what if ksp runs on the Frostbite engine.

[xnorb]

Absolutely no point in choosing Frosthype.

UE and Cry are far better choices because they aren't in constant Beta like FB.

You can also make stunning effects in Unity - just check some Interstellar Marines gameplay.

[allmappedout]

It's a simple "strength" formula that's causing a lot of the joint connections with KSP, IMO if they make the joints stronger and remove some of the "mass" of the wobble calculations it will get rid of a lot of the bugs. It's not a problem with the engine. The engine will do what you tell it to do. The problem is with the calculations.

Ferram's Kerbal Joint Reinforcement mod shows that unity is capable of building tall, strong, stable spacecraft, so it is just a coding issue. One that, personally, I hope will be addressed.

[Flowz0r]

What if KSP ran on its very own engine that does exactly what it needs to do and nothing more?

I realize that KSP probably wouldn't exist right now if Harv took the time to develop his own engine from scratch, but a man can dream.

[NASAFanboy]

Which makes them EA. They have proven again and again that they like to take full control of every company they buy out. Just look at what happened to companies like maxis and their latest simcity game.

Just he very thought of EA makes me shudder.

They're going to force tons of crappy micro transactions and DLCs down our throats.

[Stealth2668]

It would probably be a crashing mess like BF4 was for months.

[leax256]

Did I hear right? EA is not the developer of Frostbite Engine, it's DICE, here in Sweden. EA own DICE though...

when you work for a company, any thing you make on their time belongs on them. that's why dug companys make new drugs and not the scientists.

[Max Grant]

I thought that was squad themself not adding that feature yet.

The RSS mod has it, or so I thought.

[PDCWolf]

n-body would deteriorate game play. Why would they want to do that?

Care to explain how a non-playercentric model that would allow for native implementation of dynamic systems (Barycenter objects on a coherent manner for starters) and more gameplay options could deteriorate the gameplay? As long as you keep complex forces away, the game would stay the same except for a few fixes as to keep the planets the way the are (Otherwise vall would fly off and become a loner followed by bop as soon as the game starts).

You have mixed something badly here.

- floating point calculation - changing engine would not change this

- CPU only physics - this is Unity limitation, hopefully will be changed one day

- no axial tilt for planets - HarvesteR's choice not to overcompilcate things, gameplay over realism

- no complex forces - care to elaborate? Or is it tied to the next one?

- no n-body - HarvesteR's choice not to overcompilcate things, gameplay over realism, also tehnical dificulties no matter which engine you pick

- no optimization - how is tis tied to an engine?

- no good particle effects - don't know enough to comment this

- buggy joints - not sure if this is engine problem or if it can be fixed

-It could potentially solve it (Depending on the engine), given it's a bug in how unity stores the floating points and unity not being able to manage a space bigger than 2.5km² and requiring work-arounds like what scaledspace is.

-One day

-It was tested by modders and the engine/code can't handle axial tilt without dying. Let's say its a 50/50 between the actual engine and gameplay reasons.

-Complex forces refer to gravity bumps, gravitational gradient torque, tidal forces, etc. Unity would pretty much commit seppuku with the ammount of calculations required for that which are all floating point and CPU driven.

-n-body is not an overcomplication, unity just can't handle it thanks to what I mentioned in the first point.

-Optimization has a limit on unity and on every off-the-shelf engine.

-Unity doesn't have the potential to implement volumetric effects and/or 3d-particle effects.

-The workarounds discovered until now (like KJR) have their failures, no method is able to fully circumvent the bug except for welding which is just turning multiple parts into one eliminating any joints present.

[Boris_T_Roach]

Ferram's Kerbal Joint Reinforcement mod shows that unity is capable of building tall, strong, stable spacecraft, so it is just a coding issue. One that, personally, I hope will be addressed.

But I LIKE big wibbley wobbley spaceships that have a 50/50 chance of making 100 meters off the launchpad without something falling off/failing/exploding, and the sheer fear as your ship gets to 20km up and doing a gravity turn with 7 mainsails on full thrust and the G meter bouncing off the 4G marker where every one of your previous designs has an existence failure....

More seriously, if you want to see what can happen when you change game engines , take a look at duke nukem forever, it took forever to get a 1999 game out .. at the point where everyone else had moved on.

And the current engine is ideal, because it can be scaled down to run on less capable systems ( not everyone has a I7 2600 CPU with a 8 gig of RAM and GTX 560 1gig RAM vid card* )

Boris

* and I get lag when the part count gets beyond 500... but it looks too good to reduce the settings from 'max'

[TUFOM]

Cryengine would be the ultimate choise for ksp I think but I am definedly not expecting game engine change. Star Citizen on its time will show ultimate space game graphics but yeah its not kerbal space program. All things considered I am very happy there is even ksp and coming up cool space games. Time for space games has come again!

[qromodynmc]

Well unity is not the best choice,but Frostbite is not the best either. We need well optimized engine for openworld games,Opengalaxy actually

[Idobox]

The problem with N-body is not so much the calculation, they're not that difficult. The problem is that you absolutely have to solve numerically, and you have no stable orbits.

In terms of gameplay, that would mean you'd be stuck with physical time acceleration, like when in the atmosphere. You can forget the 10'000x speed to wait for a launch window.

It would also mean your satellites would not stay in orbit and crash. I don't know about you, but correcting the orbit of my satellites every few weeks, and replacing them when they run out of fuel is not my idea of a fun game.

Tidal forces can be very easily implemented without N-body physics and without changing engines, and are pretty much the same as gravity gradients. I have no idea what a gravity bump is, though.

I assume they are not implemented because they are not relevant for typical size and time scale involved in playing KSP, and because the devs didn't find that interesting. When they have better aerodynamics and more realistic reaction wheels, maybe it would make sense to look at tidal forces.

[malkuth]

All set with sleeping with a sleeze bag company like EA... They can keep there engine, and put it someplace I can't mention. Keep with what we have.

[TYRT]

Man, don't be crazy It means Squad must rewrite everything.

[PDCWolf]

The problem with N-body is not so much the calculation, they're not that difficult. The problem is that you absolutely have to solve numerically, and you have no stable orbits.

I was talking about calculations on Unity's side, unity is not ready for such heavy processing. Crafts' orbits could remain railed, after all your craft is not going to exert a force on anything.

In terms of gameplay, that would mean you'd be stuck with physical time acceleration, like when in the atmosphere. You can forget the 10'000x speed to wait for a launch window.

It would also mean your satellites would not stay in orbit and crash. I don't know about you, but correcting the orbit of my satellites every few weeks, and replacing them when they run out of fuel is not my idea of a fun game.

N-body can be implemented in a separate way from complex gravitational forces (orbital decay, gravity gradient torque, etc), also remember orbiter is pure n-body and it handles 1.000.000x timewarp while maintaining possibility of control (even if it's unusable). Also, gravity gradient torque is not bad at all, the most useful thing you can use it for is to make a long probe and have it self-orient vertically pointing to Kerbin or whatever you are orbiting.

Tidal forces can be very easily implemented without N-body physics and without changing engines, and are pretty much the same as gravity gradients. I have no idea what a gravity bump is, though.

I assume they are not implemented because they are not relevant for typical size and time scale involved in playing KSP, and because the devs didn't find that interesting. When they have better aerodynamics and more realistic reaction wheels, maybe it would make sense to look at tidal forces.

They are not implemented for gameplay reasons probably, you know, imagine in 2036 the devs coming and being like "Hey guys, we showed you tidal forces, complex forces and we got up to the point to playtest them and they weren't fun".

btw, gravity bumps are small (or not) differences in gravity. For example, the Moon basins exert more pull on an orbiting object than the rest of the surface.