[EPIC] DataFusion on Ray 0.1.0 release

[andygrove]

Manage donation

• Propose donation of Ray SQL into Apache DataFusion - Proposal to donate Ray SQL to the DataFusion Project (not into the Python subproject) datafusion-python#872
• Vote to accept donation - https://lists.apache.org/thread/2cff6k3jtzg8ob1x1qz22150swwnkqsn
• Complete the IP clearance process - https://lists.apache.org/thread/hytm7fl0z3c24sdhf9wbdvnpk2o5o9or
• Merge Donate Ray SQL to the DataFusion Project #1
• Update the README in the ray-sql repo to state that the project has moved, and archive the repo

Initial tasks once the donation has been accepted

• Update packaging and documentation to reflect the new project name and purpose (it is no longer a research project, for example)
• Upgrade to use DataFusion 41 (based on Updating datafusion to 41.0 datafusion-contrib/ray-sql#51)
• Get the Rust unit tests in planner.rs running in CI and update the expected plans under testdata/expected-plans
• Manually test on a distributed cluster to confirm that everything is working - Fix FFI errors, and add script for running TPC-H #20

Benchmarking

• Add documentation for running benchmarks
• Automate running benchmarks against PRs using compute infrastructure provided by @andygrove
• Update performance charts in README

Set up release process

• Add scripts for creating and publishing source releases
• Add CI scripts for building Python wheels
• Add CI scripts for building Docker images
• Set up RAT checks in CI

First Release

• Release datafusion-ray 0.1.0
• Write a blog post to announce the new project - Google doc: https://docs.google.com/document/d/18Yd10y9fhIKiAfUPifQk5BSFytOGV0tcbmnK2Xy7_Ro/edit?usp=sharing