1) Let's start with your failed notion of trying to defend space time of only light flashes as an "affine space". In an affine space, there is no distinguished point that serves as an origin. Hence, no vector has a fixed origin and no vector can be uniquely associated to a point.http://en.wikipedia.org/wiki/Affine_space Obviously, each light ray has a point of origin, hence, your assertion that an affine concept saves Minkowski space from failing. Now, I do not think you are a troll, I think you don't understand how this all works or you would not have attempted to shoehorn light rays into an affine space. Now, since you have retreated from defending space-time under the Minkowski metric as valid exactly as I explained it to you, let's review why. 1) The minkowski space is not a metric space. Therefore, it is not a valid geometric space. 2) The minkowski space is not a Hausdorff space. Hence, it does not have a definition of global continuity. As indicated by mathpages and me, along a light ray, the minkowski metric (distance) is always 0 from the origin. Anyone in this universe would disagree with any math definition of a space that does not support global continuity as does the minkowski metric. So, any reasonable person would refute this silly space definition. 3) The LT function does not preserve the light postulate. a) The light postulate requires any ray moves at c in the frame. b) The light postulate requires given 2 light rays with different y/z coords that are simultaneous in one frame will not be simultaneous in another frame. c) If a light pulse is emitted from the origins of two frames, each frame origin remains as the center of all light spheres. LT preserves 1 and 2, but does not preserve 3 since for any time interval, frame X views concentric light spheres, but the LT mapping on that time interval is not a set of only concentric light spheres. So, LT does not preserve the geometry of concentric light spheres. You have nothing left.