[McKay (OP)]

I understand that throwing matter from wall to wall inside a moving box, perpendicular to the boxes motion vector, would be exactly the same as throwing matter from wall to wall inside a stationary box (stationary relative to what, anyway??) because all the matter inside would have [whatever velocity it has]+ velocity of the box. But light/photons, if launched from wall to wall, perpendicularly to the motion vector of the box ..  hmm?
I started thinking about this and realized I know nothing when thinking about time dilation and how photons are usually animated to take a longer path trough space between, say, two plates moving trough space, making triangles, but, as the photon has a constant velocity and cannot change(in vacuum, of course, sorry),  everything else must bend around that fact and thus time dilation is born. Well, at least, that's how it is usually illustrated. Pop-science, eh?
From what I understand, there are two scenarios:
1) the photons "fall" or, at least, bend towards to the bottom of the moving box, if the box is moving fast enough
2) a very, very  fast spaceship moving with v=0.9(9)c , starting 1 light year away in x axis and 1 ly away in y axis (sqrt(2) ly in total) shooting a laser pointer perpendicularly to the ground where an observer is standing on (assume the ground is not a planet and is not rotating and is in such a way that it will always be perpendicular)  - since the laser pointer is aimed perpendicularly down to the ground, after roughly 1 year the spaceship with its laser pointer will be right above the observer, 0 distance on the x axis, and the observer will see the laser pointers 1 year (roughly) old light for the first time, but it would appear to come from the current position of the spacecraft, not the original 1 ly (x axis) away. (?) But that is not what supposedly happens, is it? In actuality the laser light would be detected one more year later, when the spaceship is (roughly) 1ly past.

I know nothing.

[yor_on]

What would that box be made of? And how would it reflect 'photons'? Are you imagining a perfectly elastic box? in where 'photons' bounce around?

[jeffreyH]

I understand that throwing matter from wall to wall inside a moving box, perpendicular to the boxes motion vector, would be exactly the same as throwing matter from wall to wall inside a stationary box (stationary relative to what, anyway??) because all the matter inside would have [whatever velocity it has]+ velocity of the box. But light/photons, if launched from wall to wall, perpendicularly to the motion vector of the box ..  hmm?
I started thinking about this and realized I know nothing when thinking about time dilation and how photons are usually animated to take a longer path trough space between, say, two plates moving trough space, making triangles, but, as the photon has a constant velocity and cannot change(in vacuum, of course, sorry),  everything else must bend around that fact and thus time dilation is born. Well, at least, that's how it is usually illustrated. Pop-science, eh?
From what I understand, there are two scenarios:
1) the photons "fall" or, at least, bend towards to the bottom of the moving box, if the box is moving fast enough
2) a very, very  fast spaceship moving with v=0.9(9)c , starting 1 light year away in x axis and 1 ly away in y axis (sqrt(2) ly in total) shooting a laser pointer perpendicularly to the ground where an observer is standing on (assume the ground is not a planet and is not rotating and is in such a way that it will always be perpendicular)  - since the laser pointer is aimed perpendicularly down to the ground, after roughly 1 year the spaceship with its laser pointer will be right above the observer, 0 distance on the x axis, and the observer will see the laser pointers 1 year (roughly) old light for the first time, but it would appear to come from the current position of the spacecraft, not the original 1 ly (x axis) away. (?) But that is not what supposedly happens, is it? In actuality the laser light would be detected one more year later, when the spaceship is (roughly) 1ly past.

I know nothing.

You need to understand the Lorentz Factor

http://en.wikipedia.org/wiki/Lorentz_factor

[alancalverd]

Mirrors. Never mind bouncing around, what happens to one photon reflected once?

[jeffreyH]

Mirrors. Never mind bouncing around, what happens to one photon reflected once?

What is your view of photon reflection?

[yor_on]

It's a interesting question, in so many ways. Does a photon 'bounce'?

[alancalverd]

1) the photons "fall" or, at least, bend towards to the bottom of the moving box, if the box is moving fast enough

Moving relative to what? Not the photons, clearly, as they have no knowledge of anything outside the box. And not relative to anything outside the box since that can't affect the photons inside.

[yor_on]

If you define the box as having a uniform motion, close to light from some 'inertial', far away observers definition. Then place a observer inside that box, at rest with it. Would that observer now find light inside it to be blue, relative red shifted depending on direction . And what about light perpendicular (At right angles) to the overall motion of the box, as defined (the box overall motion) by our far away observer?

[yor_on]

the question can be reduced to if uniform motion exist, and it does, it's real and can be of different magnitudes, speeds. But inside that room you won't find find light blue respective red shift due to uniform motion, although you can find it blue shift relative some other source outside the room, assuming that you've changed the rooms uniform speed to another, via some acceleration, whereas the other, 'outside' source, haven't.

so there is no change inside as I know of. Then there is the light perpendicular to the motion, I don't expect it to 'bend'. If it did you could imagine the box , still moving uniformly but now infinitely close to lights speed, in which case you now would have reason to suspect a very strong gravitational field inside that (black) box, at the same time as you are in a 'free fall', finding no gravity acting upon you. The elevator experiment is about a 'stationary' extern (far away or not) light source, propagating through through a 'window' in the elevator wall, 'bending' the light according to the outside observer.

If the light is released inside, at rest, with the box there is no bending, and no blue or red shift. You need an external 'far away' observer for this to be seen. and so it won't matter what speed that far away observer define, for the observer inside that box, as long as all inside are at rest with the box, at a uniform motion.

[yor_on]

So all uniform motion is relative, locally defined, and you can exchange that for non existent if you like. But it is weird as you can prove different uniform motion, when comparing between suns for example. We can find them to have different 'speeds' relative us, and so we define different uniform motions to be existent.

What I think is important there is to note that it is about two different views though. The one using astronomy to define different uniform motions is 'comparing' between 'frames of reference'. Or better expressed: He's using at least two suns, comparing them to Earth, relative his local clock and ruler (which then becomes a comparison between three objects) so proving different uniform motions to exist.

The one locally measured (inside a black box) is 'at rest', within a same 'frame of reference' (loosely described).  This frame is also what repeatable experiments use, all measuring relative your local clock and ruler. We define them as repeatable when we get consistent, equivalent, results from others, at other locations in space and time, doing the 'same' (local) experiment.

It's also the frame that defines what we call constants, and there we find that even when measuring lights 'propagation' between frames of reference it still gives us a same value, all relative our own clock and ruler. There it won't matter if we set up three observers in different uniform motion, measuring on light 'propagating' from the same far away mirror (two mirror experiment). All of them will agree on 'c', and as this light, ahem, 'bounced' from the same position (mirror) you either define three types of time dilation 'simultaneously existing'  at that mirror (depending on different uniform motions for your observers, relative that mirror) or you accept that some things is related to measurements between frames of reference.

to see the difference there one has to remember that each one of our three observers only measure at one object. The proof for different uniform motion uses several observations, relative ones local clock and ruler, then comparing those results. The three observers above do one observation each, all finding 'c' in it, then compare their results to find time dilations related to a 'common source', meaning the same far away mirror.
==

Hard to make it make sense, at a second read here So I added some stuff.

[yor_on]

If we now get back to that box we discuss. If we agree on that a inside observer neither will see light bend (perpendicularly), nor will find it blue respective red shift. Then the 'energy' it has is the same, no matter what uniform motion, and energy, the far away observer might define looking at the box, and light inside the window. It's no different from you meeting a car, being on the hi way. Depending on your speed it will seem to pass you slower or faster, with less, or more, 'energy'.

what is interesting to note though, is that depending on how you define a universe, as a 'container model', or not. You in the first case will have to define a energy, and actually 'absolute motion' for that energy to exist, although it (still) won't be measurable (existent) locally. That is interesting because you can, if you like, avoid a container. Instead using those local definitions as being 'equivalent', as repeatable experiments and constants actually do. They use local experiments, what I call 'locality'. Doing so you still need some mechanism connecting those local 'points' to each other though. and maybe that will demand this 'energy' to exist, in some particular way.
==

In fact: When you see people use this 'light box' the way I do, then define it as if the photon 'knows' the motion and so will blue respectively red shift, and so present different energies colliding with a wall, they also presume motion to exist, (which I do too:) but further more they define it to be 'absolute' in all frames of reference, which I don't agree too. I don't think Einstein defined it that way either.

you can't have it both ways, define motion as locally non existent for a uniformly moving frame at the same time you expect a photon inside that frame to 'know' it is moving. Moving relative what?

[yor_on]

There might be a third way, if we assume that this light 'propagating' can 'know' it comes from a moving light source. How it would do that, and correct itself for three differently moving (uniformly moving) observers, all measuring on a same mirror, I don't know? Such a reasoning will give us 'absolute motion', although in very weird way, and also, as I suspect, a definition of a sort of 'container universe' in where you then can assume a defined amount of 'energy' existent.
=

This third way would then demand of this light to not blue or red shift inside that black box we discussed, no matter what 'speed' a far away observer want to define to it. So it's a truly strange idea, maybe

Or 'it's a field' in where we have rules and regulations, allowing light to 'know', without needing to show an increase/decrease of 'energy' for the local (inside) observer?

[yor_on]

But there is a caveat to the three observers 'two mirror experiment', using the same far away mirror to 'bounce' their light. Uniform motion is real, and will give a real time dilation, as I see it. So when I write "or you accept that some things is related to measurements between frames of reference." I don't mean that uniform motion doesn't have an impact. You can take that and the idea of measuring different uniform motions, using three suns for example, as two proofs of there being a 'real motion' even when uniform. But it is still about two views, one about that 'container universe', the other about a strict locality. Both exist, and we use them mixed. A little as physics treat light as waves first, then call it photons, then let it become a wave again. Myself I think it's worth thinking about what differs them though.

[yor_on]

Actually, describing it as I do 'time' seems to become a result of frames of reference, but it can also be connected to a local constant 'c'. And that's where QM becomes really interesting, as well as any ideas of a discreteness. Because using frames of reference my way, there either will be a discreteness, or it still will be a discreteness (as I think then:) although now as a result of a 'flow'. And the last one is closely related to how one might want to see lights duality, and the ways linearity and non linearity are able to exist, 'intertwined'.
=

My way looking at it is from locality. In there your clock always 'tick the same'. any time dilation found will then be related to a comparison with another frame of reference, not you aging 'slower' or 'faster'. It's just a relation between two frames giving you that impression. It's simple and fits all evidence I know of. You will never become immortal, locality will get you

[yor_on]

If you use the last definition of time, strict 'locality' then any and all time dilations has to be placed between frames of reference. What is strange with this notion is that you theoretically then has to assume that the same will hold for all frames of reference, meaning that each one of them locally never change its time keeping. that at the same time as we have the twin experiment in where we presume one twin, the traveling one, to find himself younger when coming home again, than its twin that stayed at home.

so if you use that one, you need a way to define what a universe is. From locality I think of it as equivalent points, if I do so it should somewhere present us with a discreteness. Maybe you could define it as a 'equivalent flow', but that does not give us a position in room and time. A point makes it possible though, The first (a flow) would be as trying to define a time and space position to part of a vacuum, being inside a infinite matter less vacuum, no 'anchors' to define it by. We define this universe (SpaceTime) through anchors, of different kinds. So 'points' are easier, and fits the idea of a equivalence in a simpler manner. But if time dilations and Lorentz contractions is a result of what exist between two locally equivalent points, then 'energy' also should be recognized as part of that 'in-between'. Or it's a dynamic puzzle, joining into a seamlessly existing 'common' SpaceTime. But it still questions what 'dimensions' would mean. Degrees of freedom is a very nice concept, when thinking this way.
==

Why I still prefer this one is because the other becomes impossibly complicated, as could be seen by our three observers defining different time dilations 'simultaneously' for the same far away object (the mirror they used). So the idea of islands of space and time, where 'time' goes (and so light, as I connect 'c' with your local clock) 'slower' or 'faster' actually becomes the worse model here, as it will be observer dependent.

also, this idea of locality is what we have built physics on, and it has worked very well, so far.
=

One more thing, although we can't speak of a simultaneous happening, except theoretically, we still can define it such as: while doing the two way mirror experiment, by our three observers able to, at all times, observe each other, as well as that far away mirror. A 'simultaneity' can be assumed to exist, in so far, as they all are causally linked to each other and the mirror, and its various outcomes. We do not need to set some exact simultaneity for this to be true, well as I see it. To accept that they all are linked in a same experiment, at a approximately same time (for example defined by a fourth observer), or by any of the three doing the experiment, assuming they could observe each other.

so even in situations where we can't reach a consensus on space and time we can use causality to prove that we did it 'together'. Now, if you take causality away I don't see you having anything left, to organize the idea of a universe, or SpaceTime,  around. I think it could be called the least common nominator for a universe.

[alancalverd]

If you define the box as having a uniform motion, close to light from some 'inertial', far away observers definition.

Hold it right there, pardner! The box is travelling at a uniform speed from A to B. How does the photon know whether it is supposed to act as if it is leaving A, or approaching B? Indeed if the box has no windows, how does the photon know it is moving at all?   

[yor_on]

It doesn't Alan
I think that is your view, and it is mine.

It has no possibility to 'know', although this is how I think of it. But there is that weak possibility, if we now assume some 'field', and also a 'flow' (no discreteness) that it can 'know', possibly? But if it does, it do not give any signs of it locally measured, our 'system' and us in a uniform motion sharing a same frame of reference. doesn't matter what 'speed' one give ones system, as long as we're discussing uniform motion.

[yor_on]

But it becomes a very strange universe defining it as me. Because we have the twin experiment and we have experimental proofs of different frames of reference, as compared relative ones own local clock and ruler, able to exist longer than they would sharing the same frame of reference as oneself. Instead of those islands, we now find a plasticity defining that 'common universe' we assume us to share, and if you use my definition, connecting 'c' to ones local clock, then that clock never lies to you, neither does it do it for me. Still we will find one 'identical twin' biologically younger after that journey. So now you have a puzzle, according to my definition their clocks are as always, 'c' and all other constants are 'as always' for both of them, locally measured, no matter who traveled.

[yor_on]

some want to place the puzzle at the 'turn around' of the twin, getting back home, expecting accelerations/decelerations to be the culprit creating a time dilation. I don't, to me we should be able to reach the same behavior through different 'relative motions', and as there is no 'absolute frame' defining what, and when, a relative motion is (locally defined), more than the limit we find locally defined through 'c', then the puzzle becomes even weirder
=

You can use three objects to define that there must exist different uniform motions, but there is no combination that I can imagine, proving that 'you' are one of those. 'moving'. You could be the one being still , and defining it locally you're free to do so in any uniform motion as far as I see.

[yor_on]

On the other tentacle. Physics builds on the concept of local measurements, we define them as equivalent and repeatable when anyone can do a similar experiment, the only thing differing the experiment being ones position in space and in time. Constants builds on this too. Want to throw away our physics? (uniform motions this is, as Earth, although someone might want to argue that Earth is gravitationally 'accelerating')
==

To argue that Earth is gravitationally accelerating would become interesting, if accepted. Would give a whole new twist to how to define what a repeatable experiment is.