I can see how this better abstracts some stuff out but it makes api discoverability quite a bit worse. I’d rather create a node and end up with no op interface implementations. Maybe there were performance reasons for this but it does break with convention in a pretty bad way.
Hi, can you elaborate on why it makes API discoverability worse, and what you mean with ending up with no-op interface implementations?
Are you referring to having to implement all the various IUpdate etc?
Yes, this is the reason we’re doing this.