Residue from #305 / PR #505 (2026-08-29), deliberately left out of that PR.
PR #505 added a cargo test --workspace backstop to the rust job after finding rust/dl/tests/picker.rs (7 tests) and rust/dl/tests/terminal.rs (1 test) compiled but never run in CI. The rust-coverage job keeps its own hand-written suite list and is missing the same two suites.
Left alone on purpose in #505: that list caps a coverage number rather than what runs, and both suites drive a pty, so widening the coverage run is a flake risk that deserves its own decision rather than a rider. The question here is whether coverage should include them (and eat the pty flake risk under a coverage harness) or whether the omission should be recorded as deliberate next to the list so it stops reading as decay.
Context: the same silent-omission class as #305; the backstop now guards the rust job but nothing guards the coverage list.
Residue from #305 / PR #505 (2026-08-29), deliberately left out of that PR.
PR #505 added a
cargo test --workspacebackstop to therustjob after findingrust/dl/tests/picker.rs(7 tests) andrust/dl/tests/terminal.rs(1 test) compiled but never run in CI. Therust-coveragejob keeps its own hand-written suite list and is missing the same two suites.Left alone on purpose in #505: that list caps a coverage number rather than what runs, and both suites drive a pty, so widening the coverage run is a flake risk that deserves its own decision rather than a rider. The question here is whether coverage should include them (and eat the pty flake risk under a coverage harness) or whether the omission should be recorded as deliberate next to the list so it stops reading as decay.
Context: the same silent-omission class as #305; the backstop now guards the
rustjob but nothing guards the coverage list.