I have thought about some questions for a while. Two results pop up suddenly after fermenting for a long time.

The first is about the amount of light bending by gravitational force. In 1804, Soldner published a paper on the amount of light bending by gravitational force. In 1911, Einstein got the same result. In 1915, Einstein modified the result. The amount of light bending is doubled. Can Soldner's result, derived from Newtonian mechanics, be modified? 

Electromagnetic waves comprise two types of energy: Electric energy and magnetic energy. Each contains half of the energy. Electric energy is potential energy. Magnetic energy can be thought as kinetic energy. The total energy is mc^2. Kinetic energy, as half of the total energy, is 1/2mc^2. This is classical physics, not consistent with relativity. But if we think in this way, the amount of energy interacting with gravitational force is halved. Consequently, the amount of light bending is doubled. A modified Soldner's method can give correct answer. 

People may immediately point out that the presentation is not consistent with relativity and hence it is wrong. But relativity should not be the ultimate arbiter in research.

The second is about fine structure constant. It is quite a mystique in quantum theory. Let fine structure constant be called alpha. I figured out that alpha^2 in the micro world is equivalent to GM/(rc^2) in the macro world. Some years ago, I derived an equivalent of Planck's constant in the macro world. With that equivalence, I derived alpha^2 is equivalent to GM/(rc^2).   GM/(rc^2) is a very important quantity in macro world. Specifically, it is half of the formula representing Schwarzschild radius in general relativity. By establishing a correspondence between micro and macro world, it makes fine structure constant a much less mysterious number. Indeed, it gives a very clear understanding of the physics related to fine structure constant.