We'd like to make some combination of language and tooling improvements to better support serialization & data modeling.

More details to follow. We'll be asking for input on what people want to see--those interested can watch this issue for updates.