Looking back over my previous posts, I found a couple of errors that need to be corrected:
PeterDonis said:
a curve going from t = 30 seconds on the Earth's worldline to t' = 30 seconds on the missile's worldline would be timelike, not spacelike. In fact it will be timelike for a missile traveling at any speed fairly close to that of light (off the top of my head I think all that's required is a gamma factor of 2, which requires a missile speed of 0.866c).
I was wrong here. If the Earth and the missile both set their clocks to zero when the missile is launched, then the two events "Earth clock reads 30 seconds" and "missile clock reads 30 seconds" (each event located on the appropriate worldline, Earth's or missile's) will *always* be spacelike separated. So I was wrong to say that a light pulse launched from Earth when Earth's clock reads 30 seconds could ever reach the missile before the missile's clock reads 30 seconds.
However, I was still correct in the original example (when I assumed the missile's speed was 0.99c) when I said a light pulse launched from Earth when Earth's clock reads 30 seconds would reach the missile before the missile hits Tau Ceti (at a distance of 12 light years in the Earth frame). That's because if the missile's speed is 0.99c, the missile will take a lot longer than 30 seconds, by its own clock, to reach Tau Ceti. That is, I was wrong to assume that the missile's time of flight, by the missile's own clock, would be only 2 seconds if the missile's speed was 0.99c.
So I also got this wrong:
PeterDonis said:
in the missile's frame, the time between launch and the President issuing the order is *much* less than 30 seconds; in fact it's 30 seconds divided by the time dilation factor, which is something like 10^8)
The time dilation factor for v = 0.99c is only about 7, so the time between launch and the President issuing the recall order, in the missile's frame, is 30/7 seconds, or about 4.3 seconds. That doesn't change the rest of my conclusions; the light pulse will still catch the missile well before it reaches Tau Ceti (since, as I said, that result is an invariant and I derived it in the Earth frame without using any values in the missile frame). The missile's time of flight, by its own clock, will be the time of flight by Earth's clock, 12/.99 years, divided by gamma = 7, or about 1.7 years; and 4.3 seconds is still *much* shorter than that, so in the missile's frame, Earth is still much closer than Tau Ceti when the light pulse is launched.