Realism in Stock KSP

[Tyko]

20 minutes ago, Laie said:

I do not think that sheer delta-V equals difficulty, not in and of itself.

At some point it creates difficulties, granted, because we're only given tanks and engines up to a certain size, and duct-taping hundreds for a "Rescue Burberry" kind of rocket is indeed an engineering challenge. Whenever I'm promoting a real-scale system I'm silently (and often not so silently) assuming that dry masses etc will be adjusted to fit the scale.

Try building a stack that can launch 100 tons to LKO with a DV budget of 3400, then build a rocket that can lift the same 100 tons but with a launch DV budget of 6120 DV.

Then consider the challenges in Career or Science mode where you have limited access to parts. Just getting into orbit during early career is really difficult

Edited January 6, 2019 by Tyko

[Gargamel]

12 hours ago, Tyko said:

Currently it takes 4 different mods  .......I'm not sure if there's enough demand to support it,

 

The first part rules out the option of the second.   People are cheap.  If there's a way to do it for free, they will.   Plus, look at the % of players that actually want to play with a realistic universe, and then those who would pay for it.   That leaves a very small group of players left who are going to pay for such a DLC.

Yes, it would be better off as a stock setting.  Yes, a lot of the more vocal members of our forums would be some of the few who would pay for it.  But, sadly, No, it is not a economically viable project for Squad.   

Note, I've used this argument on many of the DLC threads, but I will be gladly wrong on any of those occasions. 

[cantab]

My 2 pence: KSP should have realistic physics and mechanics, that's what's made it such a great game. But it should not copy every little detail of real-life hardware.

Kerbin is too small, I do think it would be better if it was about twice the size, but I think the chance to fix that flew when it was left unchanged for the 1.0 release. We're stuck with it now, and the heavy parts and underpowered engines that go with it. Real scale makes launches take too long though, and I say that as someone who plays RSS. System scale should definitely be revisited for KSP2.

The basic Newtonian physics and the orbital mechanics (using the patched conic approximation) are fine. I also think Squad have done a nice job of the comms system, though I turn it off anyway because lots of vessels makes the game lag.

The structural mechanics is awful, it's ridiculous the extent rockets and other stuff can bend and flex. Easily the least realistic aspect of the game I think. Then again this does make it more visually obvious what's caused a failure, and makes for Youtube-friendly RUDs. I wouldn't mind seeing the joint stiffness tweaked though. A more significant overhaul can wait for KSP2.

The aerodynamics is still a mess. If you build something simple it's acceptable, but it's so easy to get tripped up by oddities or deliberately make a mockery of it. You can make two identical looking craft that fly very differently thanks to clipping and gizmo tricks. FAR's voxelisation approach makes things much more sensible and intuitive, stuff flies like it looks. Of course anyone who's flown in FAR will have encountered its own problems, high landing speeds and difficulty slowing down, but that comes about simply because the Kerbal parts are so overweight. Build a Cessna-sized plane and the Kerbal cockpit alone weighs more than an entire real-life Cessna. It can be corrected by fudging the exact lift and drag amounts.

The OP reaction wheels are an acceptable break from reality I think. Ditto the deep throttling unlimited-restarts engines. That said I think adding a few light-once fixed-throttle liquid fuel engines would be neat, they would "fill the gap" between solids and the existing liquid engines, make them powerful to compensate for the drawbacks. They could fit nicely into a "Moar Boosters" DLC alongside some bigger solids and other new engines.

I do think life support should be added to KSP. Not for realism as such, but rather to make time an important factor in your mission planning. Currently you can take as long as you like with Kerbals. Any stock life support should be kept straightforward, it really just needs to be enough that the player thinks about how long a mission will be. (Of course, those who don't want to can just disable life support).

[c4ooo]

I think more "easter eggs" and land marks should be added to make exploring Kerbin and other planets more interesting. Not sure what those would be, but maybe caves you could fly through or something would be awesome.

Also more realistic terrain like forests with dense vegetation that you could crash into would be awesome to see, although this seems technically challenging to implement, and not really relevant since this is a *space* game after all.

[Puddle Jumper]

1 hour ago, c4ooo said:

I think more "easter eggs" and land marks should be added to make exploring Kerbin and other planets more interesting. Not sure what those would be, but maybe caves you could fly through or something would be awesome.

Also more realistic terrain like forests with dense vegetation that you could crash into would be awesome to see, although this seems technically challenging to implement, and not really relevant since this is a *space* game after all.

I partially agree with you, but I think priority should lie with things related more to spaceflight like clouds. Also, not really related to realism, but every time you hit atmosphere at Kerbin the music cuts out. I really would like to have the option to let that happen when I hit the atmosphere of another body, like Duna or Laythe. Especially Laythe was weird to me, flying in an atmosphere with an airplane while having space music. Really got me out of the flying mood.

[VoidCosmos]

10 hours ago, c4ooo said:

I think more "easter eggs" and land marks should be added to make exploring Kerbin and other planets more interesting. Not sure what those would be, but maybe caves you could fly through or something would be awesome.

Also more realistic terrain like forests with dense vegetation that you could crash into would be awesome to see, although this seems technically challenging to implement, and not really relevant since this is a *space* game after all.

Easter eggs seriously? Nooo!!!!

[Gapone]

It's KERBAL space program. Kerbals do not exist in the Solar system. RSS is wrong. We don't need that.

[VoidCosmos]

1 hour ago, Gapone said:

It's KERBAL space program. Kerbals do not exist in the Solar system. RSS is wrong. We don't need that.

 

18 hours ago, Puddle Jumper said:

I partially agree with you, but I think priority should lie with things related more to spaceflight like clouds. Also, not really related to realism, but every time you hit atmosphere at Kerbin the music cuts out. I really would like to have the option to let that happen when I hit the atmosphere of another body, like Duna or Laythe. Especially Laythe was weird to me, flying in an atmosphere with an airplane while having space music. Really got me out of the flying mood.

RSS dosent change the kerbals

Stock visuals are the things that this game needs and also post processing.

Edited January 7, 2019 by Nigel Cardozo

[Gapone]

1 minute ago, Nigel Cardozo said:

RSS dosent change kerbal

W...hat? There are no kerbals on Earth!

[MaltYebisu]

I think it's essential to keep KSP a fun game where realism takes a step back to gameplay. I believe leaving a lot of realism to mods is a great way to get the best of both worlds. Never the less, there are still some things I think should become stock.

For me the most important one is persistent rotation during time-warp. I thought my game was bugged after I managed to reach orbit for the first time and the time-warp killed all rotation in the craft. It bothered me because time-warp is supposed to be a way to fast forward time to make the game playable, not something that exists in the game world. Because it is impossible to play the game without switching vessels or time-warping, it is not even something you can refrain from doing to avoid cheating.

I know changing this now would perhaps force a lot of people to change their play style a bit. But adding a kill all rotation button in the debug/cheat-menu would be a better solution in my opinion.

I'm not familiar with the technical reasons for killing all rotation during warp, but if the Persistent Rotation Mod can manage rotation with little to no performance issues, then implementing it in stock would not be too hard I imagine.

[mikegarrison]

Here's what should be changed for KSP 2.0:

Real scale solar system -- it's just too easy to get anywhere with only a single booster stage for launch. The rest of "realism overhaul" is a bit too much, but a real-scale solar system would give a lot more room for creative ship design.

Life support -- It's hard to do this without risking a mass kill-off of Kerbals, but life support requirements also add a lot of fun challenges to the game. However, the existing life support mods are all pretty broken due to having to work around the way KSP turns off pretty much everything for ships that are timewarped or out of focus. The whole timewarp/focus system would need to be reworked around life support.

Give Kerbals something to do in space. -- getting to Mun or Duna the first time is AMAZING, but wow, it only takes you a couple times before you start saying, OK, now what?

Fix the part count problem. -- The whole game only works because of the snap-together lego-style construction, but that really bogs everything down once you get big stations or bases or interplanetary motherships. Somehow the snap-together design and construction has to be retained but with entire stages turned into a single "part" once construction is finished.

[VoidCosmos]

Any other suggestions?

[Hotel26]

There is an 'opportunity cost' here.

I play Sandbox religiously.  I have just detected that KSP 1.6 has effectively changed the default Advanced Setting, "All SAS modes on all probes", to OFF.  I am not amused.  (Because my investment has failed.)

This is just pure Fetch-Quest to me, and completely deadening for creativity.

Compare the beauty, simplicity and elegance of Go with chess: in the latter, professionals spend two decades learning the "play book" for the opening gambit.  It's all nailed down.  Decades of rote learning (all now obsolete to machine learning).

The more onerous rules you lay on a domain, the more you constrain the degrees of freedom within the system...  KSP is probably renowned for how creative its community has been.  (Ask the authors how much they've been surprised by the kind of exploits KSP's users have made of it -- unexpected purposes  to which it has been put.  E.g. submarines...  None of it, 'realistic'!  But, O, so surprisingly and deliciously creative...)

Fetch Quest.  Cargo Cult.  "Make it look like NASA and it will fly."  Otherwise, no.  Fooey.

Edited January 13, 2019 by Hotel26

[VoidCosmos]

What about visuals and realistic physics

[Puddle Jumper]

3 hours ago, Nigel Cardozo said:

What about visuals and realistic physics

I think the physics are okay as they are now. The visuals could be better though. I especially miss cloud cover. Seeing cloud cover in Kerbal Space Program makes the planet feel much more like an actual planet. Ascending above the clouds makes the planet feel much bigger for some reason. I guess it has something to do with distance perception. I never use mods, but seeing Kerbin-videos on Youtube with cloud cover makes it so much more beautiful.

[Starman4308]

Would I like to see Squad create a commercial-grade, less buggy, robust version of Realism Overhaul? Yes.

Would it be in their best interests, or accessible to new players? No. It would be a huge amount of work to rebalance basically the entire game to match real-world scale. Even ignoring the many additional mechanics of Realism Overhaul (ullage, boiloff, avionics limits), a relatively small scale helps players get to space with suboptimal designs and imperfect gravity turns.

Overall, what I would like to see out of stock KSP: an optional 2-3x scale version of its solar system (relying on overweight parts to make up the remaining difficulty) for advanced players, bug fixes, and another pass at the aerodynamics... though the armor-plated tanks of stock are always going to mean aerodynamics are a bit wonky.

Also: launches in RO can take a very long time. My last career, I was working with an F1 -> 3x RD-58 -> RL-10 stack that took probably 15 minutes to get to orbit... partially because that second stage was hideously underpowered.

One minor issue the small scale of stock does impose, thinking about that, is that it's much less practical to loft a low-thrust second stage on a high suborbital trajectory to make up for low second-stage TWR; by the time you have a good loft in stock KSP, you're almost kinda in orbit.

[Tex]

I think the level of realism is pretty good as is. Devs have to walk a line between representing spaceflight accurately and making the game playable and game-y enough for the average new player to get into and enjoy. Realism is a whole different ballgame, and requires a certain amount of chutzpah.

[VoidCosmos]

What about real exoplanets

 

[MaltYebisu]

On 1/16/2019 at 9:44 PM, Puddle Jumper said:

I think the physics are okay as they are now. The visuals could be better though. I especially miss cloud cover. Seeing cloud cover in Kerbal Space Program makes the planet feel much more like an actual planet. Ascending above the clouds makes the planet feel much bigger for some reason. I guess it has something to do with distance perception. I never use mods, but seeing Kerbin-videos on Youtube with cloud cover makes it so much more beautiful.

It's beautiful, but you need a powerful computer to play with visual mods. (At least mine is not powerful enough)

If clouds become stock, they need to make it a setting you can turn on or off. I hope they do though. Maybe they will find a way to make it less demanding on the system.

[Tyko]

9 hours ago, Nigel Cardozo said:

What about real exoplanets

 

The problem with adding exoplanets is that they're light years away from your home system. Even at max time warp it would take a crazy amount of time for any Kerbal craft to transit to the other star. Some mods add another star orbiting Kerbol - like the Grannus expansion pack. I'm not sure if that's technically an exoplanet though since you're really dealing with a binary system then, not a different system 

 

[Puddle Jumper]

16 hours ago, MaltYebisu said:

It's beautiful, but you need a powerful computer to play with visual mods. (At least mine is not powerful enough)

If clouds become stock, they need to make it a setting you can turn on or off. I hope they do though. Maybe they will find a way to make it less demanding on the system.

Yeah, I guess it shouldn't be a problem to have different settings in that respect. Maybe one can have different settings making a trade-off between realism and graphics demands.

8 hours ago, Tyko said:

The problem with adding exoplanets is that they're light years away from your home system. Even at max time warp it would take a crazy amount of time for any Kerbal craft to transit to the other star. Some mods add another star orbiting Kerbol - like the Grannus expansion pack. I'm not sure if that's technically an exoplanet though since you're really dealing with a binary system then, not a different system 

 

A binary system should be a good idea as well. Even trinary or quaternary systems should be possible, although travel times in these systems would consequently rise as well. For even a binary system to work the orbital axis of the second star would be about five times the size of Eeloo's orbit.

[Nebbie]

Casual players want to see lots of explosions until they somehow manage orbit without once stopping to think. They actually might prefer the added challenge of realistic scale and more deadly reentry heating.

Unrealistic things that KSP really needs are Infinite engine restarts and deep throttling, because otherwise, it becomes very hard to stay in control.

As for other realism areas, what KSP does tends to be bad for gameplay because it creates confusion:

1. KSP's flawed aerodynamics modeling means body lift doesn't exist (although it's simulated on Mk2 parts only) and there are all kinds of stupid bugs with whether something is occluded from drag or not and whether a shroud on an engine actually reduces drag. Needless to say, a source of weird bugs and making imitating real planes not work well at all is bad for casual and serious players alike. FAR is a mod that fixes this, and it basically makes rockets and planes act the same 90% of the time, and the remaining 10% is mostly stuff you can do in it you can't do in stock.
2. The the lack of a damaged state (except on wheels) before exploding causes severe issues with landings in that legs will often pop and cause catastrophic failure, making things like SpaceX style landings and even Munar landings severely frustrating. Casual players certainly like explosions, but they also enjoy "WAIT JEB SURVIVED!", so I think this'd be a wash for them and just benefit more serious play.
3. Reaction wheels being way too strong (combined with deep throttling) means RCS is basically pointless. They don't need to be nerfed entirely to realistic levels, but I think they're ridiculous right now. Also, having them nonexistent on the MEM and spherical command pods is...it's okay, but completely inconsistent.
4. Inclination's a mess. You can go to the Mun and back without learning about it, which creates serious bad habits when you suddenly want to go to Minmus or Duna. This I think is actually the biggest reason most players never go past the Mun, you can't really go to Duna without dealing with inclination and it compounds on the existing issue of understanding transfer windows. I think a very mild inclination on the Mun would go a long way.

Now, of course, "there's a mod for that"...except there isn't. RO, the collection of mods that puts KSP very close to Orbiter for realism has just recently updated to 1.4, and it can't mimic the axial tilt Earth has because KSP doesn't support axial tilts. It also takes so long to update to begin with because it relies on a ton of mods that go deep into KSP's guts, because KSP was never really designed with things in mind such as:

1. Changing terrain textures
2. Fuel boiloff (this is really complex and is basically the thing that held up RO for 1.4)
3. Realistic solar system scale (the bigger system size means floating point error becomes significant, you can end up with Kerbals appearing to levitate on the surface of Mars)
4. Configuration of things like what fuel a tank uses (really stupid considering liquid fuel is a big deal in stock and there's tons of spaceplane parts where there's just a version that's the same but with different fuel)
5. Realistically-low-thrust ion engine burn times
6. Keeping spinning things spinning (timewarp on rails kills spin in stock, which is a serious issue when you don't have the reaction wheels to spin up rapidly and want to use spin stabilization)

Finally, I'd like to mention that KSP really is dropping the ball in regards to recreating the Space Shuttle and Falcon 9. For the Shuttle, the problem is engines not pointing toward CoM, so over the course of the flight, torque changes and gimbaling oscillations can develop. There simply aren't the good landing legs for a Falcon 9, meanwhile.

[steve_v]

4 hours ago, Nebbie said:

all kinds of stupid bugs with whether something is occluded from drag or not

I (and others) have been complaining about "cannot foo while stowed" since the day it landed, and it still causes me untold problems. Unfortunately FAR doesn't (and can't, IIRC) override this particular aggravation.
I really have no idea how anyone could fail to foresee the issues that would arise from abstracting cargo bay shielding as a spheroid zone projected from the centre of the part.

And the rest, 100% agreed.
 

 

[VoidCosmos]

Why do some of you like stock ksp only?

keeps on wondering.........................

[KSK]

9 hours ago, Nigel Cardozo said:

Why do some of you like stock ksp only?

keeps on wondering.........................

Some folks don't have any options other than stock due to their choice of gaming platform.

Speaking personally, back when I was playing KSP, I'd have limited play time. Hence I'd prefer to fire up the game and actually play it, rather than fiddling around with the eleventy-umpteen mods out there. Speaking personally again, I found the attitude amongst a certain section of the playerbase that everyone should be playing modded and/or that Squad should just focus on making the stock game a platform for mods, to be tedious, patronising, and presumptious on other people's goodwill and free time.

After that I can't speak for anyone else but I can speculate.

• Some folks just prefer to play the game, balanced as the devs intended.
• Some folks prefer the challenge of doing everything in stock.
• Some folks spend more time building random vehicles/dioramas/buildings/mechanical contrivances on Kerbin and don't find many (or any) mods to be that useful.
• Some folks find the performance penalty of running mods to be unacceptably high.

I'm sure there are other reasons. Incidentally, if you want a less grumpy answer next time, you might try framing your post as an honest question rather than tacking on a passive-aggressive snark in small font.

Edited January 25, 2019 by KSK