This PR moves more core operators into core. This includes almost all of the operators of #550, minus some that I think we might want to prune instead (delay and broadcast). There are other operators that I think are worth retaining (partition stands out) but several of the remaining operators should probably be pruned.
This PR has stronger opinions than #550 on how little should be left for Stream (vs StreamCore) and they may need to be walked back once there is an alternate container implementation. I think for the moment there is some type inference ease because there is only one implementor, and various annotations that occur in #550 are not needed here (yet). So, perhaps some more convenience methods / trait re-thinks (ToStream stood out as hard to specify the container type, as it is a trait argument rather than a trait method argument).
This PR moves more core operators into
core
. This includes almost all of the operators of #550, minus some that I think we might want to prune instead (delay
andbroadcast
). There are other operators that I think are worth retaining (partition
stands out) but several of the remaining operators should probably be pruned.This PR has stronger opinions than #550 on how little should be left for
Stream
(vsStreamCore
) and they may need to be walked back once there is an alternate container implementation. I think for the moment there is some type inference ease because there is only one implementor, and various annotations that occur in #550 are not needed here (yet). So, perhaps some more convenience methods / trait re-thinks (ToStream
stood out as hard to specify the container type, as it is a trait argument rather than a trait method argument).