A thought experiment I was pondering - Say there's a photon source and two observers, one at rest and one in motion relative to the photon source.
It would seem that at the point the photon is emitted, it 'knows' where it will be observed, because it will travel at c for the particular observer.
I was thinking that this might be because the wavefunction for the photon hasn't collapsed until it is actually observed. Does that mean that the wavefunction of a photon contains the possibilities of all paths including all observers at the endpoints of each path, and is that any different to saying the wavefunction includes all paths?