The last clause of your Wikipedia quote is another repetition of what I've been saying this whole time, so Wikipedia is fine with me I guess. Note that the "se

The last clause of your Wikipedia quote is another repetition of what I've been saying this whole time, so Wikipedia is fine with me I guess. Note that the "separated in space" requirement doesn't require any particular distance scale. A moving observer could disagree with you about simultaneity of events on opposite ends of your desk. And yet, coming back to the original point, we don't need to worry about the infinity of hypothetically possible moving observers to notice that we and a not-too-distant galaxy have approximately the same reference frame, and thus can approximately establish a shared timeline. That's plenty to make it meaningful to say we're looking at things that happened 20mya. (Please note that I was explicit about this being approximate in my first reply.) Hell, we don't actually have to care about the galaxy's frame at all. Even within our own reference frame, we can calculate what nearby events we observe to be simultaneous with distant events we see (not observe) now, and it'll still be basically the naive time-in-the-past estimate based on light travel time. The worst you can say about that is that it's not wrong, right? We don't need to consider all possible frames before giving a time in the past of our own frame. > But a moving observer could perceive _our_ now as occurring before the events that we now observe in light from the hypothetical 20 million light year distant object. Actually, if by "our now" you mean after we've received the photons from the distant event (or really even the same time), I'm pretty sure that would be a superluminal observer. So, uh, lol no. It's kind of important that no real reference frames disagree about causal order. Also, quit telling me relativity is "unintuitive" like I don't realize that. I've literally done the math on this.