we fully respect it. 🫡🪼

How did you solve for "Relays MUST ensure the deleted events cannot be re-broadcasted into the relay."? Seems like a burden.

jellyfish uses immortal as it's software. we have 2 layer of protection for this issue which help us to stop double processing of an event, stop rewriting of deleted and vanished events. 1. a redis bloom filter which keep track of all events seen by relay. so if someone send an already processed event, we will reject immediately. so, a deleted event is a event that were processed in past! 2. we don't delete a record on mongo db, we update it and remove all fields of it except its id. so it's basically just a nonsense hash. when someone try to write it again, (if we lost redis data for any reason.) we don't write duplicate ids on database and write operation will fail. you can take a look at code:

GitHub

GitHub - dezh-tech/immortal: An advanced Nostr relay implementation.

An advanced Nostr relay implementation. Contribute to dezh-tech/immortal development by creating an account on GitHub.