Skip to main content


Should I hack Nostr support into Pleroma just to enable zaps?
This entry was edited (9 months ago)
@realcaseyrollinsRebased, of course. But then somebody will just copy my commits to Pleroma and then Akkoma will reject it on the basis of Alex Gleason.

@eb2881406ad19ba7cf01c210ce002f4fe53e8ce6d84e77df5a2319f9f00a8005The main reason to build an entirely new backend is that "mix phx.server" takes 2.5 minutes. I leave and got coffee. Change one line of code, wait 60 seconds to recompile. Dammit, that didn't work, and type checking in my editor isn't very good so it didn't tell me. I change a line of code, wait 60 seconds, and dammit it still didn't work. I decide to just write unit tests instead. So I spend 5 minutes writing tests, run "mix tests", and then wait 5 minutes for the entire Pleroma suite and its million dependencies to fully recompile in test mode. I fiddle with tests for 10 minutes. And there we go, the API now says "hello world"

By contrast, I run "deno task dev" and the devserver starts immediately. I change a line of code and the server restarts immediately by itself. I make an error and the type checker in my editor catches it before Deno does. I just don't want to kill myself every second of every moment developing on this platform.

@eb2881406ad19ba7cf01c210ce002f4fe53e8ce6d84e77df5a2319f9f00a8005 meanwhile half of my pleroma hacks are written from a second node connecting to the main one, and live reloading the module in the repl, with no downtime

would be nice to replace my own Pleroma.Mostr.Relay with a proper translator though