Simulator

[DJRWolf]

A way to test designs at far flung places without having to take the time to get there. Great for testing a design out in career mode without having to spend the money (once it is in the game). I'll give a couple examples.

You want to see if a design for a lander is safe and stable for for both landing on and getting off of Duna and back to an orbiting mothership.

Seeing if a space plane design will work for getting down from orbit of Jool (or other gas giants as they are added) and back into orbit safely.

Doing all of that without having to spending any money but getting nothing in return expect knowing that once you get there in the real mission the craft will not have a design quark that causes it to fall apart and be a total failure.

[TimMartland]

+1, would be great.

May have already been suggested about 10,000 times though.

[ZingidyZongxxx]

Simulator? The world is your simulator! There are no consequences for failing when you can just revert to the VAH or SPH, so why be afraid to fail? The worst you can do is learn from your mistakes.

[astropapi1]

In the meantime, you can "export" your design from career to sandbox and use hyperedit.

Or, if you feel like it, download the Kerbal Engineer Redux and do some number crunching. You should be able to know if your craft is able to do something or not even before leaving the VAB.

[Captain Sierra]

Simulator? The world is your simulator! There are no consequences for failing when you can just revert to the VAH or SPH, so why be afraid to fail? The worst you can do is learn from your mistakes.

So you would rather time warp to a transfer window, make the transfer through what could be a several minute NERVA burn, do a dangerous aerocapture where physics could give out and the kraken rip you a new one, make another transfer and aerocapture, all to test a Lathian SSTO? And don't even get me started on the crazy things Hyperedit has done to my game. It's done some messed up s%$&.

I, for one, agree, particularly for rovers and terrain.

[DJRWolf]

Simulator? The world is your simulator! There are no consequences for failing when you can just revert to the VAH or SPH, so why be afraid to fail? The worst you can do is learn from your mistakes.

And if you happen to have already gone back to the Space Center and can't revert to launch or VAB? Plus there are consequences in career mode (or will be once reputation and money is added).

[SofusRud]

And if you happen to have already gone back to the Space Center and can't revert to launch or VAB? Plus there are consequences in career mode (or will be once reputation and money is added).

But adding a "simulation mode" would essentially allow you to bypass those consequences, rendering them meaningless. No consequences= no tension= not different from sandbox. I would think that career mode would be a rather pointless experience with a simulation mode. It might as well be sandbox

[Duxwing]

But adding a "simulation mode" would essentially allow you to bypass those consequences, rendering them meaningless. No consequences= no tension= not different from sandbox. I would think that career mode would be a rather pointless experience with a simulation mode. It might as well be sandbox

Simulation mode would reduce the time between tests without rendering the consequences of failure meaningless because with simulation mode added, players would eventually fly in the 'real' world, where even after umpteen simulations catastrophic and deadly failures can occur.

-Duxwing

[nholzric]

I can't believe that budgets will be so tight in career mode that when something goes horrifically wrong it's game over. Things go horrifically wrong all the time in KSP, it's practically a feature! I obviously don't know how it will be balanced, but we shouldn't be worried that a lost ship (or two, or three, or...) is going to end a career.

I agree it sucks putting in all the time and effort in to a deep space mission just for something to explode or get stranded, in fact that would be my number one "complaint" with KSP. That's what quick save is for in addition to dV calculations, thrust to weight estimations, and the wiki.

[Stargate525]

I agree it sucks putting in all the time and effort in to a deep space mission just for something to explode or get stranded, in fact that would be my number one "complaint" with KSP. That's what quick save is for in addition to dV calculations, thrust to weight estimations, and the wiki.

A game should not require you to check an internet guide in order to perform basic functions of the gameplay. As it stands, in the game, the only method of testing anything is to launch, test, revise, repeat. That gets incredibly, stupidly tedious when you want to fine-tune TtW ratios on Laythe or Eeloo, or run power consumption setups for probes on dark sides of planets.

I also don't think we're asking for a 'teleport to planet' option. Heck, I'd be perfectly fine with a 'simulate launch' button that dumps you onto a giant cue-ball that matches your chosen planet's gravity, atmosphere, and size.

[DJRWolf]

As it stands, in the game, the only method of testing anything is to launch, test, revise, repeat. That gets incredibly, stupidly tedious when you want to fine-tune TtW ratios on Laythe or Eeloo, or run power consumption setups for probes on dark sides of planets.

Just what I was thinking. Making sure those little things don't come up and bite you in a mission that you can't revert back to launch for. Or even in sandbox mode you just want to test one thing in a final playload without having to spend all of the time and effort to build a booster and go though all of the time to troubleshoot just getting it to the area you want to test it at.

[nholzric]

A game should not require you to check an internet guide in order to perform basic functions of the gameplay. As it stands, in the game, the only method of testing anything is to launch, test, revise, repeat. That gets incredibly, stupidly tedious when you want to fine-tune TtW ratios on Laythe or Eeloo, or run power consumption setups for probes on dark sides of planets.

There are many games have extensive internet resources that explain everything from basic to arcane aspects of game play. If using the wiki is not hardcore enough for you, I wonder if landing a probe with a Gravioli Detector and a Barometer would allow you to calculate the relevant stats. The planet tool tips in map-view lists surface gravity.

We can agree to disagree, but in a game with a Sandbox mode, a "revert to launch" option, engine specs listed in the construction screens, and an extensive online community, it's difficult to believe that a simulation mode would add much.

If I were going to add a simulation mode though, I'd make it as distinctively "simulationy" as possible. Maybe a white background with one or two neon green cubes as a planet. The cubes would scale in size as you adjust their simulated gravity. Then you could choose your location and velocity... Maybe you could even make it 2D? (Use KSP's normal physics and 3D environment but limit everything to keep it's Z coordinate constant (or whichever X, Y, Z)) My design focus would be to keep it as unique an experience from the rest of the game as possible to avoid devaluing either the sim or the "real world".

[Stargate525]

There are many games have extensive internet resources that explain everything from basic to arcane aspects of game play. If using the wiki is not hardcore enough for you, I wonder if landing a probe with a Gravioli Detector and a Barometer would allow you to calculate the relevant stats. The planet tool tips in map-view lists surface gravity.

To be clear, I wasn't saying that it isn't COMMON (Minecraft suffers from a terrible case of wiki-itis ). It's just bad design. And it's got nothing to do with being 'hardcore,' I just disagree with having to guesstimate thrust-to-weight ratios for different planets, guesstimate the number of drogues I need for my descent stage, guesstimate the number of batteries I need to bring to maintain power through the dark-side of an orbit.

It's basically that there's no middle ground. I can flail around repeatedly hurling vessels out until I get it perfect (and have done 60% of the mission a dozen times), or I can close the game, go online, break out the calculator, and sit and do rocket science. What a simulator mode lets us do is give us a shorter iteration time for payload testing. Flail around but actually get something DONE before we get both bored and frustrated.

We can agree to disagree, but in a game with a Sandbox mode, a "revert to launch" option, engine specs listed in the construction screens, and an extensive online community, it's difficult to believe that a simulation mode would add much.

-Sandbox doesn't give you anything remotely like what we're asking, and you can't port designs across saves without diving into the file hierarchy of the game.

-Revert to launch doesn't solve the problem of those three hours you just wasted getting to your testing grounds, or give any functionality for multi-part missions.

-Engine specs, yes. But the total mass isn't there. Nor is any functionality to provide dV or stage burn lengths without googling the formulas and breaking out the calculator. There's no way to test rover stability in lower gravity in the VAB, no way to test parachute effectiveness...

-Again, a game which requires you to search resources outside the game to solve basic problems has a tutorial and ease-of-use failure, in my opinion. And how many times have we asked for something like this, or to get answers to this question, only to be told 'go get Engineer Redux / Hyperedit?' A fair amount of the time, the community tells us to go add a simulation-analog to the game. Why is it such a crime if Squad just DOES it?

If I were going to add a simulation mode though, I'd make it as distinctively "simulationy" as possible. Maybe a white background with one or two neon green cubes as a planet. The cubes would scale in size as you adjust their simulated gravity. Then you could choose your location and velocity... Maybe you could even make it 2D? (Use KSP's normal physics and 3D environment but limit everything to keep it's Z coordinate constant (or whichever X, Y, Z)) My design focus would be to keep it as unique an experience from the rest of the game as possible to avoid devaluing either the sim or the "real world".

Did you just stop reading after the first paragraph? What part of 'dumps you on a giant cue-ball' makes you think I want anything different? I wouldn't say cubes, though, as close orbits will become impossible, and 2D doesn't make any sense for simulating a 3d world.

Like I said previously, I think all that's really needed is a single body that can be set to emulate any body in the game, painted white (maybe with a grid on it), and an option to either launch from the ground or start in orbit.

[DJRWolf]

Revert To Launch/VAB is all well and good but it does not stick around forever.

Example: Having to do two or more separate launches to get an interplanetary mission going with 10 astronauts and having to save and quit as they are halfway to the destination because you have to go do something in real life. Then when you do get back on and continue the mission you wind up having it fall apart just as you get there because of one small design flaw and since there is no more Revert to Launch/VAB anymore you have now killed 10 of your astronauts and your reputation has taken a huge hit (reputation is coming).

Being able to test out stuff like this will help you refine your designs in such a way that has the following benefits:

- Does not cost any currency

- Does not effect your reputation

- Takes less real world time since you don't have to start back at the launch pad for every small thing to get to a place like Eeloo that even with time warp still takes real world time

As for the stuff about math and specs I have noticed that the mod MechJeb helps out a ton, I was starting to get to the point where I would give up on this game if it were not for that mod. But it is after all a mod and not part of the core game as it should be. NASA has missions fly under computer control for a reason. Normal humans can't do some of the complex stuff that you need to do for a realistic space simulator. Some of the things the early astronauts were able to do was because only the best were choosen but the average gamer is not able to do stuff like calculus in their head on the fly like you would need to without computer controlled flight.

[Elington]

A way to test designs at far flung places without having to take the time to get there. Great for testing a design out in career mode without having to spend the money (once it is in the game).

Hi DJRWolf,

I don't want to interfere and IMHO all playing styles are good provided they give fun to the player!

However I'm jumping in because your concern is exactly what drove me to work on this:

http://ascentkomputron.blogspot.fr/

Forum:

NET-App-Ascent-Komputron-rocket-design-tool-and-ascent-trajectory-optimization

I can't say it will solve all the test cases you mentioned but at least it got me to Eve's surface and back right on the first design trial. I'm currently working on a .craft import feature to make the design description much easier (should be ready within a week).

Please don't hesitate if you have any feedback and suggestions,

Happy flying,

David

[coolitic]

Probably going to happen. Would also help to see if my rocket would be destroyed on launch without having to constantly switch between vehicles assembly and launch with the long load times.

[DaRocketCat]

A simulator for a simulator...ok then.

[Elington]

A simulator for a simulator...ok then.

Well for real-life IC design (microelectronics chips) you need a System simulator, a Schematic-level simulator, and a "back-annotated schematic" i.e. Layout-level simulator, all simulating the same electronic circuit at different levels of abstraction... so don't you dare complain at level 2!

Edit: working on the Ascent Komputron, I do support the idea of course, but I understand some players don't want this

Edited January 22, 2014 by Elington
Clarification

[xrayfishx]

I don't see why some of you people are against it. Especially back in 0.18 where I started, I'd have killed for such a feature, since I had no idea how much thrust I would need to get off the surface of Mün. Even now I'd like to be able to know that my Duna lander works, instead of relying on my gut feeling.

Also, as someone mentioned, this would make the barometer and gravioli detector (and if we get integrated re-entry heat, then the thermometer) useful. In the beginning you'd know very little of what the planets are. Say, you want to land on Duna. From the measurements made on Kerbin you know that it's gravity is somewhere between 2.5 and 3.5m/s² and its atmosphere up to an altitude of 3000-5000m. You can either simulate a lander on a perfect sphere with these parameters or send a probe there to get exact measurements and some photos to figure out what its surface might be like so you can test your lander in the correct conditions. It would make the measurements become useful for future missions instead of unlocking new stuff (because, let's face it, knowing the surface gravity of a giant rock in space won't get you closer to figuring out how to make a bigger fuel tank).

[leptoon]

You are your own simulator. If you're really that concerned about whether or not your craft will do what you intended, crunch some numbers. You can use math to simulate anything that you might need to determine. There are several helpful mods for that.

Edited January 23, 2014 by leptoon

[xrayfishx]

You are your own simulator. If you're really that concerned about whether or not your craft will do what you intended, crunch some numbers. You can use math to simulate anything that you might need to determine. There are several helpful mods for that.

I can calculate the Dv I need to get to Jool. I can calculate the TWR of my Laythe SSTO. Hell, if I have to, I can calculate my entire damn flight path to Moho, Eve gravity assist included. But I can't calculate how many parachutes I need to land a probe on Duna. I can't calculate how stable my rover is on Eeloo and I sure as hell can't calculate exactly how do I reenter with my capsule that's returning from Dres without squishing my kerbals or bouncing out of the atmosphere which I cannot afford because I will run out of food and die if I do (reentry heat is coming and I imagine some sort of life support might come, too. Even if it doesn't I can't calculate anything atmosphere-related).

There are so many things that I can't calculate. And there's a reason NASA does tons of simulations before drawing up an actual plan for doing anything. Some of us don't see Kerbals as a cheap resource. Look at my signature for a second. I'd be so happy if I could have simulated the stability of my asymmetrical spacecraft when it's low on fuel. Jeb would be alive. Or if I could have simulated my Mün mission that required switching to another craft for a sec so that Bill wouldn't have run out of oxygen. Or that I could have simulated my reentry and not have killed Bob with the G-forces from that Mün return. If I could have simulated them, they'd all be alive.

[Whirligig Girl]

Wind Tunnel

A special modified flight scene in a wind tunnel. A slider controls windspeed from 0.5 m/s to 1.5 km/s, and another slider controls atmospheric pressure and gravity.

Lander/Rover Test

A modified ground scene in a large hangar. Sliders control how rough the terrain is, how much gravity there it, and the atmospheric pressure. Cranes can lift a vehicle to up to 50 meters above the terrain and drop it.

[xrayfishx]

The wind tunnel and Lander/Rover test have nothing to do with what I want. 50m isn't enough to know if I have enough parachutes or if I could reenter into the atmosphere. Well, they could be added, but at higher techs I'd expect a computer simulation that I can rely on.

[Elington]

But I can't calculate how many parachutes I need to land a probe on Duna.

Yes you can!

=> http://forum.kerbalspaceprogram.com/threads/59899-NET-AscentKomputron-rocket-design-%28reads-craft%29-ascent-optimization-aerobraking

And of course I do agree with you, also because IMHO trying to fly a complex mission "first time right" is the most rewarding gameplay by far...

Cheers,

David

[kalor]

I presume there will be an economy at some point "soonâ„¢", so there would be a good incentive to use such a feature. How about a compromise.

We all know simulations (even BAD sims) aren't cheap to make and run. For now this isn't a problem, but could be in the future. How about adding in a cost to operate said sim? Say, 5-10% of the cost of the vehicle being simulated (gotta pay the Kerbals to code and design the simulated craft in the simulated simulation, after all!) with a (probably small, unless the tech tree is worked in...) margin for error, to still leave a good reason to test these things manually, or to fly them without simulation?