mleku @mleku - 11d
yeah, the "kinds with replacable semantics" is a retarded idea, all events should be replaceable, and the parameterised replaceables ... i mean, if you could just say "replace this event" then uh, what dis for? this is more than just a client problem kind 1 events are not "replaceable" because that concept accrues to a kind and an author, replacing events totally has not been designed in the back end, because they are dumbasses this is why i intend to eventually add a "replace" and "relay" method to the http API - then you can say "replace this" and "don't save this" explicitly instead of this voodoo bullshit
speaking of which... i can implement these now i just have to tangle with the somewhat messy websocket libraries and i'm done with my documentation/refactor adding the relay and replace methods would be pretty awesome for the time being, no clients are gonna support this but i'm pretty sure #gitcitadel and #alexandria need these features just like any normal fucking human being
gonna do the "relay" method first, because it is simpler, it just forwards it to the notify subscribers then i'll do the actual replace API call, and it will be easy to test it with using the expiring variant of nip-98 and curl
