As the title says, imagine two observers with clocks A and B in space which move at vrel relative to each other. Just when clock A passes by clock B, the clocks counter shows zero. Observer A will conclude clock B to tick slower, while observer B will conclude clock A to move slower (both using data composed by an army of observers with clocks being at rest in their frame). Observer A decides to accelerate until he is at rest in the frame B is at rest in. When he accelerates (non-instantaneous), clocks that are towards B will time shift (tick faster than before the acceleration) depending on the distance, vrel and acceleration. Or the change of the inertial reference system he is at rest in if you want when at a distance to B. Now the question. Is it possible for Observer A to chose the appropriate acceleration depending on the distance and vrel, which allows him to make up for the slower ticking clock of B caused by moving at vrel to B, in such a way that he would conclude the clock of B to be ticking at the same pace as his clock until he is at rest in B's frame? He would have to accelerate stronger at close distance to B, and less at high distance as i see it. (might see it wrong) Of course, if that was possible, then observer A (now at rest in the frame B is at rest in too, but at a distance) could do the similar in respect to a clock in his current position when wanting to travel back to B (locally), and yet the clocks would be in sync still, even thought observer A traveled and was accelerating at all times with a non-constant acceleration chosen according to distance and vrel. Is this possible or did my mind trick me? I could try to draw this in a minkowski diagram, but it would take quite a bit.