Wobbly Rockets - KSP 2 Dev Chats

[@KerbalSpaceProgramOfficial]

Additional Note from Creative Director Nate Simpson: I hope you enjoyed this chat! Since this conversation took place, David has been developing a tool that allows our team to compare multiple wobbly rocket remedies, including selective wobbliness for certain part categories, KSP1-style autostrut for the entire vehicle, and various flavors of packed vehicle physics. We are testing these now, with the goal of achieving a near-term improvement in vehicle rigidity while developing a more ambitious long-term fix that's performant at all scales. We'll share more information when we've arrived at a balanced solution. We know you've waited a long time for a solution to this issue, and we're excited to be closing in on a resolution.

[@KsNewSpace]

Whatever you do please give us options! Like switch to legacy physics.

[@pebegou]

Do a single beam along the craft with plastic deformation and smoothly bend the whole rocket along the beam. This would be realistic and would not have gaps in between parts when wobbling. Each part could have a beam strenght multiplier so you can make sturdier rockets.

[@pebegou]

Please just dont do the minimum, we deserve cool mechanics and people love simulatios.

[@BladedFish]

dead game

[@Kerballistic]

@KerbalSpaceProgramOfficial Is it possible to implement the legacy KSP1 autostut system into KSP2 even just temporarily while a modern autostrut is worked on?

[@Zeusbeer]

The joke is that irl fuel tanks have significantly lower dry mass compared to kerbal space program, meaning they have thinner / weaker structures for the same amount of fuel. And in real life rockets don't really whobble, they just desintegrate when the load is slightly too high.

[@radgrinarcher8184]

Perfect! Any time the rocket *would* wobble, we can just have it explode instead. No more wobbly rockets!

[@9bang88]

RP-0 does a very good job of portraying this

[@joshss223]

Hey, remember that time the Redstone, Atlas, Saturn I, Saturn II, Saturn V, Space Shuttle Main Engine, STS Boosters, Falcon, Flacon 9, Flacon Heavy, Starship, and Blue Origin rockets were wet noodles and blew themselves apart every launch? For fucks sake it took less time to dev some of the real life functional rockets than this shit game has had.

[@ElseeKay]

simulation is at the end of the day, billions and billions of simultaneous math calculations rather than the 'simple' task of physically throwing rocks at the moon. that's why they're ["rock"ets] :^ )

[@TechyBen]

@@joshss223 Starship did like 5 flips and failed to explode when they hit the explode button.

[@tomeshec1122]

If starship didn´t significantly flex when it was tumbling in the upper atmosphere then no rocket should flex at the slightest touch of steering in ksp 2

[@TysonSukeforth]

Just because Starship can, doesn't mean every rocket should. It only means that large rockets should have the potential to be rigid, if properly constructed.

[@KsNewSpace]

Starhsip is a well designed rocket. A long noodle is not so it should flex. Having Kerbals space walk to weld struts to space stations is also essential part of gameplay.

[@haassteambraker9959]

​@@KsNewSpacepeople aren't upset about noodles wobbling. People are upset about replicas and other generally reasonable designs wobbling.

[@tonyHern865]

@@TysonSukeforth falcon 9 rockets do not bend either

[@tonyHern865]

@@TysonSukeforth Space shuttle boosters didnt bend either...

[@sulijoo]

The message of this interview seemed to be "expect a lot of wobbly rockets because it's too hard to fix". I remember playing KSP long before autostrut, and it was a nightmare. Ten years later I still play KSP because of how important autostrut was to the game. It made the game so reliable that I can pretty much make anything now. I never have to worry about wobbly rockets. If you don't strive to do the same for KSP2, given how much more ambitious the game is, then you have failed. Interstellar, bases, won't mean jack if you don't actually deal with this basic issue early on.

[@KNR90]

They also fixed 98% of the wobbly joints

[@wck]

I have over 1000 hrs in KSP and never used autostrut. Maybe it wasn’t as needed as you say

[@Jmstraubgamer]

@@wck Ive never used it either. I think what the devs are trying to say here is that there are several different choices they could make and they wanted to shed more light on what solutions could be implimented but they want to make the right choice rather than doing it twice.

[@emp1u]

​@@wck Depends on what you build. I also have 1k+ hours and had used autostruts A LOT.

[@HitBoxMaster]

@@wck You're a liar

[@kaidanodom]

This is the first time I have ever left a KZbin comment and it will probably be the last. The reason I feel as though I need to say something is because KSP1 is part of the reason I am who I am today. I have been playing KSP1 since I was 11 years old. It was my first introduction to orbital physics as well as design reiteration. My love for this game throughout my teen years is what led me to study engineering in college. Fast forward to today, I am a senior mechanical engineering student. I would like to point out this inherent focus on “when you think about how X should behave… Y should occur because Z” because I found this very relatable. I often catch myself trying to recreate simulations in my head for how I think a system should behave. The reality is it’s very hard to recreate complex physics and fluid mechanic scenarios in your brain and you’re better off just running some numbers for estimations and then trying it out to see the results for yourself. Especially when you don’t have to actually damage anything in real life to test it out! I get where they’re coming from, but this feels like a lot of intellectual masturbation (which engineers are notorious for doing. I listen to it all day every day). At the end of the day, KSP is supposed to bridge the gap between hard science and casual gaming fun. Nobody really cares about the absolute real world behavior. Being 90% accurate to real physics and being playable is way better than being 99% accurate and a wobbly mess. I think the team may be neglecting the diminishing returns they may run into when trying to make each joint attachment behave exactly as they believe it should. Run some tests, get some intuitive rigidity implemented, and get this game on its feet!! You can always go back and fine tune after the game is playable.

[@skywatcher2025]

I don't think they're at the point of diminishing returns, yet. The solutions they spoke of seem to be generalized enough where I don't think we have to worry, but still dynamic enough to where there should still be decent realism.

[@kaidanodom]

@@skywatcher2025 I agree, I don't think they're at that point either. I'm just saying I think they will run into that if they are trying to get a perfectly realistic balance between rigidity and deflection. meanwhile the game sits in an unplayable state for months on end.

[@edust83]

It reminds me of how Kingdom Come sword fight system was delivered. Yes, I wished more real mechanism than Skyrim's or Assassin's Creed combat style but it turned out to be more like a sword simulator and that's something wasn't really necessary in an rpg.

[@familiarhuman]

very well put. i've worked on games with engineers in the past. they can have a very nasty habit of forgetting that games are supposed to be fun and playable. it doesnt matter if you have 100% physics accuracy. I too fell into this trap trying to create a 1:1 scale earth in unreal engine and making my own kerbal clone. After a few months, i realized that having a 1:1 scale earth wouldnt make the game more fun and I was taking up way too much of my time. This happens a lot in game design and needs to be talked about more often.

[@WeirdWizardDave]

No more wobbly rockets. Fuse together most parts into a single rigid body and dial up the strength and number of joint points for everything else. If you really want to keep the "fun factor" of wobbly rockets early on make overcoming it part of the science system which could unlock merging parts for example. I don't understand how this hasn't yet been decided, designed and built yet. Its been 7 months since start of early access.

[@barefootalien]

Yeah, that blows my mind, too. How did the first alpha build get play-tested in-house, and the entire team went, "Yeah, that's how it should feel"?? This definitely feels like foundational decision sort of stuff to me. I was immediately disappointed at EA-launch that really basic issues weren't solved, because the physics engine functioning as expected, and orbits being stable and reliable in a game that's basically _about orbital mechanics_ seems like stuff that needs to be baked in really deeply, not something that's "eh, we'll figure it out later and tack on some fixes".

[@davescott7680]

awesome I'm building a rocket that looks like the flying spaghetti Monster and have it fly like a rock solid block. The parts should flex, but what KSP2 released in was just dumb as they just went with nothing.

[@doltBmB]

so... struts

[@jonmedders3866]

Maybe its just me, but these types of fundamental discussions about how the foundation of the game should function... should those not have happened years ago at the beginning of development?

[@SpaceNavy90]

This is all one big pre-alpha test

[@kmsmith8934]

It's not just you. But again, it was said wobbly rockets are part of the game... Why are they even looking at it if that's the case. Sadly it's clear they aren't sure what's going on.

[@TheBigLou13]

Wobbly rockets are currently a part of the game but not intended/desired. It's probably a result of the physics calculating connected parts as lines connected with dots - and the models (around the lines) are free to intersect the other models. Around those dots the lines pivot/bend through forces, and bend back by the original/intended/built alignments of the parts in relation to each other. Between those two forces the parts oscillate back and forward. Through timestep rasterization you also have a non-perfectly-round oscillation, leading to accumulating shifts/offsets towards one of the two forces/directions. Maybe this wobble can be mitigated by having parts not interact as lines (connected with dots) but rather have them connected at multiple dots across the touching planes - but this can lead to problems/inconsistencies with rounding errors and probably cause the kraken to strike or parts to spin faster and faster on their own. Somehow all forces have to be integrated independently of time - while moving and experiencing changes through time.

[@OzoneGrif]

@@dhwardani No, it's KSP 2. The two means they should have the experience from the first opus. They don't because it's not the same team of developers and they have no clue how to solve the issues KSP 1 already solved.

[@neotuxxedo]

@@OzoneGrifthe game studio forced them to put it out. They warned us before hand that it was going to be bad, I think people underestimated how bad it would be.

[@lockscad]

Do they not realize that the vast majority of people who bought this game would rather "suffer" from single rigid body rockets while you figure out this wobbly stuff in the meantime right? You're all acting like having wobbly rockets is more desirable than 100% rigid rockets.... its really not a hard vote.... just make them flying bricks for now, at least they will fly!

[@papawushutv]

"at least they will fly"??? what kind of hot garbage are you tossing together that that is a challenge for you?

[@Schnittlauch02]

I absolutely agree. For me personally it was never fun, in whatever game I played, to fight against the rigid body simulation. I think there is no shame in simulating a flying brick, but them make certain parts visually wobble (but not physically).

[@critterjet100]

100% agree man. I hope the devs will read this. I know the devs have far more experience in game development but please don't make the same mistakes as other big game developers. Listen to the players. We don't want wobbly rockets. We don't want rockets to collapse for what would be structurally stable in the real world. Matt lowne has a good video on it. Watch it. Please. Love from a long time fan.

[@ShadowDev6969]

well if you build dumb stuff, ofc it wont fly

[@critterjet100]

@@ShadowDev6969 of course but the problem is that very sensible craft are falling apart at every seam.

[@DarkPhoenixDack125]

So when i build a scale starship and it wobbles apart but SpaceX can have real life rigidity, that a realistic physics experience?

[@OriaXu]

This is ridiculous. These problems were SOLVED years ago in KSP, but somehow for KSP2 its impossible to solve? We didn't expect a bug free game, but we did expect to at least not have the SAME problems we had to endure for years already! I love KSP, and somethings are good in KSP2, but it has been a HUGE disappointment so far. I have thousands of hours in KSP, but only 10 in KSP2.. and the first thing i did after playing KSP2 for the first time was reinstalling KSP... i guess that says it all.

[@platyhelminthes2877]

It's a real shame. But this is what I've come to expect from the video game/movie industry these days. Corporate BS and greed ruins everything. I don't know if this game ever stood a chance. I just don't see it succeeding, and that's frustrating. At this point I would much rather just have a remake of KSP 1 than wait for this game to maybe eventually get made. All I want is KSP 1 redone with better performance at higher part counts, maybe new IVA cockpits, steam workshop support...oh well.

[@Frank_SB]

@@platyhelminthes2877 stop whining

[@platyhelminthes2877]

@@Frank_SB you stop whining! 🤣

[@monkeywrench677]

You guys didn't have this already figured out beforehand?

[@Zwettekop]

I didn't find an answer to the following question: "Why can't wobble be fixed by multiplying the stiffness factor by 10x". I was really interested in hearing why that wouldn't work.

[@maxv9464]

It doesn't work even now. The stiffness factor is not the only thing playing a role. I've used the modifier, there is still an underlying issue that still causes significant wobble in a lot of cases.

[@casualbird7671]

@@maxv9464 Are you sure? I've heard lots of other players are having great success just by going into the files and changing the joint stiffness number

[@WillowLiv]

They talked about making everything super stiff, into a single rigid body, the problem they present is that why that would work wonders for the core of the rockets, that you would expect not to be stiff at all, it would reflect very poorly on radially attacthed parts, hence the nature of the problem. You can't simply make everying rigid or you lose pretty much any physics. A more nuanced solution must be found, and maybe, a short term solution, like autostrut, or some other quick solution.

[@Zwettekop]

@@WillowLiv No that's just not using joints at all, and modeling it as one rigid body. I just meant using the existing joint system but multiplying the stiffness parameter.

[@KNR90]

That's not the major problem ksp1 fixed it. But it's not the same team and they would rather sabotage the game saying it's too difficult (when they literally stimulate orbital rockets on the fly) because figuring out how the previous team solved it is too hard

[@TheDutchPhysicist]

Just do single rigid body, collisions only happen when you fail.. You will stop wobbly rockets and get a huge performance increase. Also in that case you could check for collisions on the boundary of the rigid body and in case there is a collision turn on multiple bodies and check if the collision breaks the single body into multiple bodies.

[@kerbalsomething2673]

exactly what I have been thinking. I was like: "Why haven't they done this already?"

[@critterjet100]

It may not be that simple. Unless you professionally code, that might not be a real solution . But I do agree that a rigid body would be better in the mean time than the situation it is now.

[@thorbjrnaaen3426]

they said in the video that that solution wouldn't work without a complete rework to how wheels funtion, so if that is worthwhile then i might happen, but they would have to completely throw out and remake how wheels funtion, which would take a lot of development time, a lot longer then some other solutions.

[@OzoneGrif]

​@@thorbjrnaaen3426 Oh damn, they will have to start over from scratch and sell it under the name KSP 2. If only it was an option!

[@ICKY427]

then you could just build ridiculous stuff that doesnt make sense. theres no challenge.

[@therealduckshow]

I feel like I've been so charitable and patient to this team over the years. I paid full-price on day one just to support the development as best as I could. As a fellow game dev, I know how hard this job can be, how complex a game like this is, and how long things take... But even my patience has completely dried up, and it all just makes me so sad and frustrated! Game development takes a long time, but I really can not compute how so little progress seems to have been made! A handful of bug patches between February and *October*?! If you at least kept the community well-informed on the progress being made, then... I suppose that might be comforting... Oh, what's that? *Four* videos posted to the youtube channel since the release of early access?? At this point, I don't even. Sigh, I hope you do better. I still really want this game to succeed, but god-damn you're making it hard 😢

[@Razgriz__1]

Seems obvious to me that the publisher forced them to release well before the game was ready, and not only that but gave them very short notice. I suspect that first early access announcement we got around a year ago was probably not long after they were first informed of it themselves.

[@higueraft571]

"Four videos" and all the other posts. Videos are not the only means of communication, they arent even the most thorough in this instance :V If that's your ONLY source of info, no wonder you think there's almost nothing...

[@therealduckshow]

It's the job of a social media manager to make sure information from a company reaches the community - if information doesn't reach me, that's on them. Of course, they can't be expected to reach *every* single customer, but it's not like KZbin is a niche platform.

[@Vindicator66]

Smoke! Nate Simpson is Madoff. You want to come and sell us the idea that just realized the problem of wobbly rockets and so you put the product on sale?

[@Hossdelux]

This is literally the reason I don't own KSP2. Just make it a solid body in which physics is applied to the whole lot. Make it so the animation when it fails it comes apart. You are holding onto something that is holding you back and lowering player enjoyment.

[@LenKin_0W0]

Yeah and have zero performance

[@GreyBeardGamingYT]

Over half a year has passed since release and we're talking about wobbly rockets? You had a roadmap and we're stuck on number one for over a half a year. I look forward to the day a post from you has anything to do with a content patch.

[@Jmstraubgamer]

Valheim had the same dev path. It was even delayed longer than six months and the major updates got pushed back significantly. This happens to games all the time as other things take priority so resources are shifted around. The KSP2 community seems to conveniently forget this.

[@GreyBeardGamingYT]

@@Jmstraubgamer I disagree with you. Valheim was feature rich at launch. I played with a group of friends in multiplayer exploring multiple biomes building bases and defeating bosses. It's not like ksp2 at all where at launch it's missing huge parts of its core gameplay.

[@adsfsd1]

Imagine how many hours have been wasted internally having meetings about this issue instead of just fixing the problem

[@higueraft571]

Meanwhile in another comment section: A bunch of people discuss how it's "not that simple" and there's a bunch of possible solutions to take and issues with those. "Just draw the fucking owl" and all that...

[@TheGreenTaco999]

​@@higueraft571 it's called "paralysis by analysis", this is not "Just draw the fucking owl" at all

[@Kepe]

It's amazing how broken the game was when launched into early access, and how broken it still is. The fact that the devs are still trying to figure out how to make the rockets rigid is quite telling. That's one of the few things which were supposed to be fixed from KSP 1, but they didn't even manage to do that even though the early access launch was delayed by three years. That's three years of extra development time, during which even the basic physics stuff wasn't fixed. There's no point in wasting development time on extra content if the core functionality of the game is broken. Talk about polishing a turd. I'm sad and angry that I bought this game. Maybe it's playable in ~2030, years after the devs have abandoned it and actually competent people have modded the heck out of it. Free hint: a completely rigid rocket is a lot more realistic than one which wobbles around like it's made of boiled spaghetti.

[@higueraft571]

Auto-Strut is not a long-term solution.

[@jedtech8942]

Phase 1: One rigid body! Save CPU and fix wobble. Phase 2: add wobbly nuance if it still feels important to the user base.

[@Seeraphyn]

You know, rigid bodies don't really exist in real life. You could try to simulate every interaction between atoms in the cristalline structure of the metal and get a frame per year. Or give players what they want, merge every unitary structure into a single rigid body or just strenghten every joint A LOT as a temporary fix. Also please feed Nate :)

[@Bulltoad_Real]

Why don’t you make rockets completely rigid, but have a sort of system where at each join, there is a stress level, which when exceeded, causes the rocket to wobble, though this should only be reached if the rocket has done heaps of somersaults and is barreling toward the ground at a very high speed.

[@HiddenPrior]

This. There are many simple bridge building games for example that do this for rigid joints, and that makes so much more sense to me. I think this game was just drastically underfunded, there are not enough devs and not enough quality work was done on improving the engine or designing these core systems. Not the devs fault, just the producers.

[@WolfyBoi7521]

Game was released in early access 7 months ago and you're still brainstorming over basic stuff 💀

[@jblakemore]

It's almost like they never intended to release a functional game😂

[@WolfyBoi7521]

@@jblakemorebefore release they pretty much said that the game was almost complete, that they had slayed the kraken and that it was a matter of optimization and making it modable. I'm starting to think they had absolutely nothing done and they were very aware that they would be abusing the "early access tag" to mislead buyers into paying $50 for a not even pre alpha tech demo.

[@jblakemore]

@@WolfyBoi7521 If they knew and did abuse the EA tag somebody needs to sue or this will become the standard for all gaming companies.

[@Xenotrickster]

@@jblakemore Anyone that paid full price for a 0.0.1 build, knew exactly what they were getting into. Don't hate the game, hate the player. (I think that's how the saying goes)

[@jblakemore]

@@Xenotrickster Yeah nah that would only apply if it wasn't scum like EA

[@Diggelaer]

what about no wobble? you know for the people who actually play this game longer than 10 minutes

[@sterlinganon]

I can't believe they started touting/teasing this nothingburger of a video a month ago.

[@MattLowne]

Wobbly rockets can be a real challenge in KSP2. Glad the team is working hard to tackle this issue

[@ecogreen123]

excited to see what you can do without wobbly rockets.

[@thebrick.1385]

Solid rocket booster

[@papawushutv]

yeah no they're not

[@danieldc8841]

Say what you’re really thinking

[@CoreyKearney]

That was diplomatic considering your interview with harvester is the evidence that it's a solvable problem. Especially when your not asking the joint system to something it was never intended to do.

[@Tech_Enthusiast_001]

So,... is there a solution in the near future? Next year? How about the promised content? Anything apart from big words?

[@ivmatt3142]

Thought: Have two types of joints within the game. Face joints (fj) and point joints (pj). Face joints would be used to connect stacks. (i.e. fuel tank-fuel tank, fuel tank-engine, etc) To do this, you could have nodes that are specifically face joint nodes. When two fj nodes attach, they create a face joint. Which will have little to no flex. As for point joints, they would be used to attack things to the side of stacks. (i.e. decoupler- fuel tank(side), trusses-fuel tank(side)). Parts like this would have point joint nodes. If pj nodes are attached to the side of another object, they would have significant flex, requiring the need to use struts. Also when a pj node is attached to a fj node, the joint would still be a point joint. The pj nodes could be scaleable depending on sizes of parent parts, however I do not know how that would impact performance of the game. Lastly, there are plenty of edge cases, however this could be an idea for solving the problem.

[@a.f.r.animations8873]

that's smart

[@aviFlashbacks]

Another thing to note, you don't usually attach things directly to fairings besides decouplers, which are meant to be very slightly wobbly (I think) as they are meant to be loose to make it so the stages can easily separate safely (again, I think, I'm not certain). In ksp1 for example, you just put things in the fairing, and thats why theres so much wobble with fairings, you have to actually attach the things to the fairing with struts. That, for example, might be a place where on one side of the fairing part you might want a "face joint" unless you put something in it of the same size, making it so the fairing does in fact have to go around instead of straight up. I feel like the structure of the last sentence sort of fell apart, oh well.

[@KNR90]

Ksp1 fixed this without the complexity. The team is completely new, they made the same mistakes causing the same problem. Previous team fixed it. This team says it's too hard to learn from the previous fix so deal with a broken game. They can't be bothered

[@captainahab5522]

@@aviFlashbacksI think making all parts inside the faring be a rigid body would make sense Payloads shouldn’t clip through the faring at all. But if needed the fairing- rocket joint could be made more flexible so that it’s more realistic with aerodynamic forces

[@CosmicAggressor]

Damn. I never considered that someone might actually want any part of a rocket to wobble. Nah, I still want as close to 0 flex as can be arranged. Plus there are those time warp issues when using infinite fuel. I don't know how it works without infinite fuel but I don't want my constant boost ships to need to be used without timewarp.

[@Fell-x27]

7 months passed. They are talking about wobbly rockets...

[@PaddyPatrone]

So far "killing the kraken" didn't work at all.

[@kempoficht]

If the kraken was the fanbase. Its killed

[@sekelus]

in conclusion: this is still a pre-alpha game

[@bobsaget8791]

when you're having basic game design months after the game was launched into early access, years after it was supposed to come out...

[@Jmstraubgamer]

Well considering EA is ment for games in an alpha or beta state that makes sense, or did you forget to read the Steamworks document you said you did when you clicked the buy button?

[@jjazy3439]

​@Jmstraubgamer it was supposed to be fully released 3 years before it came out in "early access" and so far 7 months later they have added nothing

[@tonyHern865]

you can already have ksp 3 today if you buy ksp 1 and you installed all the best mods available in the workshop.

[@mikond]

KSP is another great example of a great indie game destroyed because of a big comapany that doesnt care about the players at all.

[@T_Mo271]

Whoever is paying the bills for the KSP2 dev team should have pulled the plug on the project two years ago. KSP2 now is worse than having no KSP2 at all. It's a case study in not understanding your audience/market nor how to plan a product development. The KSP2 team lives in a world where original KSP need never have existed, because they've ignored everything it ever accomplished.

[@pancytryna9378]

For real, delete KSP2 and hopefully one day we get KSP3

[@Robert72z28]

You guys really should just give players a temp fix even if thats rigid bodies and unrealistic so people can actually PLAY THE GAME. It doesnt have to be your end game vision. Temp fix the game then work on your vision. Anything else is just your pride getting in the way.

[@neotuxxedo]

I just made a comment like this, seems that the rigid body fix is in high demand

[@goku445]

You can fix it in a config file.

[@ViewBothSides]

LOL this sounded to me like a classic example of teleworking not working. Nate: where can we turn this wobble crap off so it's not an issue? Trigger: I love all this complex wobble stuff and we have so many choices to keep wobble in lots of places. Maybe spaghetti is an option, maybe a rigid block of concrete. Maybe some people do want the 2nd stage wobbling compared to the 1st and the 3rd, maybe others do want the boosters jiggling about. Let's look at those options for the next decade.

[@CoreyKearney]

Harvester figured it out for his new title. The problem is you are abusing the joint mechanics in unity. The bigger problem is you are doubling down on it because a quick temporary fix was just to increase the joint strength in settings. Not even a code change FFS.

[@fabianmuhlberger6153]

Can you explain ? I am not familiar with unity, From what I understood it is an engine problem that is not easy to solve due to other mechanisms (wheels) that rely on that system too.

[@CoreyKearney]

@@fabianmuhlberger6153 No I can't, but Harvester did in his interview with Matt Lowne when talking about Kit Bash.

[@0koppie0]

@@fabianmuhlberger6153 From what I understand, Kit Bash makes your crafts into a single rigid body and calculates the physics on that one piece, breaking it in parts when it is supposed to, rather than having to simulate each piece independently.

[@KsNewSpace]

@@CoreyKearney It doesn't make much sense to repeat hearsay you don't understand. Kitbash is a RC plane simulator. They don't have to simulate stuff like fuel flow between tanks, reaction wheels, heat flow and many other things that make rockets and spacecraft unique. It's just not comparable.

[@TlalocTemporal]

​@@KsNewSpace-- None of those things have anything to do with body physics, except _possibly_ fuel mass shifts. A better argument would be the differing scales of handheld vehicles weighing single-digit Kgs vs interstellar colony craft and colony stations weighing ten-digit Kgs. Even then, five-digit craft should not be flailing like pool noodles, especially when nearly all games which consider themselves simulators (including most other rocket simulators) use full rigid-body physics. Not even the Ever Given (weighing some 400,000 tonnes, 80x that of Starship) crashing to a stop agaist the bank of the Suez Canal even brought a thought of vehicle flexibility. After being disloged, it completed it's journey under its own power. No simple stack rocket should be even noticably flexing without doing something stupid, let alone -folding in half on the launchpad._

[@Simon-px8mi]

How AAA ruined a passionate company.

[@TotesRandom]

Releasing a video now, trying to explain a basic game engine problem that should've been something you guys looked into years ago when you were designing the core functions of the games mechanics, seems like you're going for the sympathy vote. Problem is, you're into early access, and this comes across as you pointing out that, after YEARS of development, you have literally no idea how to fix this problem. The fact Nate previously stated wobbly rockets were part of KSPs "charm" shows they either knew it was unfixable way back when and tried to get away with ignoring it, or he was SO out of touch that he didn't know literally EVERYbody hated it until EA launch. Either way, this video makes them look incompetent and like they are making excuses for why we can't expect them to ever properly fix it.

[@Kuhiria.]

Guys, please. I would LOVE to support the development of the game if it was cheaper. It is impossible to find a good reason to pay 50 USD for a game that is not even halfway there...

[@barefootalien]

The way I think of it, the entire planet has had something like 100% inflation in the past five years, right? AAA games have lagged behind a fair bit; I've been surprised for a very long time that a new AAA game in my youth in the 80's and 90's cost $60, and a new AAA game in 2018 still cost $60, while in the same time period, a loaf of bread went from $0.60 to $1.60 and a gallon of gas went from $1 to $4. Now that the loaf of bread is more like $3 and the gas around $5 (artificially lowered in the US compared to most places), AAA games are starting to come up in price accordingly, with new games costing $80 to $120 instead of $50 to $70. It thus stands to reason that a AA game that five years ago would've cost $40 to $50, with maybe $30 in EA, would now be $60-$80 with $50 EA being pretty reasonable. The problem I have isn't the price, which seems fine. It's that it was just released into EA _way_ too early, following the trend of such hasty decisions.

[@Kuhiria.]

@@barefootalien Yes of course, it was released too early. They had two options: release it later when the game was more polished for the full price or release it earlier but decreasing the price.

[@barefootalien]

@@Kuhiria. Right, but what I'm saying is that $50 could easily _be_ a reduced price for this game. It's under the same financial umbrella as Rockstar now, after all.

[@comwarrior69]

There is a simple solution.... All attach nodes are 99% ridged. They don't flex but they do suffer from shearing and twisting forces. Each node type has its own maximum tolerances which when violated result in the node breaking and dissasembly occurs. Surface attach is 100% flexable and can cause flappy birds. So, surface attaching wings can result in wings bending or being ripped off in high G turns. Which is what you'd expect. Now, if you were thinking that solid nodes means you could have a part with nodes radially to attach your boosters and it would be ridged, yes, but the shearing forch caused by the engines would still apply to the side attach node which if too high would result in seeing wild boosters in thair natural habitat...

[@Vindicator66]

Nate Simpson is managing or is he simply a puppet to face the mediocre product?

[@unownyoutuber9049]

why was this not talked about 6 years ago at the start of development? Instead of 7 months post launch.

[@Jmstraubgamer]

this game has not launched.

[@unownyoutuber9049]

@@Jmstraubgamer so what did i pay 70 aud for then? Thin air?

[@chadleworthy1741]

As an Aussie all i hear was " yeah nah im not gonna stand on the shoulders of giants, im gonna milk this paycheck for all its worth" but what do i know with 4,000 hours KSP on console.

[@milkywegian]

Love how he wanted to "rebuild" the game engine "from ground up" and its just a copy pasted code from ksp1 with new bugs

[@higueraft571]

​@@milkywegianYou looked at the code?

[@milkywegian]

@@higueraft571 i didn't, others with thousands of more hours in the game than me did

[@higueraft571]

@@milkywegian in 1 or 2? Doubt if the latter.

[@milkywegian]

@@higueraft571 latter

[@anoniemw.222]

The game already calcates joint breaking when forces get to high. Why not remove wobble and have that instead. Or wobble only getting activated alfter a amount of stress is placed on a joint

[@N9GamingOfficial]

You can increase the rigidity number in the game's .ini files to decrease wobbliness to your liking. People have been doing it since day 1, and I have a feeling we'll have to keep doing it :P

[@goku445]

Does it increase performance as well?

[@CoreyKearney]

@@goku445 It lowers the part count so yes?

[@JulesVN]

@@CoreyKearneyIt doesn't reduce part count, it just makes the joints stronger. So to my knowledge it also doesn't increase performance.

[@TlalocTemporal]

​@@JulesVN-- Not by itself, no, but reducing the need for reinforcement therefore reduces the number of struts and structural parts, which does increase performance, as well as avoiding symmetry bugs with struts.

[@JulesVN]

@@TlalocTemporal yeah, that is true ig

[@Schwing27]

7 months in and there’s not even reentry effects yet… this is seriously getting absolutely ridiculous. Edit. I have 127hrs in this game now. I do enjoy it. I’m not just simply being a hater. I WANT this game to be amazing and it absolutely has the potential to do so. I’m just extremely annoyed and concerned how slow content is coming.

[@thorbjrnaaen3426]

they confirmed in a previous Q&A that reentry effects are coming with the science update, that is why they have been delayed so much, because they need to wait for science to be completely finished, which is taking a LOT longer because of the amount of bugs that needs fixing which takes time away from developing the science update.

[@davidmartinek5257]

They pretty much started working on it when they were originally supposed to release it. Realistically, it will take them one year to sort of polish the current "early access" build, and another year to finish the rest of the game. Just forget about this and come back once it's ready.

[@Orangnus]

I'm really disappointed in the devs. It honestly feels like a betrayal since we were expecting KSP 1 but with more things and they never once before early access told us that they couldn't get that on release.

[@kvasir121]

It's so over lol. Funding will be pulled before long. Colonies and interstellar travel will probably be reserved for mods if we ever get them in even that capacity.

[@Sloppyfish2015]

You know in uk law you have to give a refund if this is not fixed in 28 days ?

[@JJBeaner]

short term auto-strut would fix half the complaints on the subject. Short term fixes will add to sustainability. I think too many people are spoiled by KSP 1 and forget how long it took to get to the point it is. I want KSP 2 to surpass that, I think we all do. . none of us will be happy with the time frame it takes to get there because we want it now.

[@HitBoxMaster]

EA KSP1 was leagues ahead of EA KSP2, and we aren't spoiled, bootlicker. People are demanding what they were promised and paid AAA money for.

[@milkywegian]

@JJBeaner ksp2 is a AAA game, are you too slow?

[@jjazy3439]

​@@milkywegianit's not even playable

[@milkywegian]

@@jjazy3439 ikr

[@JJBeaner]

not trying to be a bootlicker. . I am hugely disappointed with the release myself. Sorry for wanting it to succeed. . sheesh.

[@navyrayne]

With all respect, the KSP 1 team figured this out years ago.

[@CoreyKearney]

They hacked a solution. The rest of the game was light enough they got away with it.

[@viniciuskobs935]

​@@CoreyKearneyjust like every ksp 2 feature

[@CoreyKearney]

@@viniciuskobs935 Reading comprehension not your strong suit eh?

[@LineOfThy]

@@viniciuskobs935 that's not an argument

[@basila33]

no respect for these liars. just remember how Nate talked about almost-ready multiplayer mode THREE years ago.

[@phoxpharms]

As a player, I simply expect a game to work as expected when sold. You wouldn't be happy buying a car, to find that the engine is still being worked on, so you can buy the car, and it looks like a car, even rolls like a car, just doesn't function as a car. When you sell the public a less then half baked imitation of a game, you should maybe charge less then half the final game price. KSP is 2xs the cost for 1/10th the experience. I keep holding out buying KSP2 for a game in a playable state, but I'm loosing hope and very close to giving up on the idea, and going back to modded KSP1 which functions as KSP2 claimed it would. Tired of gaming companies thinking Early Access is an excuse to sell people garbage.

[@Jmstraubgamer]

Its not an excuse to sell people garbage, but it is a tool that is used for developement. "Steam Early Access enables you to sell your game on Steam while it is still being developed, and provide context to customers that a product should be considered "unfinished." Early Access is a place for games that are in a playable alpha or beta state".

[@phoxpharms]

Yes I understand. I have a dozen, maybe 2 dozen games in Early Access. They are amazing games, that are usually $20 because they are not by any means a full/complete game. Hence the problem with KSP2. They sold us a barely working game at full game price. I would have 0 complaints if they had charged $20 during Early Access, and then jumped to their $50 price tag upon Release as ANY respectable company would do. Like I said, you wouldn't be happy if you bought the brand new Xbox, paid $300 for it to be "Early Access" and you can turn it on, you can see the Xbox main screen, there are a couple of aps, but its not done yet, the ability to play games is coming in an update that might be next week, or next year. Only with video games do we happily allow companies to sell us unfinished products and charge full product prices. Thanks for helping to prove that point.@@Jmstraubgamer

[@HiddenWindshield]

Here's a question: _"Why have joints at all?"_ You've demonstrated that you can do procedurally generated wings, why not have a procedurally generated rocket body? Let the player put fuel tanks and equipment in this body as a separate step. It would allow a lot more flexibility and creativity in rocket design, and, oh by the way, would also fix the wobbly wocket problem in a simple, performant way.

[@Blarnix]

Guys, you do not need to simulate every damned atom for this game. You just need simple rigid body sims and estimations on load bearings, center of mass, some material intersections, and heat.

[@ottosturm7224]

I think the example of the truck and the truss is what the rockets should preform like. where the structural bits are (mostly) rigid and the connectors are softer

[@Masked12-b5i]

Crazy how this game has been out since February and it's like nothing has been added at all... starting to lose some serious hope of this game ever being completed if it keeps on going like this

[@TheDutchPhysicist]

Please do a community poll on suggested solutions, including single rigid body.

[@critterjet100]

This seems like a real good idea. Devs, please let the community help you steer the game where they want it to go. For as much as you may know you cannot be omnipotent. Let us help you make these critical decisions

[@mushylog]

No... @@critterjet100 Don't let the community steer anything, but DO let the community speak. Community can speak to the team, but the team should do what they believe is right.

[@critterjet100]

@@mushylog the only problem I have with that is the big game developers, often because they are influenced or forced to do so by businesses I who simply do not care (I think this is the case for ksp2), the devs will not choose what is right for the game and enjoyment of the fans but will instead choose the best option in their limited experience. No man rules alone and this applies to game devs. They are certainly sure they are doing the right thing, but the right thing to them is obviously leading to massive community push-back. Yes the community should not steer, bad word choice on my part, but the community and the fans should have a large influence in the final output. This game is being made with investment, financially and with time and dedication from the fans. The devs may be fans themselves but they are but a few. The base ksp2, I think should appeal to the wider community and more niche sects should be given every opportunity and tool to mod the game the way they see fit. This, I think is the strength of ksp 1 and other games like early Minecraft.

[@drewcipher896]

The community if full of fucking morons. No design decisions should ever be made by the community. That's how you end up with grinds and battle passes when streamers influence game design.

[@CysicSystems]

Perhaps it's worth employing a few senior engineers into the team to focus on this issue rather then continuing to rely on a team that are likely burnt out and lack the required competency to fix the issue. The game is completely unplayable in the current state unless you want to play with extremely small rockets with few parts. Given these issues where present in KSP 1 for over 4 years, and plagued development I'm surprised we are still discussing the art of the possible without an actual road map. Who's project managing this mess?

[@rosengrenj9]

i think in the meantime, a quick fix should be just making everything rigid. it would improve performance, and ships would be much more flyable. you can just apply that in a patch and we can be happy with that until you have a more elegant solution figured out.

[@drewcipher896]

Stupid idea. They should focus on the finished product and not the early access experience.

[@clayel1]

@@drewcipher896so ignore the community that spent $50 on the game? wtf lol

[@danczer1]

Can you share videos of different methods you tried? I think lot of players can't imagine how surreal a flying rocket which has 100:1 ratio and flies like a brick.

[@orenong]

From what he says it's clear that they just use unity's physics system, which is garbage. I don't think that the old guy from QA will have the solution for it. they need to scrap it all and start over

[@SpearHead1011]

I dont think they have even tried tbh. It sounds like they are scared to even touch it. Seriously

[@brianboyd8692]

let me call your buff, hate wobble

[@UsernameInvalid48]

Wait this is still a thing? Guess I'm going to continue taking a break from KSP2 then.

[@mephisto8101]

I don't want to have ANY wobble in my planes, spaceplanes or rockets. Parts may fall of when they are over their respective limit. Parts may disintegrate. But. They. Should. Not. Wobble. This is not fun, nor in any meaningful way realistic. Make it rigid body simulation and improve performance. Please.

[@miguelvera9472]

Hope y'all are doing ok, physically and mentally

[@SuperYtc1]

You hope that people who are scamming people over £40 for a broken mess that's barely developed are doing ok? What about the people who fell for this scam and lost their money?

[@iveharzing]

@@SuperYtc1 You are being very hyperbolic. It's a difficult game to make, and they're clearly working very hard on it.

[@Draxi_1]

@@SuperYtc1 is it a scam when you buy unfinished product that literally says is unfinished on the store page? Not sure what your expectations were. Just learn to read.

[@miguelvera9472]

@@SuperYtc1 I am not talking about the company, I am talking about the developers as people, who clearly care about this game and are passionate about it, as someone whose greatest fear is being hated, I don't know if I'd be able to live knowing that there's tons of people out there who hate me, who knows how many death threats they've gotten in the past few months? I am one of the people who got "scammed", I regret it, sure, but I still want the developers to be mentally healthy, because I know they're trying their best to give us the best product they can, because we can't just assume that they're all evil because the company decided to releasse the game too early, or because it was too expensive, they're not in control of any of that. And I haven't even mentioned crunch culture.

[@danieldc8841]

@@SuperYtc1excuse me I paid £50!

[@FatovMikhail]

autostrut is not about cancelling wobbliness, it's a part of design where you configure the required rigidity for the price of structural mass and cost

[@thrasher2344]

I can't even play this game until wobbly rockets are gone. Ksp1 all day until then.

[@dXXPacmanXXb]

Its embarrassing that this is still an issue for you guys

[@captainahab5522]

A possible solution to a bit of the problem is to make everything inside a closed fairing a rigid part. And decrease the flexibility of all joints especially joints in vertical stacks Joint rigidity should be changed by the area of contact

[@temitelko4413]

when do you fix the wobbling rockets?

[@elbuglione]

Soon™

[@lockscad]

"Lose some of that novelty" = "It's not a bug, its a feature" 🙄 its a freaking game that is currently not enjoyable to play! Forget about "realism" for a second, and make it playable already!

[@JG-yk6ny]

Playability should always take the front seat over realism. You are absolutely correct, first and foremost this is a video GAME.

[@ShadowDev6969]

well if you build dumb stuff, ofc it wont fly

[@JG-yk6ny]

@@ShadowDev6969 They wouldn't be addressing this for the umpteenth time if this wasn't an outsized issue my man.

[@ez_theta_z9317]

@@ShadowDev6969 it's ksp. building dumb stuff is like, 50% of the game. and in ksp1, a majority of that dumb stuff is flyable. in ksp2, an SLS (irl planned launch system, definitely not dumb) recreation needs about a million struts to not fall apart

[@Zacho5]

That seems a little hyperbole. I can make a sls alike with like 6 or 8 struts at the most and those are all srb support same as ksp1.The large 7.5m tanks don't flex that much.

[@StoryMode180]

My thoughts during this Video: Wobbly rockets were an issue previously in KSP1 but were manageable. It appears to be significantly more of an issue from what I've seen in KSP2 footage. Novelty isn't what I would call mechanical issues. Good to hear what you are working on currently, I had not heard about an orbital drift problem, but that would cause issues. I am also glad to hear you say you plan to address the wobbly aspects; thank you for confirming. In KSP1 there was an issue of "wobble" between tanks in line with one another, where force applied would bend a tank at an angle showing the solid ends where they normally connect. In a real world scenario, this would be a rupture and would be the end of the rocket. Obviosuly not intended behavior. The footage I saw of KSP2 showed a similar issue. That is what I would call "wobble". I don't actually remember auto-strut, thanks for the explanation; and ya I don't think that's more than a short term possibility. I look forward to following your development. Once some of these issues are addressed, I look forward to purchasing KSP2.

[@memeslife-wq2tx]

"It's not a bug it's a feature!"

[@starfleetau]

Here's a hint, give us Procedural tanks.. which then allows us to design the tank to the size we want but then that would be smart. This was largely solved well before, and it SHOULD NOT be being solved at this point, you should have solved this well before now, but instead your pointing out more and more that you guys lied to the community about how far along the game was when it went into EA. And I think Starship's first launch proved that it takes a LOT to cause some of the stuff that just happens way to quickly in KSP2

[@alexsiemers7898]

Not even procedural tanks would fix everything since the joints between engines, fairings, decoupler, etc are still a concern

[@comet2085]

i think there is a very important part missing in this conversation: the main thing the wabels is showing a build up to the rocket destructing so the player can estimate how its going and if they need to intervenieren. this would mean that the wabble is only needed to tell the player not to destroy the rocket. the rocket should be destroyed because of the forces not the wabble.(I mean like an increase. not that the force shouldn't wabble the rocket apart that's fine)

[@TheGreatDrAsian]

If you don't fix the game, I'm not buying it. Simple as. It shouldn't have been broken from launch to begin with.

[@cameronharrington354]

ill buy this game in 10 years when its the coolest shit ever

[@goku445]

How do you pay devs before that?

[@jonmedders3866]

sadly, with sub 100 players, they will never have the money to continue development that long.

[@mantis1s1k]

@@goku445 I'm not paying a dev to eventually make the game they promised. If I wanted to go that route I'd buy all the ships in Star Citizen. This developer promised a full game sequel with interesting new features not found in the original game. Not only are none of the new features present, but key features in KSP1 aren't even included. They need to earn their pay and produce what they said they would.

[@goku445]

@@mantis1s1k And no one is forcing you to do so. They never said the game would have these features on early access. Bunch of clowns.. Go play Mario Kart.

[@goku445]

@@jonmedders3866 Do the players pay a monthly fee? Your argument is rather pointless.

[@PaddyPatrone]

Im just mad that I didn't refund ksp2 right away. Now I spend 50€ on something that will never be finished. I lost all hope on this game

[@goku445]

Already?

[@PaddyPatrone]

@@goku445 They are fighting the absolute basics. Like, these things should have been dealt with early on.

[@goku445]

@@PaddyPatrone If you had listened you'd know this isn't "asbolute basic". Who are you to know what is basic or not?

[@manvslife271]

​​@@goku445why would I listen to the words of a man who has been constantly lying about the state of the game?

[@goku445]

@@manvslife271 Well. See you in a year when the game reaches a playable state for your taste.

[@Uloncl]

so u can just turn off the joints? i think your idea of what this game is about is different from what we think its about, if a rocket is bending then it is literally ripping itself apart and or there has been massive structural damage and it is effectively mission over haha funny big youtuber is laughing at the flaccid rocket, i guess we should persue that and make it a million times more difficult for people who want to love this game to make anything cool edit oh great glad to see ur spending valuable development time on some fancy cosmetics and animations, glad to see so much progress being made /s

[@TlalocTemporal]

Eh, animators do animation. Trying to put them on physics will be a bad time for everyone.

[@dirzz]

This game is being made by actual kerbals, through sheer trial and error. Just let it go guys.

[@chopsu123]

Yall put this game out way too early!

[@Samuel-j9q2b]

They try theyr best, now shut up

[@iemandjwzz4614]

Got most likely pushed from the management above

[@Xeloboyo]

Sad reality of modern publishers

[@danzstuff]

yeah that's either our fault or the publishers fault

[@rixyarbrough1702]

doubt they wanted to lol

[@LeastNationalistPole]

that's right folks! long time owner Mr. Krabs is opening a new restaurant called Kerbal Space Program 2! first of all, congratulations Mr. Krabs *hello i like money*

[@mantis1s1k]

You've released a sequel game in early access for full price with fewer features than the original and only the promise of grander features coming later. I wanted this game to be amazing. The idea of colonizing interstellar space had me ready to buy on day 1. Why should anyone pay $50 for a game which is prettier, but offers so much less than the original? There are plenty of mods which will give KSP1 a glam-up. I wish you could take the 20 minutes you've spent chatting about wobbly rockets and redirect it toward heating, colonies, interstellar travel, mining and resource management etc... or at the VERY least letting everyone know when those features are coming.

[@SirWiseKnight]

Damn so i guess i really do have to wait 5 more years to even think about buying this....

[@neotuxxedo]

Just an idea, but to temporarily mitigate wobbly rockets you could add an option to be able to fuse the rocket and make it one object, it would probably make it way less realistic, but players want to actually see their creations fly, and until wobbly rockets are fixed, it would take a lot of pressure off of the devs.

[@drewcipher896]

That's a waste of dev hours. They should focus on finishing the game and not the wims of early access users.

[@WillowLiv]

They actually talked about that, although the lingo was definitely technical. It's what they called treating the rocket as a single rigid body. They said that while that works, it has some problems with wheels. Regardless, they said they don't want to do this as it would take away from the magic when in situations where you would expect wobble, or rather, bending, you would see none. Although they did talk about doing a quick solution just while they figure it out, but they said that autostrutting had some issues with performance, and nate hinted at wanting to figure out the long term solution first, rather than focusing on a quick fix. So chances are, we might not see much until they figure it out. Which will be a long time, most likely.

[@sircotty2330]

8:55 kitbash model club uses this system, and works fine. what"s stopping you from using it? starship barely even bent while it was twirling out of control! Rockets. should. not. bend.

[@vibinseal]

yeah....I'll just go on and keep playing KSP1

[@zroku123]

is it possable to allow us users to adjust the tension value on struts ourselves? like we can set high tension value struts where we want integrity and then less tension value struts to allow flexing?

[@MasonSeibel]

I've seen more impressive games made in a quarter of the time. This was supposed to come out three years ago for early access. It's four years after initial development, It's been eight months since early access came out. Little progress has been made. I am an avid supporter of this game and the franchise, but you guys are losing me and a large portion of your playerbase with this.

[@NZOCD]

It’s fascinating watching IG/PD trying so hard to justify taking your $50 in exchange for a crappy broken tech demo. Absolutely appalling progress in 7 months.

[@msmithdev1369]

Sad truth.

[@fabiopauli420]

has it really been 7 months? jesus christ

[@acathosh]

I have a crazy idea: what about making the game fun? If it's slightly more realistic to have a shitty wobbly mess on your screen whenever you scale up, then it might be a good thing to not strive for maximum realism. People love realism in efficiency, speed, engines, orbital maneuvers. NO ONE loves the realism of an accidentally wobbly rocket, absolutely no-one.

[@alan-daniel]

I'm glad this is getting the design and development attention it is, and I'm 100% agreed that auto-struts are not a long-term solution. The concern I have is that it feels like it was community feedback that got the ball rolling on fixing the exaggerated wobbliness, and it's such a fundamental problem that I would have thought it would be one of the things "solved" long before now, especially given its history from KSP1

[@rraaiin]

71% like/dislike ratio 💀

[@injoru]

If you're having trouble deciding how flexible a connection should be - why not let players decide for themselves? (in a varying degree)

[@grtninja]

I feel like you guys are beating a dead horse over this. Sure, there's lots of particulars when it comes to the game's physics, and how things should interact. But maybe we can all agree that the "wobble dial" has been up to 10 all this time, and if you dial it all back down in steps, all at once, we early access players can let you know when it seems like enough. Also, there's a bit of limitation when using a keyboard to play, it forces single hits of lateral forces, rather than a smooth analog force from say an xbox controller, I'd like to see better analog controller support.

[@TlalocTemporal]

Control surfaces that take more than a single frame to actuate would help too.

[@ioresult]

I have an idea. Instead of looking at joint flexibility, how about looking at part flexibility? Instead of flexing at the decoupler, a rocket could flex at the level of a fuel tank. For example, at launch, the fuel tanks are at their most rigid, fully pressurized and heavy with fuel. But as the rocket empties, the tank looses rigidity etc. How would the physics calculation change? Instead of calculating joint flexibility, calculate part deformation?

[@janetfaulkner4256]

I have a mod that gets rid of wobbly rockets :)

[@SliptOnAChip]

like for mystery goo:

[@johnwilson839]

When Nate says he expects a single core tube part to be absolutely rigid, I take him to mean that he expects that the real world engineers who build those kind of parts for real rockets would build that tube such that it wouldn't wobble under the expected loads produced by the engines. The thing to remember is that KSP allows the play to experiment with poor design choices. if you make your tube too long for a huge payload and a huge engine then the sim should buckle or at least wobble. The road to hell is paved with making 'core structures' absolutely rigid. KSP1 got this trade pretty good imho.

[@joachim2464]

I honestly dont know why you have to do a physics simulation on every single piece all the time. At least fuel tanks should be solid. You can do a simulations of loads instead which do not require physics bodies and explode the pieces that exceed the load, then do physics on exploded or seperated pieces. Fuel tanks in rocket do not flex much. They disintegrate.