This is maybe a discussion we should be having elsewhere, but for lack of a dedicated thread somewhere, and because I think it's interesting, I suppose we might

This is maybe a discussion we should be having elsewhere, but for lack of a dedicated thread somewhere, and because I think it's interesting, I suppose we might as well keep going on the tangent. Moving forward though, is there anywhere dedicated to talking about this kind of stuff, or should we set somewhere up? > consensus algos for removal of data is an impossibility There may be a bit of a semantic barrier here. What I mean by consensus removal is "can we, as a society, decide that it's okay to delete this piece of data?" That can be done as a zero-knowledge algorithm. The approach I've taken with it is, loosely speaking, 1. Author Alice publishes (encrypted) data, tells server "don't delete this" 2. Alice shares with Bob; Bob also tells server "don't delete this" 3. Alice tells server "now it's okay to delete" 4. Alice can see that Bob is preventing deletion 5. Alice provides social/legal/political pressure to Bob to remove his hold 6. Object gets deleted Technically it's implemented in a way very reminiscent to name binding and garbage collection in memory-managed languages. It's worth noting that no, this isn't something that's solvable strictly with in-protocol actions. The external pressure step is necessary to prevent deletion. But that's exactly my goal; we're never going to be able to solve profoundly complex social problems with algorithms.