Login
You're viewing the discuss.systems public feed.

Replies

  • Oct 15, 2026, 4:21 PM

    @owen When we got it spun up, the weapon of choice was called Mjolnir. Today, you want to look at its successor Draupnir.

    You need to know a _little_ about how Matrix communications are structured and work, but relatively little. Single-action org-wide banning, per-user or per-origin, is a delight, and retroactive redact was something that a few free-as-in-speechies objected to until the day I had to clean up a goatse flood, they came around pretty quick after that.

    github.com/the-draupnir-projec

    💬 1🔄 1⭐ 6
  • Oct 15, 2026, 4:32 PM

    @mhoye Thanks.

    Going off a quick read, it seems like a lot of its underlying thesis is in the direction that if you have to moderate someone in one spot, odds are they've been up to antisocial nonsense in several, and you might want to act on them without having to individually find and adjudicate them. That tracks - but am I reading it right?

    💬 2🔄 0⭐ 4
  • Oct 15, 2026, 4:33 PM

    @mhoye (The background radiation of most moderation on the internet is to treat each incident as a one-off; tooling for addressing repeat customers is frequently weak)

    💬 0🔄 0⭐ 4
  • Oct 15, 2026, 4:41 PM

    @owen During my full-time involvement, it was still true that moderators had to generally keep track of disparate acts across forums, and communicate appropriately.

    Early on, this was a lot of administrative work for me - most abusers get away with it by being perfectly nice to most people - but after a while I'd built up enough credibility that people came to value the safety of their colleagues in other fora very highly.

    💬 0🔄 0⭐ 7