Public Service Announcement regarding Aero

[luizopiloto]

But as I said before on another thread, new atmosphere is fine... what need a closer look are the air breath engines... they are not performing so well, loosing too much thrust at high altitudes and draining a lot of fuel... they got balanced for the 1.0 model... not for the 1.0.2...

As an example... League of Legends... every time they change something, even an insignificant one, it can unbalance the game a lot later...

Same thing now happens with KSP. Now we have a default aerodynamic cfg. All the parts have it's own performance. Nothing is dynamically connected. any change at the atmospheric settings will unbalance all the other things...

Think about the craft sharing part of the community... Today all crafts works, Squad changes atmosphere, All crafts get unbalanced and needs rework...

Edited May 5, 2015 by luizopiloto

[Brotoro]

...you have to flutter throttle and watch the temps, etc...

I don't understand the term “flutter throttle.“ Please clarify.

[sal_vager]

Fuel consumption in .90 was about fifteen times higher than the most efficient human jet engine, Kerbals might be good but that good?

And there's not a lot of air up there, jets have a limit...

[regex]

I don't understand the term “flutter throttle.“ Please clarify.

Increase/decrease throttle as needed. It's more apropos to cars/motorcycles where throttle response time is more immediate and you can "flutter" the carburetor valves (or the throttle body, or whatever) but old terminology dies hard with me.

- - - Updated - - -

But as I said before on another thread, new atmosphere is fine... what need a closer look are the airbreath engines... they are not performing so well, loosing too much thrust at high altitudes and drainning a lot of fuel... they got balanced for the 1.0 model... not for the 1.0.2

You mean, they're acting like they should? Good.

[FleshJeb]

Probably in the minority here.

"Here" is the extreme end of the playerbase. A significant and unrepresentative proportion of players here were using FAR, RSS, DRE, etc.

{Overwhelming, etc.}

tl;dr if you thought it was viable before, it's just as viable now - the only thing that changed is how comfortable you are with the game. Give it 20-30 hrs of gameplay and then consider it again and I bet you'll find it just as viable as before.

I will consider that. So far I've only got one rocket and 5 planes built post-1.0. I have about 2000 hours of stock to unlearn.

Using KER and MechJeb, I could definitely teach a pre-teen how to build and fly a rocket. The adjustment might be a matter of just saying, "Add fins to the bottom".

{Moar boosters!}

Here's how to teach Isp / thrust relationship:

1. some engines are meant to be used in atmosphere.

2. some engines are meant to be used in space.

3. don't use space engines in the air. They have bad (small) numbers here where it says "Isp (atm)". Only use them in space. (they can learn the fine distinctions once they're older)

Or even simpler - "If this engine doesn't work, try another." "Moar boosters!:!KLJ~:ljkadfajsdasdfkljad"

That might work, thanks. The two kids I'm thinking of could definitely handle that.

--------------------

I'm definitely looking forward to nuFAR, as nuStock handles occlusion badly and VERY unintuitively for tiny parts. I like my panel van planes, and the wings/skin panels SHOULD protect things under the skin.

Edited May 5, 2015 by FleshJeb
nuFAR

[Temeter]

You mean, they're acting like they should? Good.

Oh come on, save us that 'like they should'. The same mindset can be used to say 'orbital speed should be around 7800m/s and spaceplane reentry needs to take 6 hours because realism'. Atm those turboramjets are the only supersonic engine in the game and they get barely anywhere because of the high fuel consumption. And said fuel consumption 1st hardly even matters for the weight of a spaceplane and 2nd cripples every other 'long range' plane you took out to do stuff on Kerbin. Then 3rd Rapiers come around the corner, they are a bit faster and, since you'll need max thrust at the highest speed, make trj's kind of useless.

Maybe it would be different if we had some intermediate engines (note that rapiers still would replace them), but we don't. In my eyes this is a part where realism resulted in a net loss, there is no comfortable way to go explore Kerbin anymore.

[Wallygator]

It seems to me, that having a more properly aligned aero and physics model has now shown a few gaps in the various air breathing engines available.

I would rather have a wider selection of engines to fill any gaps (within reason) instead of "fixing" the physics so that the current engines regain lost capability.

DISCLAIMER: I am not a jet pilot, I forgot most of my physics and calculus before most KSP players were born, I admit I do not remember enough to judge a specific engine and its performance environment.

[luizopiloto]

when I say higher altitudes I'm not saying 20k+.. I'm saying 10k+... the phase the spaceplane normally accelerates to exit atmosphere... turbojets are fine, but the Rapiers looses a lot of power above 10k... A medium sized spaceplane needed 4 Rapiers, now it needs 6 to compensate the new atmospheric settings... that's the point... if Squad changes atmosphere default settings, they have to re-balance the atmospheric engines, wings, body parts drag... etc... anything that uses a MKIII part now feels like trying to make a Bus fly...

[regex]

when I say higher altitudes I'm not saing 20k+.. I'm saing 10k+... the phase the spaceplane normalli acelerates to exit atmosphere... turbojets are fine, but the hapiers looses a lot of power above 10k... A medium sized spaceplane needed 4 Rapiers, now it needs 6 to compensate de new atmospheric settings... that's the point... if Squad changes atmosphere default settings, they have to rebalance the atmospheric engines, wings, body parts drag... etc... anything that uses a MKIII part now feels like trying to make a Bus fly...

While some things do need to change, you can make it to orbit on two, or even one RAPIER (craft mass matters, really) by diving to build up speed and break through the transonic barrier to about 450m/s+. It's kind of tough and fiddly to do, I generally fail about 1/3 of the time, but when you get it and bust out into spaaaaaaace, oh, the feels man.

There's also a problem with the Mk3 cargo bay not occluding properly and applying a ton of drag.

[luizopiloto]

While some things do need to change, you can make it to orbit on two, or even one RAPIER (craft mass matters, really)

I use the old TurboSpike... works way better...

[Fraggle]

I feel like I started this at the right time.

At "release" instead of in the beta phase. lol.

I don't think I quite understand the arguments fully though.

Some people seem upset that the new edits are not "realistic" enough?.

Is this supposed to be a sim?. Because its not set around Earth in a known solar system.

Okay so I understand a lot of people put a fair few hours into the game too... But that was during Early release, and one always has to remember that Early release are often subject to change, often quite dramatic. (I dropped over 90 hours into Rift pre-release only to have a lot of my efforts killed by a patch and update).

I love the idea of the game version, but theres a part of me that likes the notion of a pure-sim too. But right now I can see a game that's still developing and I think it would be unfair to tie its hands to "real world" so early on. Id love to see the Kerbal Solar system develop and grow.

Im loving the game, and once I get some serious time into it Ill probably setup a second career mode and set it to more known-world variables to Simulate our own system.

[PB666]

Or, maybe, and, here me out, even if I sound a bit rude, but, just maybe, the majority of players shouldn't have to edit their game's files or access a "Cheats Menu" to experience features Squad spent a significant amount of time implementing. At the moment reentry heating has lost a lot of its functionality.

I don't mind going to the cheat menu, oddly however I have not had to do this because of drag.

IMO, the drag is not bad going up, the problem for me is that since I use a combination of gear and chutes to land, the chutes do not catch quicky enough, I am resetting the chute open height at 1500 meters. And of course if you happen to accidentally engage your chute before re-enter cool down, you are toost.

Of course I did undertaking the building of Aero transition tanks prior to the release, there are ways of building to minimize upward drag.

My big problem with the game is the

CUBE tags have yet to be explained and the bulkhead profiles tag jumps from 1 to 2 to 3 and I'de like to see something more flexible.

can we define bulkhead profiles?

[sal_vager]

You two are arguing each others ideal of KSP, not strictly aero, and I'll say it one more time, discuss it privately if it means that much to you.

Otherwise be civil and don't attack each other.

[Hannu]

These are some of the questions that come up the most in the community. The point I'm making is that you can't answer any of my questions because you don't know what my physics config is like, and all of them depend upon that. None of the answers you gave were applicable to my game if we use different physics configs.

Some of my answers are quite correct in every reasonable set of physical parameters. For example real spacecrafts have TWR about that range and they start gravity turn soon after launch. But of course somebody may want extremities (for example something like KSP's old souposphere). It should not be too difficult to answer that everything depends on settings and it is impossible to get experiences with all possible combinations. I am sure that majority of players who ask basic things would not touch complicated physical sliders.

This gives us a grand total of 4 possible combinations: those who play stock, those who play FAR, those who play RSS, and those who play FAR + RSS. And yes, the community is already split into distinct, smaller groups because of this.

I do not see any problems in such splitting. It is great that SQUAD made so moddable game that very large number of different players get a great game. Even they who likes homogenous community because probably most players do not install physical mods.

Things would be different if there were choices in this game genre, but unfortunately there are not. Orbiter have nice flight physics but it does not give an opportunity to plan and build spacecrafts. It is best in KSP for me.

[allmhuran]

Some of my answers are quite correct in every reasonable set of physical parameters. For example real spacecrafts have TWR about that range and they start gravity turn soon after launch. But of course somebody may want extremities

That's true about real spacecraft, but it's debatable what counts as "extremities" when it comes to aero. In the three versions of stock aero that we've had so far (pre 1.0, 1.0, and 1.0.1/2) there have been three very different "optimal" TWR's and ascent profiles. It would be strange to say these three actually implemented, stock versions of aero counted as "extremities". So you can have customizations that seem reasonable enough to be the stock implementations, but which significantly fracture the experience of the players.

I think there's really no denying that physics customizations split the community. They do/will. As you've hinted at, the question is whether that's a good thing or a bad thing, but... well, I'm not sure whether that discussion can be had here.

[sal_vager]

Off-topic posts have been removed.

[MechaLynx]

Its the physics framework which needs to be consistent and believable.

The game's simulation of physics cannot be both consistent and improving at the same time. with so many deviations from realism (by necessity, both technical and practical) there is a lot of room for improvements to the physics engine, allowing for a lot more interesting engineering to happen, which is why most of us play this game instead of a game realistically simulating real spacecraft.

Like any simple simulation, it stops being believable when you reach its limits and all players that explore deeper will find those limits. With such a simple implementation, necessary to keep it real-time and playable, small changes will result in large effects on craft and physics, it's inevitable.

In the end, everyone has to bite the bullet and accept that the game will never be "realistic enough" to satisfy every case or even most, and it neither will stop changing since most any improvement to the core physics will alter how craft behave. Personally, I feel a good balance is getting as good a physical simulation as possible without requiring much more cpu/gpu/memory/hdd than the average modern game while keeping the initial difficulty curve accessible (endgame can be as hard as you want really).

If you believe that the core physics should never be changed and that it's a bad thing, ask yourself this: would you now revert to .23.5 physics, given the option? do you believe that would improve the game for the average player or even hardcore players?

isn't this going to be a problem for challenges?

The whole argument that these changes indirectly force players to customize their physics settings, thus making sharing craft a labyrinth, is pointless. The solution is trivial - each posted craft comes with the relevant .cfg files, in a nice zip archive. Even simpler would be to include just the .craft file with a diff of the default .cfg and the player's .cfg, to make it easier for humans to parse. In competitions and challenges, there should be surrogates that test the .craft/.cfg combination for validity - this isn't as difficult as it may seem, since most players experienced enough will know what to look for. In the end, it's not like this is some kind of ksp progaming scene where we need a high-accuracy anti-cheat. It's sharing for the fun of it or at most for challenges - the most spectacular accomplishments will be scrutinized the most and remember: people can already cheat a lot as it is.

Players can already cheat in challenges in a way that is nearly impossible to determine. Since KSP has no anti-cheat, there is nothing preventing a player from making small invasive adjustments, just enough to get an edge and then claim it's just that other players don't know how to fly the craft well enough or something. They can use a hacked client, custom mod (super easy), memory editing, video and image editing, glitches, the debug menu and any combination of these to get the desired result.

even from a set of pictures it's not hard to tell when someone has infinite fuel on

Nothing personal, but you are very wrong on this. It's only easy if they're doing it wrong. I can guarantee you that if someone has enough of an interest in the game to be on these forums, chances are they know about delta-v maps, delta-v calculations, mechjeb and ker, common community wisdom etc. - all they need to do is build a craft with the required delta-v for a trip or challenge, hyperedit themselves into position and change their fuel with enough random variance to be believable, take snapshots of each position at the expected delta-v and add some random trivia ("I lost my right fin during re-entry") to make it even more believable. Unless someone is cheating blatantly, it won't be obvious even to the experienced player unless someone scrutinizes it closely. In essence, most players can leverage community knowledge, mods and tools to cheat in near or entirely undetectable ways. And this has been true since at least .22.

So what I propose is that if you do make a change to any of the physics related settings and anything else that effects game-play, but not things like toggling aero-forces markers, entry heat bars etc, then a little icon should appear in the corner of the screen. The icon will be an indicator to challenge judges that the physics settings are not stock and (as a useful feature for the user) clicking the icon will restore settings to default. Do you think that's a good idea?

You're proposing an anti-cheat. Unless we have one that operates at the level of VAC, nothing will suffice. Even then, players would be able to win challenges and would get banned days, weeks or months later. Custom cheats would never get detected.

The problem of detecting cheating in KSP is the same as detecting whether a player deliberately disconnects from a competitive game to prevent a loss or their computer just fails - there is no way to do it with enough certainty without violating laws and rights much more important than the certainty of legitimate play. Thus, you are only left with 2 options: treat any suspicious action as penalty-worthy (this is how disconnects are treated) or use third-party validation (or just assume no one is cheating unless it's blatant). The latter is the solution here I think: each challenge entry must be evaluated by surrogate, non-competing players using challenge-compliant installations of the game. When sharing craft just to share, include both the .craft file and a diff of the player's .cfg files and the default, easy peasy. If lazy, just include the entire .cfg.

The only extra improvement I can think of is adding replay functionality to KSP - essentially, demo recording, either through a mod or the vanilla game itself. This way, the player in a challenge includes their .craft and the recorded mission and a trusted third-party runs it on their own installation to see if it checks out. KSP's engine AFAIK is deterministic and even if it wasn't, it would be easy to resolve by just including the seed for the RNG in the demo file. Problem solved. It can even record when the player accelerates time and adjust the simulation accuracy to match, thus allowing variable time acceleration for the viewer of the replay without affecting the accuracy of the replay itself.

It should be obvious that they who want unified stock experience do not touch default values. There are always some defaults by SQUAD which could be considered as "stock physics".

This. Is saying "Stock, 1.02, normal" so complicated, with the simple assumption that no physics have been messed with and asking for or posting changes (a diff?) of physics changes if pertinent?

Think about the craft sharing part of the community... Today all crafts works, Squad changes atmosphere, All crafts get unbalanced and needs rework...

The "consistent physics framework" argument is essentially a "backwards compatibility" argument, since it would never be an issue if we had this model from the start. And we know that, in practice, those that ditch backwards compatibility early win, those that don't are stuck with being dependent on it forever *cough* windows *cough*. In the end, it's extremely easy to fix when it comes to crafts and when it isn't, consider whether it's worth holding the game back so that some old crafts work.

I have about 2000 hours of stock to unlearn.

You have a knack for exaggeration. Remember you're not talking to 8-year-olds here, you're talking to adults who actually play the same game, don't be a smartass.

Atm those turboramjets are the only supersonic engine in the game and they get barely anywhere because of the high fuel consumption.

What do you mean "barely anywhere"? I just tried it with a mk1 spaceplane, 1 trj and 450 units of liquid fuel and at an altitude of about 16.5km you can get your fuel consumption to around .15-.10 - that lets you do about 60%-95% of kerbin's circumference at ~800m/s. How much more do you need?

Here is the craft:

http://steamcommunity.com/sharedfiles/filedetails/?id=438074530

Mind you, this is just 2 minutes of building - I haven't optimized anything. Pretty sure that with some gliding and finetuning for TWR and speed you can exceed a circumference of land distance traveled.

Most every spaceplane I've tried post 1.0 has either had too much speed or failed due to aerodynamic instability. I usually don't even get to spend all my fuel going up, especially with the trjs - my problems are usually that I'm not good enough to add some rocket engines and rapiers are too middle-ground to be really good at anything.

Edited May 6, 2015 by MechaLynx