Just from a performance perspective, when your feed is being retrieved, I can only presume that the more communities you’re subscribed too the more intensive the query is?

At some point does it not get prohibitive? Is there a limit on the number of communites one can subscribe to.

If true, I see this as a good reason for user created multi or meta-communities. This way, I can look at a feed of only a subset of communities at a time which should lessen the load on the server, and also be a better interface because I honestly don’t want all of the communities I’m interested in being fed into a single feed.

    • maegul@lemmy.mlOP
      link
      fedilink
      English
      arrow-up
      3
      ·
      1 year ago

      Good point! I should have thought about that … I just don’t use that feed. There’s a chance that it’s relatively efficient because it’d be the same for everyone and so is basically calculated once for the whole instance.

      • breadsmasher@lemmy.world
        link
        fedilink
        English
        arrow-up
        4
        arrow-down
        1
        ·
        1 year ago

        Once for the whole instance

        Not necessarily. I have a very long blocked communities list which is likely unique to me. I believe thats taken into account when pulling /all

        • RoundSparrow@lemmy.ml
          link
          fedilink
          English
          arrow-up
          2
          ·
          1 year ago

          Even the person to person blocking is taking into consideration with that query. Lemmy’s SQL output isn’t well optimized for caching, it puts a lot of work on PostgreSQL to customize these views. It is only users who are not logged-in that would get a repeatable query.