>> It's hard for me to see how the "twin paradox" is a paradox in any >> sense other than being a surprising result. > >Just study Einstein's solution and see if you agree - or if you smell >a dead cat. ;-)
I'm not sure what "Einstein's solution" is, but for any problem involving traveling clocks, the prediction of GR or SR is this:
elapsed time on clock = Integral along the path of square-root(|g_uv dx^u dx^v|)
In SR, we can always choose coordinates so that g_uv is constant and diagonal, with g_00 = 1, g_11 = g_22 = g_33 = -1. Then the above expression reduces to:
Integral along the path of square-root(1 - v^2/c^2) dt
In General Relativity, or in SR with noninertial coordinates, g_uv may in general vary from point to point, so there is an apparent position-dependence to the rates of clocks.
I really don't understand how the twin paradox is a consistency challenge for GR. The fact that Einstein himself may have worried about it doesn't mean anything to me. We're not dealing with holy scripture, and Einstein is not a prophet. GR is a theory that stands or falls independently of its creator. Einstein is not the last word on GR.
What do you consider to be the *real* twin paradox?