Allow message compression automatically, per specific serializer id

[ktoso]

Could be useful esp if users do plenty repetitive JSON data messages?

[hassila]

It is also potentially useful for distributed cluster with low-bandwidth (relatively speaking) between different cluster sites (for me a common deployment scenario). Basically a set of "sub clusters".

https://github.com/lz4/lz4 perhaps?

[ktoso]

Yeah, in general message compression can be interesting.

Currently the cluster offers working with Codable and protobuf but in a weird hacky way (pretending to be Codable).
It does not make much sense to compress proto messages I think, but if one uses json then it definitely sounds compelling...

In reality we'd like a Codable or other message encoder that uses CBOR or a similar binary efficient format.
We are not planning on working on one right now, we're focused on the language and shape of the library for now -- enabling swapping the encoders in the future.