Hey /sci/, I don’t have an education physics, but I heard about the one way speed of light problem, and I tried to solve it. I sent my attempt to a physics professor, and he gave me this explanation for why it was wrong, which I am too dumb to understand. He also linked two papers which I had more difficulty understanding.
Could someone explain more clearly why I am wrong? Thanks.

>The “one-way speed of light” problem has been around for a while (see https://en.wikipedia.org/wiki/One-way_speed_of_light, or https://www.youtube.com/watch?v=pTn6Ewhb27k), and it’s true that to date there has been no substantiated experimental measurement of the one-way speed (though there have been many other tests of one-way dispersion, equivalences of synchronization conventions, etc.)

>The primary problem with your proposal is the same problem in the “circular twin paradox” (I’ve attached a couple versions for explanation). Even if one of the twins is at the center of the circle, we cannot immediately apply any of the rules of special relativity (even using anisotropic light speed) because the orbiting observer is not in an inertial frame (they are centripetally accelerating). Thus (as your first footnote mentions) we need to either use general relativity to handle the acceleration or we need to be extremely careful about each infinitesimal transformation of frames as B goes around A (made all the harder since we’re assuming light speed is anisotropic, but doable [https://en.wikipedia.org/wiki/One-way_speed_of_light#Generalizations_of_Lorentz_transformations_with_anisotropic_one-way_speeds]).

>The end result will be that B’s clock will have varying time dilations around the circle, so it won’t measure 10 minutes for each quarter with respect to A. In order to aim the pulses to be properly intercepted by B, A will have to account for these time differences anyway, which will still require the synchronization of their clocks using the two-way speed!