Why is there a conflict of interest at all between realistic and unrealistic aerodynamics?

[tater]

Simulation, like everything else, is not binary. It's not a choice between "complete realism" (whatever the heck that is supposed to be) and "doesn't matter, do whatever."

KSP is absolutely a simulation. It imitates the appearance or character of spaceflight (that's the definition of simulation in italics). Is it Orbiter? No, but neither is Orbiter KSP. Orbiter doesn't simulate the design of spacecraft at all, for example. What is simulated, and the quality of that simulation is the trick to balance. The KSP treatment of the business/operations end doesn't even imitate the appearance of same, IMO. If it meets any standard of simulation, it's a very poor simulation there compared to the orbital element of the game.

So we can all be on the same page that KSP is a simulation game, though we might disagree on the quality. The metric for judging that quality would be how close our real world expectations of actions (the physics) are to the results in game.

Now to the game bit.

The metric of a game, is gameplay. Gameplay is certainly impacted by the simulation, but choices in simulation do NOT have simple effects on it. The current career/tech system is a very poor simulation, and is also… very poor gameplay. Making the career more realistic, contrary to people chucking straw men around, would not consist if Squad forcing the players to make spreadsheets, endure boring meetings, and all the work-a-day stuff that happens in any real administrative office. No, realism in simulation would be if the type of missions presented, budgets allowed, etc "made sense" in the verbal world. The outcome is what matters, not the black box. The orbital mechanics, launch, landing, even craft design remain fundamentally unchanged, regardless of the aero model. Some new parts to play with (heat shields, and fairings), and likely some changes in design choices… that's it. Even FAR changes nearly nothing aside from that, as I have said, I don't even notice it's there, and I only tested a couple rockets (intentionally breaking them) after FAR install to realize how kooky I'd have to fly to have any problems (no comments on aircraft, I have yet to build any, I have no interest).

If atmospheric realism made getting to orbit profoundly difficult, then yeah, that would be a gameplay effect. The simple fact is it doesn't. Once again to use FAR as an example, I've barely glanced at the metrics shown in FAR (in the VAB, flight, or otherwise). If I was pushing the edge of what should not fly, maybe I'd have to, I dunno. I have yet to need them, though I'm curious enough about what is going on that I sometimes watch them in flight.

[Parkaboy]

Well, if what I say is an opinion and what you say is not, then yes.

Examples:

This is fun - Opinion

This is what people think - Statement of fact

However, since you are the only judge in the debate of what is opinion and what is fact, things get a little skewed. So I guess I have to be careful to phrase everything very, very carefully. For instance, if you are of the opinion that more realism equals more fun, I would conclude that you would be much more amused by the game if time warp were to be removed, since NASA can't possibly distort time to speed up their missions in real life. I, however would find that crazy - that's my subjective opinion, too. I would imagine that it couldn't possibly be the case. No one would like to do a mission to Eeloo in real time - or at least, in my opinion, they wouldn't like it for very long. So I would conclude that when you state that realism equals fun, you might not be talking about all kinds of realism. My impression, then, would be that you haven't given much thought to what you're talking about.

[rcp27]

I think the issue is not one of "realism is not fun", but one of how to make realism fun to play with. For the game to be fun it doesn't have to be easy, but it does have to work in a way that the player can understand (consider, for example, the arguments that have been made regarding random part failures). For those things that are intuitive, like bigger engine = more thrust but higher fuel consumption, this is simple. For those things that are not intuitive, like orbital mechanics, the game provides clear and intuitive feedback to the player. Burn the engine in a particular direction and watch the orbit change in the map mode. Adjust the orbit, and watch the encounter markers move around.

The problem with aerodynamics is that, for realistic aerodynamics, the behaviour is not intuitive and it is difficult to provide clear feedback. How do you indicate to a player that this loss of control was caused by burning fuel from the front tanks before the rear tanks, that loss of control was caused by the transition from subsonic (CoL forward) to supersonic (CoL further back) and the other one because the higher aspect ratio wings having a lower stall angle of attack than lower aspect ratio wings? If a player doesn't have a method of understanding why their super-cool looking plane crashed in a ball of fiery destruction, how are they supposed to make the next design iteration work better?

Fundamentally, people have a basic expectation of how planes fly, fed by playing arcade style flight games, watching TV shows and movies, and being flown around in airliners. People have an expectation of what a plane should look like based on real world aircraft they have seen. On the whole these expectations are incomplete or down right wrong. In producing an improved stock aerodynamics model, the developers therefore face a choice. They can either make an aerodynamics model that matches players false expectations, so that players can build intuitively and iteratively, or they can make a realistic aerodynamics model that requires the players to un-learn their false expectations and come to terms with the reality. The problem with the second approach is one of how to guide players through this process in a way that is actually fun or at least intuitive.

[Alshain]

However, since you are the only judge in the debate of what is opinion and what is fact, things get a little skewed. So I guess I have to be careful to phrase everything very, very carefully. For instance, if you are of the opinion that more realism equals more fun, I would conclude that you would be much more amused by the game if time warp were to be removed, since NASA can't possibly distort time to speed up their missions in real life. I, however would find that crazy - that's my subjective opinion, too. I would imagine that it couldn't possibly be the case. No one would like to do a mission to Eeloo in real time - or at least, in my opinion, they wouldn't like it for very long. So I would conclude that when you state that realism equals fun, you might not be talking about all kinds of realism. My impression, then, would be that you haven't given much thought to what you're talking about.

Well for starters, that isn't my opinion. My opinion is that a more realistic aerodynamic engine is more fun, not that all aspects of the game should be realistic. Even other mods, I don't like life support mods, I don't like DRE (I'm told that weird that I like FAR and not DRE), I don't like RSS. These are all my personal preferences.

Second, I have never once stated realism automatically equals fun, I will say that realism in the aerodynamics is fun for me. Your putting words in my mouth.

Finally, FAR is not absolute realism either. It's just a closer approximation.

Edited January 15, 2015 by Alshain

[Fleb]

I am happy with how Squad have decided to handle the aerodynamics personally. FAR is a great mod and I have played with it quite a lot, but it seems a little too beyond the scope of KSPs ease of play imo.

[tater]

Just stop with the "remove warp" or forcing Jeb to use the bathroom as examples of realism. We are entirely limited to discussing outcomes of narrow space flight scenarios. That's it.

[karolus10]

I think that I may be blunt with this one but I think that realism is the wrong word in this case, it's not about realism but plausibility, it's not possible to perfectly model everything so always there have to be some simplifications to make it viable.

I think that we need to ask themselves if all basic mechanics makes sense and if the model is good enough. Good enough means that it's simplified because of the game limitations but it is still believable.

Making aerodynamics to work in the similar way it would be expected in real life is actually more intuitive (especially with airplanes) and it may add some challenges at the beginning but also trial and error learning is one of the most enjoyable parts of the early game.

[Duxwing]

I get enough realism in my life

Is your life not fun?

---

Aerodynamic realism preferences seem distributed among three large groups: stock aerodynamics users, NEAR users, and FAR users. Many of each group have likely considered each option and chosen the one they most prefer, making all universal plans inconvenience some players. The question therefore is utilitarian: "What stock aerodynamic system would create the greatest good for the greatest many?"

From a recent forum poll we can infer that almost all players would prefer such a realistic aerodynamic system as FAR--even if it broke backward compatibility. I argue that we should do their will for not only their sake but that of future KSP players, who would know only realistic aerodynamics. These aerodynamics and their implications would need only such brief explanation as in KSP's v0.13 tutorial; e.g., "Turning a swiftly-climbing rocket will expose to the air its side, which will turn the rocket like a rudder turns a boat." when the player first achieves high dynamic pressure.

Players wanting simple, intuitive aerodynamics should consider that realistic ones usually are simple and intuitive. Sleek things slip through the air whereas fat things drag like boulders; payload fairings and heat shields protect delicate payloads; cargo bays and inline parts let air smoothly pass. The only complex or counterintuitive mechanics I have observed involved exotic flight regimes, which should be strange, or became intuitive once experienced; e.g., in flight, the center of lift tries to follow the center of mass. Realistic aerodynamics, I therefore further argue, are simple and intuitive during normal flight and offer whoever ventures beyond it fascinating challenges.

Dynamic instability would be among these challenges. It involves putting the center of lift before the center of mass, which the center of lift will try to follow whenever the craft turns, making tiny control inputs very powerful. The challenge is controlling the instability lest the craft should turn uncontrollably--and the challenge is optional! Putting the center of mass before the center of lift tames maneuvering enough for peaceful, steady flight.

-Duxwing

Edited January 15, 2015 by Duxwing

[Jouni]

In general the proper way to launch a rocket is to launch straight up, turn about 5 degrees, and turn off SAS. Then let gravity turn you the rest of the way as you go. That's why it's called a gravity turn. A perfectly designed rocket will do the gravity turn without you needing to touch the controls.

That's a nice goal, but it rarely happens in practice. Most of my payloads are much wider than the central stack of the rocket launching them. Without active guidance, such rockets would tumble and disintegrate within the first minute after the launch.

[problemecium]

My two cents is that there's a possible performance hit.

Now I'm sure a lot of you can tell me that you get 785 frames per second using FAR, but! I can say that there are a lot of people around here posting their computer specs in their signatures, and more likely tan not you only get 785 FPS because your computer is so beefed up that it could run stock at 1000 FPS. I can also say that there are threads all over this forum where people complain about bugs in FAR or stuff not working with FAR.

Long story short, more realistic aerodynamics are harder to compute and introduce more possibility for faults, leading to reduced performance and stability for the game. I applaud SQUAD for endeavoring to try and improve them, but I envy them not the task of doing it without hurting the rest of the game.

[KasperVld]

I'd like to take one post in this thread to go a little off-topic and leave a compliment for everyone who's partaking in this discussion, I've read some very insightful comments here, for example Tw1's comment in the beginning of the thread where he questions the very notion that we should be discussing 'fun versus realism', instead of the game using a more realistic approach to reach its potential. RCP27 also made a good comment about the delicate balance between realistic simulation and intuitive gameplay. Those are just two examples of what is really one of the best discussions I've read on the forums in a long time.

As for my personal experience I can only say that I often install NEAR with my KSP, but also play with stock aerodynamics a lot, for example when we're testing experimental bugs. The differences between them are obvious at this point, but I won't make an argument either way as I'm not a game designer or developer. What we can see in both this topic and throughout the entire discussion of this subject is that there's a giant range of preferences, and I personally think that's both complicating the discussion and a great thing to see. I'm excited to see what HarvesteR et al come up with, but perhaps just as important is the fact that HarvesteR restated his commitment to making (keeping) the whole system moddable, so that he can focus on getting KSP to work to the satisfaction of most players, and allow people to tailor the game to their specific preferences if they prefer a different balance.

Keep it up!

[Louella]

There is a whole lot of stuff that is involved in aerodynamics and the atmosphere, that you're not aware of, and there's no feedback from the game from.

like, aerodynamic drag during ascent of rockets, the terminal velocity table for Kerbin for a fuel-efficient climb. I don't know how that was worked out, and I wasn't aware of it until recently.

Knowing about it, means I can make a spacecraft reach orbit with a whole lot more fuel, by limiting acceleration in the ascent, to keep speed below the value in the table.

Aeroplanes shift balance when fuel is consumed. A plane I have that is very stable at full tanks, becomes almost unmanageable at full throttle with less than 1/4 fuel remaining. You have to pump fuel from the rear tanks to the front tanks to keep it easy to control.

If the atmosphere becomes more "realistic", whatever that means, given the dimensions of Kerbin, then there needs to be a lot more feedback and information from the game, such that players can do things, and see the results of those things, and have some idea of what happened and why it happened.

the worst situation in Kerbal Space Program, is looking at a failed flight, and not knowing why things happened that way.

Aircraft flips at high speed due to breaking the sound barrier ? well you have to know that there is a sound barrier, and why breaking it made the aircraft flip, and that sort of thing isn't shown in the game. All you'd see currently, is the aircraft flip, possibly breaking up, with little information as to why.

[MacroNova]

There seems to be an assumption among many in this thread that the current stock atmosphere is more fun than a more realistic atmosphere would be. In fact, I think a more realistic atmosphere would just be different. It might require relearning a couple of things, but once you learn them, flying in the new atmosphere will be just as easy as the old one. And heck, a new player won't even know the difference.

Some people find it fun to launch crazy monstrosities into space, which would be harder with a more realistic atmosphere. But some people also find it fun to design rockets that are aerodynamically logical and will tip over if poorly designed. Plus you get that satisfying feeling when you get to space, release the fairings and reveal your spacecraft in all its spaceworthy glory. Again, this isn't about difficulty, but people having different tastes.

So put me down for preferring more realism in the atmosphere, and also put me down for not believing this would make the game more difficult or less fun.

[DBT85]

This is how I look at it.

Realism - Not having a soup under 15km, pointy things being better than flat things.

Fun - Me not needing a degree in Mathematics to slam Jeb, Bob and Bill into the Mun.

The game can be both fun and more accurate at the same time.

I only started using FAR a week or so ago and have had no issues at all with it. Deadly Re-entry has caused me more issues.

Re-entry heat/speed damaging things really can just be an on off button. It affects it or it doesn't.

Having two separate aero models that make things FLY differently is silly. I'd be happy right now if they just made FAR stock.

Edited January 15, 2015 by DBT85

[Accelerando]

If the atmosphere becomes more "realistic", whatever that means, given the dimensions of Kerbin, then there needs to be a lot more feedback and information from the game, such that players can do things, and see the results of those things, and have some idea of what happened and why it happened.

the worst situation in Kerbal Space Program, is looking at a failed flight, and not knowing why things happened that way.

Aircraft flips at high speed due to breaking the sound barrier ? well you have to know that there is a sound barrier, and why breaking it made the aircraft flip, and that sort of thing isn't shown in the game. All you'd see currently, is the aircraft flip, possibly breaking up, with little information as to why.

Yes! This touches on something that I've been thinking about, too, and that I've given severely little thought to in the OP of this thread - that one of the great appeals of the game, and one of the ways in which its potential seems, in my opinion, the most severely neglected, is its potential to educate, and the potential fun that that is.

When I first started playing in the early, early days of KSP - back in the first releases when it was a free game - I had no idea how orbits were achieved, or that there was anything such as an optimal flight profile that could be achieved by throttling the engines backward or forward; I had no idea that rockets are supposed to fly east to get the greatest speed boost from their planet's rotation, if counterclockwise; I had no idea that a rocket's fuel efficiency could be neatly and easily described by a single number, Isp, and that I could use a simple equation to give me a rough idea of their "range". I had to learn all of this, and it was fun, and the reason I could do any of that and have any of that fun was because the way rockets behaved and behave now in the game is realistic in some way - it wasn't simply a hard game mechanic, like learning to grapple with Dwarf Fortress's arcane control system. I actually -learned- something! Why shouldn't this figure into peoples' considerations of fun?

But at the same time, and to address your point - as the game grows in complexity, it helps greatly to increase the amount of feedback, the amount of instrumentation, the amount of data and analysis the game gives to the player, in order to help them understand and make sense of what's going on around them. For example, I remember hearing something from HarvesteR along the lines of refusing to implement a delta-V readout or a set of aerodynamics readouts and instruments because "it's not fun to know everything about your rocket"; but that is not true and not good gameplay. It simply creates an artificial obstruction for players who want to learn the real physics, and makes the game seem more obtuse and arcane to players who are unfamiliar with such concepts.

EDIT: As a further example, take the orbital Map view and all the data that it currently provides. The orbital map provides fairly plentiful information on your orbital characteristics, and by doing so, it allows me to set clear goals for myself and understand whether or not my maneuvers have succeeded - and by looking at the ∆V left, at the inclination and dimensions of my orbit with regards to the planet, the ecliptic, other orbits, and suchlike, I can understand why they did or didn't succeed. If the game didn't have these readouts, it would not be more fun - it would be infinitely more frustrating. The fact is, I need to know this data anyway if I am going to advance my understanding of the gameplay; you can't just thrust randomly in space and expect to "hit" Duna without knowledge of phase angles, much less knowledge of the dimensions of your very orbit. Likewise, I think that providing better, clearer, and more insightful instrumentation and data for all other aspects of the game - some kind of ruler in the vehicle assembly building and spaceplane hanger would be muchly welcomed, for instance - will open up far more opportunities for education and creativity than denying them to players. It is infinitely more enlightening to be able to see your orbit grow and shrink, and to be able to watch the numbers go up and down for its periapsis, phase angle, and so on, at the same time, than to simply watch a blue curve move around a planet, or even worse to have no orbital display at all.

And it's not as if people don't run out of fuel even though they know their rocket's delta-V, or crash aircraft even though they have all the readouts anyway - people run out fuel all the time in cars IRL that don't deplete their supply in such unintuitive ways, and people crash test aircraft in real life even though they have plenty of computer data. Knowing and being able to keep track of more or less how much range you can expect out of a car or the flight characteristics of an aircraft does not make driving a car or flying a plane less fun! But not being able to know without running the calculations yourself or installing a mod does make it more frustrating for people who want to be able to learn.

It certainly helped me a lot when I started using MechJeb in .17 and I could begin turning to its readouts to learn about phase angles and watch my delta-V fall. So yes - it's definitely very helpful, very educational, and fun to play the game with realistic elements for the purposes of learning, and it definitely helps and increases that fun to have bountiful data available on my flights! And if you want a challenge, you can go without the instrumentation - but just because you've played it long enough for that to be a challenge doesn't mean that that's the only challenge there is, or that's worth undertaking.

I'll reiterate this in the OP, because it's absolutely exactly something I want to address, but completely forgot to.

Edited January 15, 2015 by Accelerando
Adding in an example

[[email protected]]

Fun vs Realism is a false dichotomy.

The game is fun in part because of its semi-realistic physics.

Slapping things together and seeing how they work in a realistic environment is fun. Obviously most of us don't want to have to micromanage all the way down to individual nuts, bolts, and o-rings, but we do want the manner in which we slap something together to have an impact on how it performs. Stock arodynamics are absurdly forgiving and punishing at the same time. I can fly a giant brick into space with better efficiency than an actual rocket, but if I take those same rockets and accelerate into a full power dive I will actualy start slowing down...substantially.

[Crimson Sunrise]

Well, there's only one conflict I see, and this one happens every time which is a conflict between what people want and what the developers think people want. Since AFAIK, developer posting seems to be 'business as usual' down to trying to sound as corporativistic (sic) as possible, it's no wonder this communication failure happens and you end up with people complaining about features that were mentioned but are not present when the time comes...Or the abundance of beating around the bush.

[Nemrav]

Well, here is from what I see the discussion should be a bit closer too, given the arguments.

-At some point all components of physics will have to be cut to make sure you don't need paper and pencil to work things out (n-body physics and stable orbit, about)..

-everyone agrees that at some point too much realism isn't fun, n-body physics advocates would like n-body physics but might not want solar wind and stray particules effecting your orbit type simulation.

-so the discussion should be closer to at what would be a compromise the FAR and the Stock-NearIsh people could work with, remember you must give in something to make the argument solvable no matter which side of the argument you are on

Maybe for a starting point we could have -shielded and pointy > flat and exposed , Mach effects if you go ludicrous speeds higher than the sound barrier and have them deadened down (have publicly modable variables for this so its adjustable) mybe in a setting slider...

In return, make control surfaces op and keep sas op ?

*that was just a draft, I don't have 5 hours to consider everyones opinion....

[Alshain]

Mach effects if you go ludicrous speeds higher than the sound barrier and have them deadened down (have publicly modable variables for this so its adjustable) mybe in a setting slider...

Very good post but I wanted to point something out on this specifically. You could just redefine mach. The speed of sound is the speed at which sound travels through air... on Earth. Earth isn't in this game and for all we know Kerbal's breath Hydrogen Sulfide and their "Air breathing" engines do too. That would alter the speed of sound.

[cadaverific]

I'm replying to this just because you mentioned dwarf fortress, my favorite hated game. The only other game where losing is !FUN!

oh and here's my opinion on the aero debate:

[Scramble]

True realism should not be a target for the stock product. Why? Because you as a player would have to become NASA. Real space programs are not run as a part-time activity by one single person. Even if you quit your job or studies to do KSP full time, you could never run a full space program. You don't design a truly realistic space shuttle as a non-expert as an evening project. There is no question there is a balancing between realism vs fun.

For instance as OP is asking for true aerodynamics, does that mean there should also be truly realistic weather effects with seasonal variations? Would you want to cope with having to use complex weather forecasts to plan launches and landings, and having to land your already barely stable airframe with unstable crosswinds and near ground turbulence by using the not so realistic "pushing asdqwe keys" control method? How would you eyeball how to enter the atmosphere in order to get where you want based on what is supposedly going on in the upper atmosphere? Miss a launch window due to bad weather and wait a few years for the next opportunity? And should KSP be updated to include a autopilot module, which you would have to program yourself to do automatic landings as is my understanding is what the real shuttles did?

I do appreciate high levels of KSP realism, but for that I let Scott Manley get paid from YouTube ads to install various mods and do all the heavy work, and just enjoy the added complexity by watching some of his videos.

[Rocket Farmer]

At the end of the day gameplay can be made much better without having to stick to reality. Without fast forward this game would lose a lot but that isn't exactly realistic now is it? Again kerbals needing air/supplies would greatly reduce most people's ability to play the game.

So instead design the game for the common player (you know, they guy that probably will never figure out how to get to Duna) and let the mod community fulfill the requirements of the elite 5% that want more realism.

That's where Squad will go and it makes sense financially as well.

If you expect otherwise prepare for a disappointment.

[tater]

True realism should not be a target for the stock product. Why? Because you as a player would have to become NASA. Real space programs are not run as a part-time activity by one single person. Even if you quit your job or studies to do KSP full time, you could never run a full space program. You don't design a truly realistic space shuttle as a non-expert as an evening project. There is no question there is a balancing between realism vs fun.

As every proponent of having things be as realistic as feasible within a game says in every one of these threads, no one is asking for "true realism." The "you'd have to be NASA," or "you'd have to model bathroom breaks for Jeb," or "you'd have to model the office workers in the admin building standing around waiting to go home" straw men arguments need to just stop. No one is arguing for what anyone characterizing the "realism crowd" says they are.

Squad is removing the awful, placeholder atmosphere model, that is a given. their goal is to have it more realistic than it is now, also a given. Many of us want them to do it as well as they can within the limits of having the game run smoothly, that's it.

For instance as OP is asking for true aerodynamics, does that mean there should also be truly realistic weather effects with seasonal variations? Would you want to cope with having to use complex weather forecasts to plan launches and landings, and having to land your already barely stable airframe with unstable crosswinds and near ground turbulence by using the not so realistic "pushing asdqwe keys" control method? How would you eyeball how to enter the atmosphere in order to get where you want based on what is supposedly going on in the upper atmosphere? Miss a launch window due to bad weather and wait a few years for the next opportunity? And should KSP be updated to include a autopilot module, which you would have to program yourself to do automatic landings as is my understanding is what the real shuttles did?

I do appreciate high levels of KSP realism, but for that I let Scott Manley get paid from YouTube ads to install various mods and do all the heavy work, and just enjoy the added complexity by watching some of his videos.

Another straw man. OP is arguing nothing of the sort. One, the thread is about why people have the (entirely unsubstantiated) idea that somehow which algorithm is used to calculate drag---that no one can see---is assumed to "reduce fun" if it is arbitrarily choice A over choice B. Two, OP is right, it's a false dichotomy in general (otherwise a flight sim with no gravity would always be more fun than one with gravity, for example).

I'd wager that Squad could drop in FAR (minus the graphs and stuff that lets you know it's there)., tell everyone they opted for "fun" at the expense of doing all they could do… and none of you would even notice. Yes, you'd notice the atmosphere was changed, but I bet you'd have more fun than if they told you they added FAR in.

[OhioBob]

Very good post but I wanted to point something out on this specifically. You could just redefine mach. The speed of sound is the speed at which sound travels through air... on Earth. Earth isn't in this game and for all we know Kerbal's breath Hydrogen Sulfide and their "Air breathing" engines do too. That would alter the speed of sound.

Although we don't know the exact composition of the atmosphere, we know its molecular weight. We are given following:

Sea level density = 1.2230948554874 kg/m³

Sea level pressure = 101325 Pa

Sea level temperature = 293.15 K

From this we can calculate R_specific, the specific gas constant, which is 282.6 J/kg-K.

And from R_specific we calculate the molecular weight of the air, which is 29.42 kg/kmol.

We know oxygen is present (mol. wt. = 32) so, given the above molecular weight, the bulk of the remaining atmosphere is almost certainly nitrogen (mol. wt. = 28).

The only additional piece of information we need to calculate the speed of sound is the specific heat ratio γ. Given the known composition and temperature, γ = 1.40.

The speed of sound is, therefore, about 340.6 m/s at sea level.

Although the temperature on Kerbin changes with altitude (as recorded by a thermometer), this is ignored in the thermodynamic model. Pressure and density are computed as if there is a homogeneous atmosphere at constant temperature. Because of this, the speed of sound does change and is constant everywhere.

(ETA) If we assume the atmosphere consists only of oxygen and nitrogen and no other gases, then, given the average molecular weight, the atmosphere must consist of 35% oxygen and 65% nitrogen by volume.

Edited January 16, 2015 by OhioBob

[Accelerando]

For instance as OP is asking for true aerodynamics, does that mean there should also be truly realistic weather effects with seasonal variations? Would you want to cope with having to use complex weather forecasts to plan launches and landings, and having to land your already barely stable airframe with unstable crosswinds and near ground turbulence by using the not so realistic "pushing asdqwe keys" control method? How would you eyeball how to enter the atmosphere in order to get where you want based on what is supposedly going on in the upper atmosphere? Miss a launch window due to bad weather and wait a few years for the next opportunity? And should KSP be updated to include a autopilot module, which you would have to program yourself to do automatic landings as is my understanding is what the real shuttles did?

Another straw man. OP is arguing nothing of the sort. One, the thread is about why people have the (entirely unsubstantiated) idea that somehow which algorithm is used to calculate drag---that no one can see---is assumed to "reduce fun" if it is arbitrarily choice A over choice B. Two, OP is right, it's a false dichotomy in general (otherwise a flight sim with no gravity would always be more fun than one with gravity, for example).

Yes, Scramble, please do not put words in my mouth. I never said anything about having to explicitly deal with every fine minor detail of rocket construction and launch - I have even, multiple times, explicitly supported the "armchair commander" model of flight control and voiced my dissent against simply requiring the player to push every possible button at every second of the mission, give an alternative, and explain why this is neither un-fun nor unrealistic. Even in real life, not every person in mission control has to worry about every other aspect of a rocket's systems or its flight plan in order to be able to do their part, operate the controls, and make it work. Everyone splits off some work somewhere; that's true of any job. There's nothing fundamentally incompatible between making a game realistic and choosing the elements of realism you want to include and in what way to implement them that they'll be fun. Since I don't seem to have gotten through, I'll point out the places where I tried to clarify my opinions, for you:

I remember that thing HarvesteR said a while back along the lines of how if you wanted to play a perfectly realistic game, you'd go outside. But how many of us can actually reasonably expect to go out and be able to fly a spaceship tomorrow? Very few if any; much as most FIFA players can't expect to go outside and start a career in pro football tomorrow, or how most Tycoon players can't expect to go outside and start growing a business to massive proportions; you get the idea. I think it is neither un-fun nor unrealistic to gloss over certain grating parts of the challenge or leave it up to someone else or a background mechanic to take care of; leaders and executives in real life do this all the time, for one. There's clearly elements of real life that are fun and people want to do, and ways to present them so as to skip over the parts people don't want to do.

That's true, and I edited your second post in to account for that, since it lined up very well with what I was saying to John. I personally consider it more than realistic to simply kick back behind an armchair and direct someone or something else to move their arms to push the controls of a faraway vehicle; real pilots do it with drones today. I've never heard anyone criticize, for instance, Dwarf Fortress for being unrealistic by not forcing players to manually carry out calculations for moving every pint of blood in every artery of every animal in order to advance the wildlife simulation going on around them; it just happens, as is realistic and as is conducive to gameplay. Such players' opinions lie in a different arena of realism, I think, rather than being a matter of simply more realism.

So, yes - please stop trying to strawman my posts and use slippery-slope fallacies. This idea of "all realism advocates are clearly being sticklers for 'perfect hardcore realism of actually putting on astronaut boots and stepping into vacuum' or else they don't know what they're talking about" seems far too common among some people, and I agree with Tater here 100%:

As every proponent of having things be as realistic as feasible within a game says in every one of these threads, no one is asking for "true realism." The "you'd have to be NASA," or "you'd have to model bathroom breaks for Jeb," or "you'd have to model the office workers in the admin building standing around waiting to go home" straw men arguments need to just stop. No one is arguing for what anyone characterizing the "realism crowd" says they are.

Squad is removing the awful, placeholder atmosphere model, that is a given. their goal is to have it more realistic than it is now, also a given. Many of us want them to do it as well as they can within the limits of having the game run smoothly, that's it.

Edited January 15, 2015 by Accelerando