Great, I didn't know that! I can see that there's indeed some space for providing other Encoders with different outputs. I haven't seen it applied to a direct ByteString encoding though. I suspect that this might change the Encoder interface a bit to be closer to what is presented in my post and in aeson. Actually even with aeson I don't understand why list is not taking Series as a parameter: hackage.haskell.org/package/aeson-2.2.1.0/docs/Da…