near-team oxidation plans

[cgwalters]

This project has existed for a long time, and we've gone through a lot of cycles where we've added major features, like lockfiles, apply-live etc.

Our slow conversion to Rust is an important aspect of reducing technical debt accumulated from these features as well as allowing us to confidently add new features.

This issue will track near-term plans for more Rust:

• WIP: Start rewriting origin code in Rust #2388 will help rethinking origins #2326

This will also lead into touching rpmostree-core.cxx and ideally removing the "treespec" abstraction in favor of a subset of the treefile that's shared across client and server.

Another major one if we can pull it off will be switching more of the main.cxx code to Rust and making it just work to write new CLI entrypoints in Rust.

[lucab]

One more target (with rationale and reverse dependencies) noted at #2719.

[cgwalters]

OK, the origin stuff turned out to be a lot more entangled, this is like my 3rd attempt at it that improved things but stalled.
For one thing I realized that treespec is entangled in the extensions code too. And refactoring all this is very fragile - it's mostly covered by tests but it's easy to break things.

[cgwalters]

Closing this in favor of more targeted issues.