n-body physics? We don't even have 2-body physics!

[cantab]

I think the only thing you lose by representing a craft by its point mass is tidally-induced rotation and breakup. Rotation doesn't really matter with our godly reaction wheels, and I suspect a ship large enough to break up tidally would be too large for other aspects of KSP's physics.

[Monger]

Assuming your ship is built of 4000 jumbo tanks, it has mass 144,000 tons.

Gilly has mass 124,205,120,000,000 tons.

Well... I think it doesn't really matter Gilly is on rails.

Exactly. Every simulation needs some kind of abstraction from reality.

This is why KSP doesn't need 2-Body physics, and it definitely doesn't need N-Body physics. If course the simulation is not correct (no simulation ever is), but it is correct enough to achieve your goal (i.e. have fun while playing).

[Camacha]

This is why KSP doesn't need 2-Body physics, and it definitely doesn't need N-Body physics. If course the simulation is not correct (no simulation ever is), but it is correct enough to achieve your goal (i.e. have fun while playing).

I am not too sure about that. KSP is best at introducing people to typical problems when dealing with spaceflight. Tidal interactions, drift and similar phenomena are actually a big part of that.

If I would follow your line of reasoning, I have no idea what Keplerian orbits are doing in the game. You do not need them to have fun and are only confusing. Yet they are one of KSP's strongpoints. There is little reason that would not extend to n-body physics, unless someone can explain very clearly why the gameplay challenges caused by that could not also be part of the fun. The line is completely arbitrary.

[CaptRobau]

Wouldn't the biggest impact be that in time all orbits would change? Other bodies would pull at your spacecraft and change their orbits. KSP is a game in which most players only focus on one mission at a time. You couldn't take a probe on a year long mission to Eeloo, for fear of your Kerbine or Munar space stations losing too much altitude.

[Camacha]

The biggest change would be that KSP would become an even better laboratory for all kinds of eduction and shenanigans On-rails objects just give a user a little less freedom.

ou couldn't take a probe on a year long mission to Eeloo, for fear of your Kerbine or Munar space stations losing too much altitude.

Although I do not think that is true, people also tend to forget the tools we have for dealing with current problems - we have a beautiful map view and navigation nodes to help us around. You really think n-body physics would be implemented without some aid to help the user along a bit - while still getting to solve the problems posed by it?

You were the ones that talked about a game having to be fun, right?

[Nathair]

Anyone could mod a Jumbo Fuel tank to have the same mass as Gilly and then HyperEdit it into Gilly orbit.

I'm sorry, I don't see the point here. Until the simulation is absolutely exact, which is not possible, there will always be the potential to hack, mod and/or cheat yourself into a position that exposes or takes advantage of the discrepancies between the model and reality. That is simply not a solvable problem (if it even is a problem at all.) Once we accept that the simulation will never be absolutely perfect it becomes a simple matter of deciding how close it needs to be. "Needs" is the key word there. How close it needs to be depends upon what you want from your simulation. For KSP the current gravitational model seems to meet the needs of the devs, myself and, judging by the comments here, the vast majority of the playing community. By this standard it just ain't broke, so why should time be invested in trying to "fix" it?

I think that, perhaps, the problem here lies in the expectation that KSP should be written to meet your personal wishes. If you want better physics, Orbiter is still out there.

[Camacha]

I think that, perhaps, the problem here lies in the expectation that KSP should be written to meet your personal wishes. If you want better physics, Orbiter is still out there.

This argument is easily reverted - if you do not wish to use n-body physics, why not play another game? There has been a consistent discussion regarding n-body physics in the community, so I would not easily dismiss the need or desire for it. For the rest, see my earlier comments on drawing an arbitrary line. I would like to see substantiated why Keplerian orbits are good and n-body physics are a bridge too far.

[Whirligig Girl]

Note that I'm not saying that for it's purposes the game's gravity doesn't work, it does. But I'm just clearing up a misconception when people talk about n-body physics, they say that KSP has 2-body physics, when the only real physics simulation is 1-body physics.

[m4v]

Honestly, why do we need n-body? Lagrange points? come on, I would rather see effort put somewhere else than go into some feature that does mostly the same thing we have now.

[Nathair]

This argument is easily reverted - if you do not wish to use n-body physics, why not play another game?

Er, no. My point is that n-body physics just wouldn't make enough of a difference to the game to greatly influence me one way or another. I don't see the need to enormously increase the complexity and computational load, however if KSP had been written that way I would still be playing and enjoying it. If 0.29 comes out with perfect Newtonian simulation I will still be playing and enjoying it. (Of course then I'll be in the forum reading the complaints that it's just crappy n-body Newtonian mechanics and what about relativistic effects?! Why is that a bridge too far?!)

For the rest, see my earlier comments on drawing an arbitrary line. I would like to see substantiated why Keplerian orbits are good and n-body physics are a bridge too far.

If the enormous computational increase in going from a 1 body to an n-body simulation isn't immediately obvious to you then I don't know what I can say that would make a difference. As to it being "a bridge too far", I would suggest that it is, instead, a decision about resource allocation. This is the oberth design effect; where are you going to apply your resources to have the greatest overall impact. In my opinion converting to a new physics model would be far too much effort for largely trivial results. Obviously YMMV. (And just as obviously this particular horse is dead, dead, dead.)

[Accelerando]

I think that, perhaps, the problem here lies in the expectation that KSP should be written to meet your personal wishes. If you want better physics, Orbiter is still out there.

It's terribly ironic how you are so content with shunting someone else's wishes for the game on the basis that they would like the game to meet those wishes, while you clearly have a such a wish of your own in that the simulation shouldn't change from what you deem "good enough".

[Pecan]

I'll keep this short ... Gilly is not attracted to ...

I know it's upsetting and frustrating but I'm sure in time you'll find n-bodies that are attracted to you[r ship]. Don't worry, just try to move on Platonically.

[Er, sorry, I may have missed something there ...]

[Nathair]

It's terribly ironic how you are so content with shunting someone else's wishes for the game on the basis that they would like the game to meet those wishes, while you clearly have a such a wish of your own in that the simulation shouldn't change from what you deem "good enough".

As I have already pointed out, that's a false dichotomy. I am not opposed to n-body physics. (I am, on the other hand, opposed to casting every difference of opinion as being between perfectly valid and balanced opposites.)

[*Aqua*]

It's terribly ironic how you are so content with shunting someone else's wishes for the game on the basis that they would like the game to meet those wishes, while you clearly have a such a wish of your own in that the simulation shouldn't change from what you deem "good enough".

Let's look at the facts:

- N-body-physics make orbits instable.

- Some kind of autopilot is required to keep a stable orbit.

I guess the autopilot needs electricity and/or fuel to work. If it hasn't that it will stop working. At this point the orbit will gradually decay shooting the craft into a planet or out of the system. Due to lack of fuel/electricity the craft can't do anything anymore. Usually the player will mark it as debris and forget about it.

Summary:

- With working autopilot you'll have orbits like those of the current system.

- Without working autopilot you'll have chaotic orbits but you don't care. You'll discard the space ship anyway.

So what's the point of having n-body-physics?

Edited March 12, 2014 by *Aqua*

[vetrox]

N body physics just sounds like a big waste of time for ksp.

The whole point in ksp is that its almost a pick up and play. All you really need to get started is the tutorials that teach you how to get off the ground and what pe/ap are.

add n-body physics and you have to start doing crazy calculations and tedious planning. Its all well and good making the game super realistic but this turns it into less of a game and more a pure simulator such as orbiter.

KS, from a physics perspective, does its job well, it teaches you the basic principles of spaceflight and still allows you to enjoy yourself. Personally n body just sounds like an idea that sounds good but is probably boring, and tedious and boring as **** for an actual game.

its like trying not to fast travel in skyrim. It sounds like an awesome super cool roleplay/realism idea but after the first few trips across skyrim on foot, u soon get bored and activate fast travel again.

It would deffo need to be a toggleable option. Then all the elitists could shoot everyone who doesnt enable it down.

[Guest]

You're completely, utterly wrong. What autopilot is needed for is precise, long-term stationkeeping. Barring edge cases, orbits in N-body simulation are not chaotic, but they do slightly change over time. Unless you screw up, the orbit won't go out of control. Also, autopilot would be required for maintaining unstable, edge-case orbits (which you might want for whatever reason), but nobody says you have to use those.

Clearly, you people don't know just how big deal N-body (or more accurately, a physics model based on integrators) is. It's not "only Lagrange points". There are sun-synchronous orbits, orbital decay, realistic planet motion, thrusting during timewarp, tidal forces, binary systems... With this kind of simulation, forces can be applied during timewarp just fine. It also fixes inaccuracies when crossing the edge of a SOI (currently one of Kraken's nests) and gives you a lot more options than with the current simulation.

I don't think it should be toggleable, either. It's just too big of a deal. Disable it, and say goodbye to ion engines, tidal stabilization, sun synchronous orbits, solar sails... The possibilities are so enormous that you wouldn't be able to stop using this once it's added.

Edited March 12, 2014 by Guest

[Camacha]

So what's the point of having n-body-physics?

Please read the previous posts made on the subject.

I am not too sure about that. KSP is best at introducing people to typical problems when dealing with spaceflight. Tidal interactions, drift and similar phenomena are actually a big part of that.

If I would follow your line of reasoning, I have no idea what Keplerian orbits are doing in the game. You do not need them to have fun and are only confusing. Yet they are one of KSP's strongpoints. There is little reason that would not extend to n-body physics

My next post after that also applies. To be honest, it all sounds a bit like we don't know it so it scares us.

If it were up to me they would drop the whole multiplayer thing and improved physics across the board.

Edited March 12, 2014 by Camacha

[Lack]

It's worth pointing out that HoneyFox has a n-body capable plugin released: http://forum.kerbalspaceprogram.com/threads/70881-0-23-Orbit-Manipulator-Series-%28WIP%29-%28Updated-March-12-2014%29

It's not a super-accurate implementation, and it's only for the active vessel (although it does some cool things like atmospheric decay for non-active parts and warp while throttled for ion engines). It is interesting however, I reckon the planets (or at least moons) would have to be shifted around considerabley. You see how quickly a 100km by 100km Kerbin orbit decays.

EDIT:

From a perfect 100x100km orbit

I should point out that'll continue on for quite a while. It just quickly becomes lop-sided and hits the atmosphere.

Edited March 12, 2014 by Lack

[Red Iron Crown]

thrusting during timewarp

You have my attention with this, all the other stuff doesn't matter to me. How does n-body help with thrusting during warp?

[Camacha]

You have my attention with this, all the other stuff doesn't matter to me. How does n-body help with thrusting during warp?

N-body is always applied, there is no on-rails That is pretty much the point of the system.

[Red Iron Crown]

N-body is always applied, there is no on-rails That is pretty much the point of the system.

So, it would be equivalent to the physics warp we have now? No simplification possible, thus computationally expensive for high time compression, if high time compression is even possible?

[Camacha]

So, it would be equivalent to the physics warp we have now? No simplification possible, thus computationally expensive for high time compression, if high time compression is even possible?

You could roughly compare it to that. Computational expensive should not be an issue, as games like Universe Sandbox do this on a huge scale with great succes. You should also not suffer from all the weird effects physics warp causes.

[NecroBones]

If I'm understanding correctly, it would allow for applying physics in terms of the forces acting on the ship (gravity from multiple bodies, solar wind, engine thrust, etc) at all times, relatively accurately, no matter the time scale / warp factor, in a computationally inexpensive manner? That alone would make it worth it, I think. Otherwise, I think the game is doing what it set out to, in terms of teaching basic orbital mechanics and building fun gameplay around it.

[Guest]

You have my attention with this, all the other stuff doesn't matter to me. How does n-body help with thrusting during warp?

When you're doing N-body physics, you need to preform numerical integration in order to derive object motions from forces applied to them. You need to do this both during timewarp and out of it, as in that case, integrators handle everything, including planetary motion. With that in mind, thrust becomes just another force acting on a ship that has to be taken into account. So, not only you could thurst during timewarp, but also change the throttle and toggle engines. Solar wind, ion engines and upper atmosphere drag would all become able to act during timewarp.

The same also applies to drag and torque, BTW. So not only it fixes "timewarp stabilization" and allows for orbital decay to be implemented with minimum fuss, but allows simulating the need for attitude control while at warp, and various different ways of achieving it. While you do have to work around KSP unloading unfocused vessels, it's still very possible to do.

[Nathair]

Clearly, you people don't know just how big deal N-body (or more accurately, a physics model based on integrators) is.

Or, just maybe, everyone who disagrees with you isn't completely ignorant, maybe we just disagree with you on where the "big deal" line should be drawn in this case. De gustibus non est disputandum, you know?