nx_fissure_gate.nx

nx_fissure_gate.nx -- GATE for aesthetictwin rung AT16: the PALPEBRAL FISSURE, so that a canthus exists to be measured at all. WHY THIS RUNG IS FIRST. AT4 (canthal tilt) ranked #1 on this board for weeks and was NOT BUILDABLE: a canthus is the junction of the upper and lower lid margins, and the face carried a socket, a globe and a lash with no lid at all. Worse, the obvious proxy was provably vacuous -- the socket carve and the globe are axis-aligned and share cx exactly, so their intersection is mirror-symmetric, both ends land at the SAME y, and a tilt read off it is identically zero for every face forever. An axis that cannot fail is not an axis, so shipping fa_canthal_tilt against that geometry would have lit the board green with nothing behind it. THE LOAD-BEARING TOOTH IS T4 AND IT IS THE ONE THAT COULD NOT HAVE PASSED YESTERDAY: the two canthi must sit at DIFFERENT y. That is exactly the assertion the old geometry made impossible, so it is the tooth that proves the prerequisite actually landed rather than that a name was written. T5 pins its SIGN -- lateral above medial, a positive canthal tilt -- because a non-zero number of the wrong sign would still pass T4 while describing an anatomy nobody has. T7 IS THE NEGATIVE CONTROL AND IT IS THE PRE-AT16 FACE ITSELF: dropping the part count back to 27 reproduces exactly the model that could not be measured, and the ruler must refuse BY NAME rather than return the most flattering reading available. A measure that answered anyway would be reading its own parameters back out.