Everyone seems to use the slope definition of stability for vector bundles without making any mention to the fact that this should be the correct definition describing that a stable equivalence class of vector bundles is then equivalent to a stable orbit under the action of the gauge group on say, the space of all connections.