Please no wobble (Solved!!!)

[Davidian1024]

On 4/17/2023 at 4:57 PM, rjbvre said:

How bad are people's rocket designs on here? And lets be honest, the reason rockets don't flex significantly in real life isn't because they can't, it's because there are large teams of well educated professionals designing them not to.

I think this is a big part of the issue.  And a legitimate one.  Structural engineering is hard.  But for some reason alot of people seem to think it shouldn't be, or maybe they just don't want to deal with it.

But then, KSP 1 and 2 are drastically over simplified structural simulations.  Now in terms of video games they're probably more complex than anything out there.  I'm hard pressed to thing of a game that comes close.  If there are games that are more complex, they probably don't simulate anything else.

People want to fly big sophisticated craft that look cool.  It's not at all easy to do this realistically.  Trying to simulate a complex craft who's underlying structure accurately reflects it's visual appearance is just too much for people's computers.  A single fuel tank could consist of thousands of individual parts for example.

All that said, there is definite room for improvement here.  And the devs know that, and they're working on it.

Now to the people demanding to have the wobble removed, I say give them what they want.  In the form of a difficulty option.  An easier level that removes all flex from parts/connections perhaps.  Makes everything perfectly rigid.  I know a ton of people would use that.  And they'd probably love it.  But it needs to be a reduction of difficulty because real world objects are not perfectly rigid.  Wings should not be able to withstand any amount of drag and not get ripped off.  Long slender rockets should not simply be able to stand up to any amount of opposing drag and thrust forces.

I for one want the wobble in my rockets to remain.  But please do try to improve the situation.    By the way, I'm loving the direction KSP2 is heading in.

 

[whatsEJstandfor]

On 4/18/2023 at 2:54 PM, Davidian1024 said:

Now to the people demanding to have the wobble removed, I say give them what they want.  In the form of a difficulty option.

"It should be a difficulty option" is not a solution and I'm getting bored of how often it's suggested as a compromise. The difference between fun and not fun shouldn't be a difficulty slider. And building systems that can perform in multiple ways adds exponentially more QA work which, I hope it should be obvious, the game does not need more of at the moment.

[thewhitemetroid]

1 hour ago, whatsEJstandfor said:

is not a solution and I'm getting bored of how often it's suggested as a compromise.

I agree, I'm not a fan of a difficulty option simply because I feel it is a poor way to represent a real problem. I'm 100% on board with there being in game queues that indicate bad design, just not wobble because it in no way represents a real scenario for this particular instance in engineering and design. I understand that this is a game and cannot fully represent real life physics. However, I believe there are much better options which were shared in another thread that I will share here in case anyone wanted to read some more opinions on it.

Some of the suggestions include increasingly loud noises, sheet metal and other bits starting to fly off the vessel and I think also adding shaking effects to the UI as the vessel is approaching a failure state are much better representations of what happens when a there are design flaws and a failure is about to occur. I also like these ideas better because they would add a lot of tension to the launch when things start to go bad.

[pandaman]

It's jot so much the wobble itself that bugs me, it is after all a good visual clue to the weak points, but the resulting enforced 'gimballing' that completely wrecks any chance of recovering stability.

And it's even caused by using the right parts for the job.   Surely a part designed to rigidly connect two rigid cylinders should be able to do that!

[DancZer]

I see lot of comments about why not and almost nothing as a suggestion how should. Here is my suggestion for the improved wobbling, I hope some devs will see it and they could react.

Generally rockets are pressured gas tanks. Tanks has pressure and wall thickness. Since object does not bend in KSP, technically the wobbling is the only way to add a "bending" of object to the game.

• Pressure could wary based on the volume and amount of fuel (not necessary, because they usually fill the tank with another gas to keep the pressure on, but it could be a factor if you wish (probably it shouldn't be linear)).
• Thickness could be a parameter for structural parts + fuel tanks, as a tradeoff changing thickness should increase the mass
• Decuplers/docking should have strength, which would require more energy as a tradeoff

By adding similar options fo the parts, I think the Struct would not needed at all.

Benefits:

• Player has control over it, no tweaking needed form the development side (basically it would be a UX for the player to control what they internally try to achieve)
• It would be visible for the player, not like autostruct, which is a bit strange solution
• Struct on the vessels are not fancy at all

[RalphKerman]

Please watch today's video of one of KSP's most famous content creators: Matt Lowne

He explains in a very good way why wobbly rockets that do this on purpose have no place in KSP!

I agree with him 100%:

• Rockets that are built with serious intentions and follow a solid design pattern should not wobble at all and should not need 100s of struts. KSP (despite being in a cartoonish style) can be a very serious and respectable physics sim. And if people spend time to build stuff correctly the game should award this by being physically correct and not introducing artificial wobble.
• [A KSP 1 instance with RSS/RO is no "game" anymore, it teaches science and it really needs smart thinking to get a vessel into orbit let alone fly to another planet or moon. And that is amazing! This point is not in the video, it is my own opinion and emphazises the first point.]
• Fun creations of stupid things may wobble all the way and can provide entertainment but as Matt says: "This is fun for 10 minutes when you encounter it the first time". Not a gameplay loop for a game that so much more to offer!

And here is the link to his excellent video that came out toady (July 1st 2023):

https://www.youtube.com/watch?v=IpK_jGTJG3k

[Lisias]

With all the due respect, Matt Lowne is an ophthalmologist, not a rocket scientist neither a game designer.

If your rocket is wobbling, you are doing something wrong. The alternative is punishing bad designs by just collapsing them and letting the gamer trying to figure out what they did wrong.

That said, I'm not a fan of wobbling by itself, I want something to punish bad designs as well the abuse of the parts' limitations, being it wobbling or not. But if you remove the wobbling, you need to replace it with something else - perhaps flexing of the rigid bodies? I surely want my wings to bend under higher loads instead of just snapping out once their limit is reached, and now that the wings are a monolithic block being procedurally generated, we don't have wing bending anymore (and this is bad).

And whatever this wobbling replacement would be, it needs to be "auditable" so the user can diagnose the problem by analysis and thinking (and posterior trial and error) - but without demanding the user to get a rocket science degree first, this is still a game, not a training simulator for professional rocketeers.

Edited July 6, 2023 by Lisias
tyops as usulla...

[Vortygont]

Did someone notice, that it will be good if radial attached tank structures don't wobble? I made several planes with radial attached tanks, and problem was in wobble of these tanks. When you attach other tanks to radial tanks, they connected inline not radial and they wobble. Maybe it will be good to use modified wing system with multiple attachment nodes and apply it to the radial tanks, but game need to determine when and with which details use this soliution

[Sylvi Fisthaug]

To breathe more air into this discussion again: 

 

[Clóvis]

Wobbly rockets are not fun and should be removed. That said, if wobbliness is to become a feature it needs to be reliable and controllable, and not an wild stupid random thing that is only there for annoying reasons and to impede you to make the rockets you want. I mean, if wobble is feedback for a too long rocket, or a poorly designed aerodynamics, or you forgot to configure the stage and one of the four planned thrusters didn't ignite, or if you misaligned something something, if its used as feedback for something useful that is bad if you do with an rocket and its possible to understand and avoid it from happening because there is a physical based gameplay plan for it to be happening the way it happens I think is acceptable to have it in the game, and even could be a fun opportunity for game youtubers to generate content. But the way it works now its just frustrating and time consuming, making planning a rocket not as fun as it could be.

 

[whatsEJstandfor]

(Copy/pasting this here from the Wobbly Rocket bug report thread at @Anth12's request due to it being better classified as feedback)

 

My kingdom for a Beam.NG-style soft-body solver but, until then, let's get these rigid bodies to stay, ya know, rigid!

So far the arguments I've seen FOR wobbly rockets are "players should get feedback on when they've designed something poorly". Agreed! However, I think having elastic joints are just the wrong kind of feedback. I think @Kavaeric's Discord post here is right on the money:

Groaning/creaking and (imho, a purely cosmetic) rattling (like in the launch trailer) could tell the user that the joints are nearing their failure point, followed by catastrophic failure. I ran some tensor calculations on my Hello Kitty calculator and plotted this helpful chart:

[MechBFP]

On 7/1/2023 at 7:04 AM, Lisias said:

With all the due respect, Matt Lowne is an ophthalmologist, not a rocket scientist neither a game designer.

If your rocket is wobbling, you are doing something wrong. The alternative is punishing bad designs by just collapsing them and letting the gamer trying to figure out what they did wrong.

That said, I'm not a fan of wobbling by itself, I want something to punish bad designs as well the abuse of the parts' limitations, being it wobbling or not. But if you remove the wobbling, you need to replace it with something else - perhaps flexing of the rigid bodies? I surely wants my wings to bend under higher loads instead of just snapping out once their limit is reached, and now that the wings are a monolithic block being procedurally generated, we don't have wing bending anymore (and this is bad).

And whatever this wobbling replacement would be, it needs to be "auditable" so the user can diagnose the problem by analysis and thinking (and posterior trial and error) - but without demanding the user to get a rocket science degree first, this is still a game, not a training simulator for professional rocketeers.

I really like the idea of the rigid bodies bending, although I don't know how practical that would be to implement. Definitely doesn't seem easy.

[Kavaeric]

I went and wrote a whole analysis in its own thread about my own personal thoughts on the problem of wobbly rockets, as well as a possible solution. Might echo some of @Lisias's suggestions for bending rigidbodies.

 

[Superluminaut]

On 7/5/2023 at 11:07 PM, Kavaeric said:

I went and wrote a whole analysis in its own thread about my own personal thoughts on the problem of wobbly rockets, as well as a possible solution. Might echo some of @Lisias's suggestions for bending rigidbodies.

 

Interesting solutions. I have also thought about the artistic side of rocket fail states.

On 3/9/2023 at 1:27 PM, Superluminaut said:

I actually think kerbals are that gauge of intensity you're looking for. Kerbals shake, vibrate, and have panic attacks. As a player you mirror their emotions and get a feel for the crazy excrements your vehicle is going through. Also kerbals are happy-go-lucky and have different personalities, so you want to save them, you're emotionally invested in the success of your missions. In orbiter if you get stuck on the moon you just try again. In KSP you launch the rescue mission, 10x more powerful than the original mun lander mission. There have been a lot of rocket sims, but KSP is the one that really connects with people, and I think the differences is the kerbals. Both KSP1 and KSP2 could do a much better job making use of the power of kerbals, in my opinion.

I'd also like to to see your thoughts on additional fail states.

For instance

• Have segments disconnect if too much force is stressing a joint. 
• Have fuel tanks crush like soda cans if they have too much weight or force on them.
• Have shearing animations for parts.

Given only a handful of failure animations for parts, every system of rocket parts failing would be like a fingerprint. I feel as though this  would dramatically add to the joy of failure in KSP.

[Superluminaut]

I did a little testing, and I'm calling it solved.

How is everyone else's experience?

[rogerawong]

10 hours ago, Superluminaut said:

How is everyone else's experience?

I got KSP2 when the For Science update came out, and was like "What wobble are they talking about"  until it happened to me.  Put a rocket on the launch pad, and then kraken came and entire cylinders of my rocket started oscillating backwards and forwards in resonance until it blew up right there on the launch pad.

I've since had other experiences where upon loading a game, parts have been moved inside other parts, or a part just randomly snaps off while in orbit, or suddenly a ladder or something that deployed just fine suddenly is "blocked".   And my 10-kerbal-to-Eve craft shook itself to pieces in orbit as well.

 

Edited February 10 by rogerawong

[steveman0]

The new joint physics are a massive improvement. Most "traditional" rockets I build never need struts. Only some really massive asparugus staged rockets might need a strut or two to reinforce the boosters to keep them from flexing and colliding with other parts. This seems perfectly reasonable to me.

There's definitely some remaining physics instability with craft in orbits though. I too have seen craft shake themselves to oblivion spontaneously. This suggests the algorithm for the wobble is still mathematically unstable. A stable craft would have mechanical dampening from friction and bending stresses that should prevent spontaneous shaking to pieces unless a substantial input force was involved. These flexing forces should include some kind of dampening factor to ensure the runaway wobble isn't possible.

[Superluminaut]

On 2/10/2024 at 6:38 AM, rogerawong said:

I got KSP2 when the For Science update came out, and was like "What wobble are they talking about"  until it happened to me.  Put a rocket on the launch pad, and then kraken came and entire cylinders of my rocket started oscillating backwards and forwards in resonance until it blew up right there on the launch pad.

I've since had other experiences where upon loading a game, parts have been moved inside other parts, or a part just randomly snaps off while in orbit, or suddenly a ladder or something that deployed just fine suddenly is "blocked".   And my 10-kerbal-to-Eve craft shook itself to pieces in orbit as well.

 

 

4 hours ago, steveman0 said:

The new joint physics are a massive improvement. Most "traditional" rockets I build never need struts. Only some really massive asparugus staged rockets might need a strut or two to reinforce the boosters to keep them from flexing and colliding with other parts. This seems perfectly reasonable to me.

There's definitely some remaining physics instability with craft in orbits though. I too have seen craft shake themselves to oblivion spontaneously. This suggests the algorithm for the wobble is still mathematically unstable. A stable craft would have mechanical dampening from friction and bending stresses that should prevent spontaneous shaking to pieces unless a substantial input force was involved. These flexing forces should include some kind of dampening factor to ensure the runaway wobble isn't possible.

There are certainly still many problems with the game. Loading bugs and kraken encounters, I don't believe they fall under wobble. However I have not seen exactly what you have seen. If at all possible please share a video so that we can see if sudden, unexpected instances of wobble do occur.

I have not put a whole lot of time into KSP2, it's still too frustrating for me to relax over, so from my perspective it may be entirely possible that that it's still out there.

[steveman0]

6 hours ago, Superluminaut said:

 

There are certainly still many problems with the game. Loading bugs and kraken encounters, I don't believe they fall under wobble. However I have not seen exactly what you have seen. If at all possible please share a video so that we can see if sudden, unexpected instances of wobble do occur.

I have not put a whole lot of time into KSP2, it's still too frustrating for me to relax over, so from my perspective it may be entirely possible that that it's still out there.

I don't have any video on hand for this, but it basically is a spontaneous wobble of a ship in orbit that oscillates with increasing magnitude until the craft shakes itself into pieces.  At first I suspected SAS, but I've seen it on craft with SAS disabled.  Time warping can suppress the effect as this halts the physics sim for this motion.  Some stray calculation is causing a runaway feedback loop until the craft tears itself apart.  It's most prone on large interplanetary craft.  My largest had 3 instances of this effect in one game session.  I'd have to see about setting up OBS again to try to record it.  It's quite random when this occur so it might not be something that can be easily reproduced.