kernel-CPoint.menelaus_collinear_of_ratio_product
- Kind
- kernel-term
- Status
- checked
Supports: `CPoint.menelaus_collinear_of_ratio_product` is admitted with exactly the type in formal.statement. The proof rests on a polynomial identity derived symbolically BEFORE any kernel work and confirmed with zero remainder: cross X Y Z = (p*q*r + (1-p)*(1-q)*(1-r)) * cross A B C. MENELAUS NEEDS NO NON-DEGENERACY HYPOTHESIS AT ALL -- not even the ratio condition Ceva requires. When the hypothesis makes the left factor vanish, cross X Y Z vanishes REGARDLESS of cross A B C, the fully degenerate A = B = C included. Ceva's forward direction, by contrast, needs PosBound (D*D) k on the ratios where D := (1-q) + p*q. Two configurations, two different answers to what degeneracy matters; the second was checked rather than carried across from the first.
test "$(scripts/cargo-serialized.sh test -p axeyum-lean-kernel --lib creal_point::creal_point_tests::menelaus_collinear_of_ratio_product_statement_is_exact -- --exact 2>&1 | grep -Ec 'test result: ok\. 1 passed; 0 failed')" -ge 1 Evidence notes
A footprint check cannot carry this claim, and here the two candidates are a SINGLE `neg` apart: Ceva's ratio condition and Menelaus' are the same product equation with opposite sign, and both are axiom-free. Stating Menelaus with Ceva's sign would be false and no sweep in this repository would notice. The pin asserts the occurrence count of the negation BEFORE relying on it, because a substring guard for a shape that was never present reads as a dead guard rather than a red flag. Mutation-verified by the establishing lane: weakening the guard killed exactly one test, then reverted. The command is anchored on the exact `1 passed; 0 failed` count so an unmatched filter -- which prints `running 0 tests ... ok` and exits 0 -- cannot read as success.