I'll describe the problem in my own words. The co-accelerating ships problem: -There are two identical spaceships which are initially at rest in "lab-frame". -These two ships perform IDENTICAL acceleration procedure. -The ships are launched simultaneously in the lab-frame. Spacetime diagram - LAB FRAME Red ones are the space ships. Right one is the "front ship". Blue line is an observer who stays at rest, added in there just for convenience. The acceleration events are instantaneous here, but you can imagine any sort of real-world acceleration instead) Now, since the accelerations are identical in lab-frame, it should hold true that the distance between the ships does NOT change during the acceleration; As measured by the lab-frame, the distance between ships before and after the acceleration is exactly the same. Obviously in the mechanics of SR, the acceleration procedures are NOT identical from the point of view of the ships. The distance between the ships should stretch in their own POV: From the "POV" of FRONT SHIP: Immediately after acceleration, the front ship exists in such an inertial frame that the rear ship must not have been launched yet. If you assert it has, then the front ship will receive information about the launch of the other ship at speeds slower than C From the "POV" of the REAR SHIP: Vice versa happens. Immediately after the acceleration, the rear ship exists in such an inertial frame, that the front ship must have been launched much earlier than the rear ship. If you assert it wasn't launched earlier, then the rear ship will receive information about the launch of the front ship at speeds faster than C. POV is used here to refer to how things "are" in the inertial coordination system of an observer, according to Lorentz-transformation. POV is NEVER used in the meaning of what the observer can SEE; This talk concerns Lorentz-contraction. In other words, the front ship will accelerate away from the other ship while it is still sitting on the launch pad. Here's what it all looks like from the inertial coordination system that the ships will end up to after acceleration. Spacetime diagram - FRAME OF SHIPS AFTER ACCELERATION Black lines are planes of simultaneity Even though the ships did go through the same acceleration procedure, after the fact these procedures exist in different moments in time. We should arrive at the same conclusion even if there is a steel rod between the ships. And that's not all. Since the ships are performing an identical acceleration, they should stay in the same inertial coordination system at all times. In other words, their notion of simultaneity should be identical at all times. In other words the ships (& the rod) should keep their length from their own perspective, and contract from the perspective of the blue observer. So it seems that SR very concretely requires that from the POV of the front ship, the rear ship must still be on launch pad after the launch, AND it must be in the same inertial coordination system at all times (i.e. NOT on the launch pad). There is odd asymmetry in Lorentz-contraction. It occurs to external objects when you switch inertial frames. But at the same time, it doesn't seem to occur if it is the external objects that are switching frames. And if the observer has any volume, he should stretch by the same mechanic that causes contraction (At least if he is being accelerated from the front and from the rear simultaneously). --- When I first asked about this, it was noted that there is no standard way to construct a coordinate system where non-inertial observer is at rest. However, if we place the launch pads at far enough distances from each others, and/or use rapid enough acceleration, it is trivial to show that AFTER the acceleration the front ship can exist in such an inertial frame, that the rear ship must not have been launched yet, IF it is true that light approaches the front ship in this new inertial coordination system at speed C. I was also informed of a FAQ-page regarding this problem. I didn't understand the explanation, if any was even offered: Physics FAQ - Bell's Spaceship Paradox They first give instructions to draw similar spacetime diagrams as I have offered above, but with non-instantaneous acceleration. Then they note: It doesn't make any difference who thinks the accelerations are constant and who doesn't. All that matters is that the acceleration procedures themselves are identical. Surely one acceleration procedure looks the same when performed in any location of the lab-frame. Then they say: What does this mean? Just pick the curve to be something else? Just decide that of the two identical accelerations, the other one is not identical? What is the actual explanation?