all-day trip and couldn’t access my instance all day because of it, but let’s not talk about that).

  • RotaryKeyboard@lemmy.ninja
    link
    fedilink
    English
    arrow-up
    4
    ·
    11 months ago

    Most people pick an instance by collecting a list of instances that are physically close to them, and then reviewing the /instances page to see how big the block list is. Lemmyverse.net can tell you how many users are on those blocked instances.

  • RoundSparrow@lemmy.ml
    link
    fedilink
    English
    arrow-up
    4
    ·
    11 months ago

    Something to consider… Lemmy runs fast with no data in it. Latency issue is tied to how much data they have stored in PostgreSQL. If they aren’t holding full copies of all the remote communities, sure it is faster, but your searches and All aren’t going to turn up much.

  • azuth@lemmy.world
    link
    fedilink
    English
    arrow-up
    3
    ·
    11 months ago

    I don’t think latency is relevant in an application like lemmy. If you can’t get posts from the server it’s not latency but the server being overloaded.

  • Philip@endlesstalk.org
    link
    fedilink
    English
    arrow-up
    2
    ·
    11 months ago

    If you have the ability then self hosting is the best option

    Otherwise I think instances with less users and content will be quicker, but then the all feed will probably be smaller.

    I will also recommend my own instance endlesstalk.org. It is very small(10 users) and the server is located in Germany. Hopefully that is close to where you live, so the latency is low.