• Head admin @ lemm.ee, a general-purpose Lemmy instance
  • Creator of lemmy-ui-next, an alternative Lemmy frontend
  • Lemmy contributor

ko-fi

  • 6 Posts
  • 89 Comments
Joined 1 year ago
cake
Cake day: June 9th, 2023

help-circle



  • Regarding your question:

    Lemmy federation basically works by copying stuff from their source instance to all other federated instances. So if I write a comment on lemm.ee, other federated instances will get their own copy of my comment. They will also all know that the “authority” for this comment is lemm.ee.

    If an admin on another instance decides to delete their local copy of my comment on lemm.ee, then they are always free to do so (for example, some instances might want to moderate more strictly), but any actions they take like this are limited to their own instance - for the rest of Lemmy, lemm.ee remains the authority for this comment, so individual remote instance admins taking actions won’t have any effect on any other instances.

    As for the original topic of modlog federation, basically it just boils down to this: just like with the comment example above, Lemmy instances also save a local copy of incoming federated mod logs. The Lemmy software does not yet have 100% coverage in terms of federating mod logs (for example, there are no federated logs yet for instance admins banning remote users), but this coverage has been increasing, and I expect this will eventually get to 100% (just needs more dev time really).

    Also, if some instance admins try to tamper with their mod logs, then other instances can still see the real history, because there is no way for an instance admin to delete copies of their mod log from other instances.



  • Most actions federate, any exceptions which aren’t federated yet are generally just there because the federation logic has not been implemented (but improvements are constantly being worked on).

    Generally federating the modlog is mostly just there for informative purposes. As in, we can check what mod actions were taken on instance A through the modlog on instance B (and there is no mechanism in Lemmy for other instances to retroactively remove or hide federated modlog items, btw).












  • I think community discovery can (and should) be improved for sure!

    Currently it’s true that you can use topic-centered instances for this, I do this myself as well, but I do think it has quite significant downsides in terms of creating pockets of centralization. For example, if you’re a user who is ONLY interested in french cinema (or any specific topic) on Lemmy, and all of the related communities and other invested users are on a single instance, then for you, the experience is absolutely no different from any centralized platform - the french cinema instance admins have 100% control over your Lemmy experience.


  • IMO, in practical terms, 3 key things should imapct instance choice:

    1. Basic instance rules (including things like community creation policy, nsfw allowed, etc)
    2. Federation policy
    3. Instance infrastructure (hardware & how it’s managed)

    Content specialization really shouldn’t matter IMO, because as long as the federation policy is OK for you, then you can participate in any communities, regardless of what instance they are on. In other words, even if you’re super interested in french cinema, there is no need to centralize all users interested in this topic on a single french cinema instance. Thanks to federation, users from all instances (accounting for federation policy) should be able to become fully fledged participants in any french cinema communities.

    Of the points I listed above, #1 and #2 are easier to include in an instance introduction, I’m not sure how to properly and reliably reflect #3 in any kind of overview. At the end of the day, I think most users tend to figure out their long-term home instance a while after they first join Lemmy, and quite often, it’s not their original instance, so maybe it’s not that important to emphasize the initial instance choice too much?


  • If I have several backends that more or less depend on each other anyway (for example: Lemmy + pict-rs), then I will create separate databases for them within a single postgres - reason being, if something bad happens to the database for one of them, then it affects the other one as well anyway, so there isn’t much to gain from isolating the databases.

    Conversely, for completely unrelated services, I will always set up separate postgres instances, for full isolation.