Having said that, is it really the end of the world if large Lemmy instances have ads to make up for any shortfall in donations? Otherwise, how are large instances expected to be sustainable long term, especially if they’re going to ever reach the kinds of traffic Reddit sees?

  • RuudA
    link
    English
    81 year ago

    The donations to Lemmy.world have been enough in the first 2 months to last at least 6 months. I’ve been getting donations for mastodon.world since November, combining those I have enough for the next year. I don’t see a problem yet.

    • Izzy
      link
      English
      1
      edit-2
      1 year ago

      Lemmy is new and exciting for a lot of people, but in 3-5 years from now I doubt the level of donations will keep up. Of course, I could be wrong. Maybe the cost doesn’t scale linearly with the user count and having 10 times the users but 2 times the donations will be sustainable.

  • Izzy
    link
    English
    61 year ago

    No single instance should get to the size of 100 million users if we are doing federation correctly.

    • RuudA
      link
      English
      51 year ago

      That’s true. (But I think 100k servers with 1000 users will be more expensive than 1 server with 100M users. But we don’t want to become the new R*ddit)

    • @[email protected]OP
      link
      fedilink
      English
      11 year ago

      So if the ideal Lemmy structure is a large number of medium sized instances, would you say there should be a mechanism (either at the API level or handled by clients) to randomly select a general purpose instance at sign up?