On KSP2 and Special Relativity

[LitaAlto]

20 hours ago, Incarnation of Chaos said:

Because of the scale of KSP I actually think even 10% of C will be less useful than people think; what use is getting where you want in days or minutes if it turns a lander into a kinetic impactor because you can't slow down fast enough.

I personally think that will be part of the challenge, and would be most educational on its own. After all, this is a problem that all interstellar flight will need to reckon with.

[HebaruSan]

6 minutes ago, rmaine said:

And time raises further issues than "simple" dilation. Time being relative to the observer ends up implying that you can't even say whether two things happened at the same time or not. Also, the thing about light speed as a limit applies to information - not just physical objects. So if you wanted to do time "correctly" (or even some approximation to it), you would not be able to switch back and forth between focusing on your interstellar craft and the space center.

I wonder if they could solve some of that by declaring that all measurements are in the KSC's reference frame. So if you switch to a relativistic-speed vessel, it would use KSC's sense of simultaneity to pick the instant and adjust the local flow of time to be what KSC would see.

[LitaAlto]

17 hours ago, Incarnation of Chaos said:

The solutions tend to ship with the mods; if someone develops a crazy planet pack that has planets multiple parsecs out then I can almost guarantee that someone will make an SR implementation.

I think I'd rather see it baked into the stock game, personally. Forward-thinking is not a bad thing in game design, especially if the game is designed with modding in mind--and KSP2 is claiming to have enhanced modding support.

10 minutes ago, rmaine said:

And time raises further issues than "simple" dilation. Time being relative to the observer ends up implying that you can't even say whether two things happened at the same time or not. Also, the thing about light speed as a limit applies to information - not just physical objects. So if you wanted to do time "correctly" (or even some approximation to it), you would not be able to switch back and forth between focusing on your interstellar craft and the space center.

And to be clear, I'm not even talking about adding time dilation, only an upper speed limit and possibly nerfing acceleration as you approach the limit.

[Marachdrifter]

20 hours ago, PTNLemay said:

I feel it will kind of depend how they implement the interstellar aspect of the game. 

In the current game we are able to travel far enough and fast enough that eventually you leave the Kerbol sphere of influence, but I forget what happens after you do that.  I think that altitude and relative speed are all still measured according to Kerbol.  So if not that, what would the distances and speed be relative to?  A distant galactic center of some sort? 

If we assume the nearby solar system will be based on Alpha Centauri, and we continue using 10% the scale of the real world, then this new solar system would be around 0.44 lightyears away.  So... that's really not worth implementing any kind of speed limit.  Some flights in the Kerbol system can take several years anyway.  With the new super propulsion systems that trip could be done in a bit over 2 years while going at 0.2 C.

But... if they open the door for modders to make planet packs, and have available "slots" where new planets could go.  Or even slots for future official expansion packs (planet packs), and assuming some of these could be much more than 1 LY out.  Then... maybe we would want some form of SR. 

That is if they go for interstellar navigation. It's totally possible that you set your ship trajectory, fire your engine, and once you leave the kerbol System you ship disapear from the game. at that point, you unlock a new "stage" that is independant from the kerbol system. I say this because i wonder if the game can actually keep track of several systems at the same time. with that model, if you are playing in anothger star system, the others are paused, and each system would have their own time, or that the other systems are on pause and that the game will ajust itself for the time you spend in another system when you return to kerbol for exemple. that way it is not tracking several star systems in real time. 

[rmaine]

1 hour ago, HebaruSan said:

I wonder if they could solve some of that by declaring that all measurements are in the KSC's reference frame. So if you switch to a relativistic-speed vessel, it would use KSC's sense of simultaneity to pick the instant and adjust the local flow of time to be what KSC would see.

Of course then you are basically establishing KSC as an absolute reference frame, which is pretty much exactly the opposite of the whole idea of relativity. And once you try to go down that line, everything falls apart. That's what lead Einstein into time being relative - it was needed to make it all hold together. Hmm, speaking of things falling apart when you try to enforce an absolute reference frame, I just now thought to ask what people would mean by an upper speed limit in KSP. Relative to KSP as an absolute frame? That's not how most speeds in KSP are given and it could result in odd artifacts when near light speed; granted the real world has odd artifacts near light speed, but different ones.

[LitaAlto]

33 minutes ago, rmaine said:

Of course then you are basically establishing KSC as an absolute reference frame, which is pretty much exactly the opposite of the whole idea of relativity. And once you try to go down that line, everything falls apart. That's what lead Einstein into time being relative - it was needed to make it all hold together. Hmm, speaking of things falling apart when you try to enforce an absolute reference frame, I just now thought to ask what people would mean by an upper speed limit in KSP. Relative to KSP as an absolute frame? That's not how most speeds in KSP are given and it could result in odd artifacts when near light speed; granted the real world has odd artifacts near light speed, but different ones.

I'd argue that it'd only seem like an absolute reference frame from the perspective of the the KSC, which is where you'd be spending the bulk of your time. I wonder if it'd make sense to switch reference frames between the active ship and the KSC, but no other objects in the game.

[HebaruSan]

6 hours ago, rmaine said:

Of course then you are basically establishing KSC as an absolute reference frame, which is pretty much exactly the opposite of the whole idea of relativity. And once you try to go down that line, everything falls apart.

Does it, though? I guess I'd need to see some examples worked to buy that. Maybe it would work out fine.

[Lach_01298]

I think the "best" you could hope for is there is an absolute speed limit and that it takes increasingly more and more energy to accelerate to the speed of light giving the effect that you can't go faster than the speed of light. But there will be an absolute reference frame from which this speed is measured. I know that is completely against relativity but i don't see how they could do it. Like how would you do time dilation or lengthy contraction or difference in simultaneity? I've never seen a game implement relativity and i don't know how it would be possible.

But as stated many times in this topic in 99.9% of cases relativity won't be relevant. The general rule of thumb is you don't see any significant deviation from Newtonian mechanics until about 0.3c. So really in the end you can only hope for there being a speed limit that you can't exceed just for hardcore players or Dannys out there.

[purpleivan]

From a developer's point of view, the simplest way of doing intersteller travel would simply be to cheat... i.e some kind of wornhole/stargate/ jump point (whatever sci-fi IP rocks your boat).

Just get to the location of "point of magic stuff happens", maybe with some requirements for travel thrown in, and "poof... you're in a new start system". Not saying that this is the way that KSP2 will handle it, but it must be a tempting method for the developers to use.

[chaos_forge]

1 hour ago, Lach_01298 said:

I've never seen a game implement relativity

Check out A Slower Speed of Light

[Lach_01298]

31 minutes ago, chaos_forge said:

Check out A Slower Speed of Light

Cool! But this doesn't support multiplayer which creates the hardest part about relativity: difference in simultaneity. This video explains it well 

 

[ModZero]

Just properly implementing light delay (so not like in RT2, lol) means your game state explodes in size because you have to keep a way of tracking what has seen what. Add relativity to the mix and your game is now best played on a blackboard.

[Lisias]

13 hours ago, ModZero said:

Just properly implementing light delay (so not like in RT2, lol) means your game state explodes in size because you have to keep a way of tracking what has seen what. Add relativity to the mix and your game is now best played on a blackboard.

So don't implement it properly.

Make things happening faster "outside" the the Current Frame of Reference for things under Relativistic rules, and make things happening slower for them when the current Frame of Reference is under Newtonian rule.

You don't have to implement the thing - it's good enough to mimic their effects on artifacts under the user's perception.

The only extra complication I see is having to track time separately for artifacts under Relativistic rules.

Edited August 25, 2019 by Lisias
make a bit of a confusion on that Frame of Reference thing

[nikokespprfan]

When it comes to relativistic effects, there are two area's where I'd just like to see how the universe behaves: 1) on interstellar travel, and 2) around bodies such as black holes. I just want to see what happens.

However, by carefully considering and tuning the content of KSP2, it will not be necessary to have it. We don't have to have black holes, we don;t have to have drives that get a significant enough portion of the speed of light.

Yet, I'd feel this as something missing.

-------

Two notes on our discussion about relativistic mechanics;

The principle of relativity is a thing that exists in the game. Your velocity can currently be measured in many reference frames (even rotating ones), broadly grouped into the categories "with respect to the rotating ground of a body", "with respect to a body moving through wider space" and "with respect to a target object moving through wider space". In interstellar space "body" will mean the centre of the galaxy (if the galaxy is set up as in real life) and "target" will be relative to the target star (which is what we all mean when we talk about reaching sppeds of X% of c). No need of a universal frame here.
 

As for simultinaity, I personally would choose for a KSC reference frame, since most of kerbalkind (under certain assumptions about the wider kerbalkind) will be experiencing the state of affairs in this way. You might say, but relativity but relativity simply says that all reference frames (inertail ones that is) can be seen as the true reference frame, it just becomes problematic if you start messing up what speeds belong to what reference frames. Gamewise, I don't see any of this becoming an issue except maybe for stuff like a universal game time needed for saves and such. Physics simulations can be sped up/slown down in terms of that UGT, but stay the same under the hood, computationally speaking. It doesn't matter for focussing on a vessel that moves with the camera. And well, vessels that don't focus on the camera, they don't need complicated physics, unless you crash things at relativistic speeds.

[rmaine]

Crashing things at relativistic speeds, eh? Sounds fun. :-)

 

[nikokespprfan]

I like your thinking, but good luck keeping your aim to the other vessel while moving at those speeds.

 

scratch that, it's not about the crashing, it's about the moving relative to that craft at relativistic speeds, while you are loading the craft.

 

No, scratch that as well, it is just about crashing, since you can just statically move the other vessel by without physics for the split second it is in the loaded zone around your current craft. No-one will notice a single thing. If you slow down just right or crash, you can start simulating normal physics as soon as the relative speed difference gets below the SR-speed treshold that the game uses to know when to start simulating SR. It's an edge case sure, but one with a very simple solution.

Edited August 25, 2019 by nikokespprfan

[nejc]

There is one effect of fast travel that becomes noticeable below 10% c as demonstrated in A Slower Speed of Light. Stars in front blue-shift and concentrate together, stars behind red-shift and expand. It is subtle below 10%, but since the effect is purely visual with no effect on game-play, it would be easier to implement. 

The effect should be noticeable using nuclear pulse and inertial fusion drive. However, assuming antimatter rockets, time dilation should come into play. But, it would be too unwieldy for game-play, so I suggest only implementing certain consequences: lower life support expenditure at higher speed and higher fuel expenditure (faster you go, more fuel required to accelerate).

Edited August 29, 2019 by nejc

[nikokespprfan]

6 hours ago, nejc said:

 lower life support expenditure at higher speed and higher fuel expenditure (faster you go, more fuel required to accelerate).

Genius

[Daniel Prates]

That whole interstellar travel in KSP is promising to be one colossal goatfudge in the desert (ha! Two movie references in one). 

One thing that has been bothering me is that stars are not just laying in space. I mean, Kerbol is, but that is because up until now it was alone in the whole universe. Now that we are getting more stars, the very first  thing that crossed my mind is: are they orbiting around some galactic baricenter? I don't mean that this baricenter has to be a proper object in the game world, but the stars are obviously moving in different speeds and will that be modeled? If Kerbol is the lower-altitude one,  it is slowly overtaking the other stars. Will that be taken into account? This may be more relevant than it seems. You see, if we have three stars that are mere static, fixed points in the universe, travelling from one to the other is a straight line - specially so if each star has a SoI, which you can leave. But if modeled more realistically,  travelling from Kerbol to AlphaKentauri (let's call it that) is the very same mechanic as leaving Kerbin towards Duna.  A rendezvous with another moving object, with a interstellar travel path which is actually an elliptic orbit within a gravitational field (the baricenter's field), with all implications (you slowing down as you approach a higher altitude target object, for instance - a  fact that, alone, completely changes the idea that you have to accelerate than decelerate when you arrive at  your destination star).

How is it going to work in the game? This has been pestering me ever since I read they were going to include more stars. 

 

Edited August 30, 2019 by Daniel Prates

[nejc]

10 minutes ago, Daniel Prates said:

How is it going to work in the game? This has been pestering me ever since I read they were going to include more stars.

I think the relevant topic is here:

Regarding this topic, the stars would probably have to be static for relativistic effects to be manageable for game-play, assuming they were to be implemented at all.

 

[HebaruSan]

1 hour ago, Daniel Prates said:

But if modeled more realistically,  travelling from Kerbol to AlphaKentauri (let's call it that) is the very same mechanic as leaving Kerbin towards Duna.

Not quite as "same" as you might initially think. A Hohmann transfer takes on the order of half an orbital period to execute (more for one body, less for the other). The Sun's orbital period around the galaxy is 240,000,000 years. So if you wanted to perform a Hohmann transfer to another star just like going to Duna, you're talking about a 120,000,000-year trip. Brachistochrone trajectories will be necessary to bring that down to a manageable duration, at which point it's much more like the stars are standing still.

[nikokespprfan]

Astronomers keep tables and databases of positions of stars in the milky way galaxy. These databases get updates every 50 years to account for all kinds of phenomena, from changes in the earth's rotation to the stars moving around. I don't know the average change that these tables have, what I can say is that that is the change in position you notice in between a trip of 4 lightyears on a craft at 10% speed of light. That can give you an idea as to how much the stars really move.

 

What I'm trying to say is, in the KSP universe we are used to bodies further out moving larger distances over time, but my gut says that this just isn't the case for stars, everything is very static on human lifetimes, although admittedly you might not completely be bounded by human lifetimes in KSP 2.

Edited August 30, 2019 by nikokespprfan

[Daniel Prates]

@nejc yes, I was trying to tie-in both topics, not to debate a different topic. I too think that for game purposes they most likelly keep all stars static, so that is something to consider in this discusison as the perspective may differ.  No sense in arguing to death an interesting topic only to find out later that the devs took a different approach. This is relevant. How things play out, including the application of the laws of physics (relativity too) relates to whether interestellar travel is a staight line in a void of forces, or if it is done in a gravitational field centered on a baricenter.

@HebaruSan you are right, of course. I merely meant the physics are the same. But the scale of things change what is feasible or not. A hohmann transfer to another star takes too long in our real-life. On the other hand, a kerbal section of the "k"alaxy would probably be a lot smaller in scale. Guess we will have to wait and see. If the vincinity of Kerbol is closely knit together, maybe our orion drive analogues could allow a trip like this: decrease altitude closer to Kerbol, burn main drive there, hohmann from there here to Kentauri. I feel I am veering off topic, but my point was, all depends on knowing if the void between stars is a forceless void or if it crosses the SoI of a baricenter somewhere far away. That's the part I wanted to add to this discussion, nothing more.