How should rockets flex?

[Guest]

4 hours ago, cocoscacao said:

However, take a different perspective on this. If wobble is an indication/punishment of a stupid design, then why include struts in the first place? They negate the punishment part. If they're there to patch wonky game mechanic, in KSP 1 that's fine by this point, but in KSP 2, either fix that mechanic, or remove it completely.

It's not  "punishment" of "stupid design", it's part of an engineering problem. Struts provide one way to solve that. KSP1 tended to remove all engineering problems and I originally thought this was a good idea but on coming back to check out vanilla before KSP2 launch it just trivialized part of the game; all I needed to do was click on autostrut (and even that, in a lot of cases, wasn't needed), no thinking about how to fix an issue or desperately and carefully shepherding a floppy rocket to orbit. Failure points were so few and far between. If you were playing a heavily modded RO this worked fine, there were many other challenges to overcome, but in vanilla, with the cartoony solar system and reliable to a fault vehicles, it helped make the game boring.

[Dakitess]

KSP1 without struts / autostrut is still wobbly, way enough, way too much actually, to me.

It can be 3 times as rigid and would still allow for "understanding" that a 50m tall 1m diameter wide rocket is not a good idea, especially based on a discrete stacking of 10 tank rather than 2 monolithic stages.

[MechBFP]

On 7/1/2023 at 7:41 PM, regex said:

I actually completely agree. If we were talking something like RO I'd be all in favor of autostrut or rigid joints because I have other constraints on design to contend with. With stock KSP1 and autostrut there is very little engineering challenge, it's way too simplistic and trivial a game with it, especially for experienced users.

Most issues of wobbly rockets in the current build can be fixed with sparing and intelligent use of struts. I am not at all opposed to reducing the wobble but KSP1 level autoatrut is, quite frankly, baby mode KSP.

Yup. There is very much a “no fun allowed” vibe in some of these arguments, quite reminiscent of the RPGCodex crowd lol. 

[Geredis]

17 hours ago, Lisias said:

Oh, no. I already handle CRUD (Create, Read, Update, Delete) features all the time at Day Job™, I don't want to handle SQL statements on KSP too!!!

Lol, no I meant it more as "Catastrophic Rapid Unplanned Disassembly" - basically a step beyond the "standard" Rapid Unplanned Disassembly.

[cocoscacao]

5 hours ago, regex said:

If you were playing a heavily modded RO this worked fine, there were many other challenges to overcome, but in vanilla, with the cartoony solar system and reliable to a fault vehicles, it helped make the game boring.

I didn't. I'm fairly new in terms of playing hours. RO is for players who can basically do anything in vanilla game. Keep that in mind. While landing on most bodies is a boring chore for me now, I still remember the crowning achievement of (very clumsy) 1st time landing on the Mun.

Non faulty engines are good. Base game should be deterministic, because punishing a player by a coin-flip for doing everything right for the first time is a bad game design. Keep masochism for mods

I'm hoping that KSP 2 planned scope will bring a whole different set of challenges for veterans to enjoy, but that doesn't include struts.

My biggest concern with wobble isn't launching a rocket, but making a space station to which you can dock dozens of ships. That's not possible in KSP 1, and while space stations are just a gimmick there, I understand they'll have bigger role here (OAB 'n stuff)

Edited July 3, 2023 by cocoscacao

[Guest]

30 minutes ago, cocoscacao said:

My biggest concern with wobble isn't launching a rocket, but making a space station to which you can dock dozens of ships. That's not possible in KSP 1, and while space stations are just a gimmick there, I understand they'll have bigger role here (OAB 'n stuff)

Space stations that have colony functions will be treated as a single-part, near as I can tell, so that shouldn't be an issue.

[PDCWolf]

On 7/2/2023 at 2:38 PM, LoSBoL said:

I don't think SRB's in real life are just fitted to their decoupler only

You can see the lower attachment points (the ETA Ring) in this video:

Real life doesn't have a hard part limit based on simulation performance, or node tree based construction. Real life SRBs are attached at multiple points via exploding-bolt or hydraulic based decouplers. This is probably the worst case, as radial attachment wobble is effectively a punishment on the player for what's essentially the game's own fault: not having the right tools to build with, nothing to do with their ability to engineer a rocket. Heck, adding struts to anything detachable nullifies the ejection force from decouplers and that's been an unsolved bug since the launch of KSP1.

Need to attach 2 struts to SRBs? that's 4 extra parts on x2 symmetry, 6 on x3, 8 on x4, 16 on x8 and so on. It quickly adds up if you need more struts or want to attach stuff radially anywhere else.

Back before autostrut, KJR, and big parts, when you had to use octogonals + struts lenghtwise to make big rockets strong, the craft were pretty much a third rocket and 2 thirds struts and their structures. I feel this is being grossly ignored for this entire thread, specially since struts punish people for having a bad computer as well.

On 7/2/2023 at 2:38 PM, LoSBoL said:

how have you learned playing KSP?

That people can learn by trial and error doesn't mean it's the right way. It is clearly denoted that trial and error is not something the dev is pursuing by many other aspects being communicated in game, like dV, TWR, heat management, electricity, action groups, and so on.

On 7/2/2023 at 2:38 PM, LoSBoL said:

No they are not 'those arguments'

When you ask "how did you learn to play?", I'm 100% sure you expect me to reply "I figured it out". Like yes, I'll definitely take the blame for taking those kinds of arguments to the absurd but at the core, it's what's being said, just with prettier words. Some more examples:

• Me & others: "Your rockets are clearly not as well-engineered as you think they are".
• With stock KSP1 and autostrut there is very little engineering challenge, it's way too simplistic and trivial a game with it, especially for experienced users.
• It just seems rather logic to me that when a rocket wobbles, somethings wrong in the build.
• I have been playing KSP since 2012 and have never had trouble with excessive flexing in my builds.

These arguments do not help the conversation for either side, only attempt to ridicule people who find an issue in wobble by attacking them, and not the wobble or the arguments those people have. Funnily enough, doing it to me is no problem (for me), but Matt literally made a video with a simple vertical rocket, and they're telling the guy that his rockets are bad, lmao.

12 minutes ago, regex said:

Space stations that have colony functions will be treated as a single-part, near as I can tell, so that shouldn't be an issue.

Source?

[Alexoff]

36 minutes ago, regex said:

Space stations that have colony functions will be treated as a single-part, near as I can tell, so that shouldn't be an issue.

Has this ever been announced?

[Guest]

53 minutes ago, PDCWolf said:

Source?

Pretty sure it's been sussed out in earlier conversations on the topic but, true to this forum, I'll let you find it yourself and draw your own conclusions.

[Scarecrow71]

I think what people are forgetting is that in a science or career game struts aren't available until you get General Construction...and autostruts aren't available until then either.  So players do get a lesson in wobble up to that point, at which point technology makes it obsolete.

Edited July 3, 2023 by Scarecrow71

[Pthigrivi]

I love Matt but I always find it a bit funny when folks speak on issues of canon as if there is some hard and fast truth there. Kerbals are fictional, we are playing the game, so there's nothing real about them being stupid or sophisticated. They are however we play the game. Likewise the meaning of the word 'kerbal' used as an adjective is entirely subjective. To me it means fun, experimental, and iterative. There's nothing necessarily janky or negative about it. Im sure others have other connotations in mind.

As to the main question I think whats getting lost a bit here is that we're talking about a game thats all about physics and its absolutely essential that these machines respond to physical forces and aren't carved out of diamond. In real life basically everything flexes and bends under force, it's just so subtle you can't see it unless something has gone catastrophically, tacoma-narrows wrong. Its important for a game like kerbal to exaggerate that so players can see what is happening and where the weak points are, but, and Nate was very clear in this in the last dev-post, it shouldn't happen or cause launch failures or frustrating flight characteristics for vessels that are designed in a realistic way. Matt's absolutely right that many of those simple, single stack vessels should not be wobbling and it sounds like Nate agrees. But yes if you have a side booster that's only attached at one point the thing should flex to show the player that they haven't adequately reinforced it. 

I personally do not use autostrut in KSP1. Most vessels don't flex because I have a pretty keen idea how to place just a few, usually diagonal struts to efficiently counteract lateral forces. The one place this breaks down for me is on really huge orbitally assembled vessels where I can't use struts and need to rely on single-point docking port connections in stock. Hopefully down the road in KSP2 orbital assembly platforms will help with this too. The one place I might say some kind of autostrut might be called for is inside fairings and cargo bays where once again you're often relying on a single docking port for structure. 

I do also agree that there should be a bit of training here too. Just because a thing has a strut on it doesn't mean its being strutted in the right direction. Even inside a fairing you can create a lot of stability with regular struts if you know how. The key here is to connect these things in triangles or x-bracing  to create 3d trusses so that you're resisting side-to-side stresses rather than forces that are straight up and down. 


 

2 hours ago, regex said:

Space stations that have colony functions will be treated as a single-part, near as I can tell, so that shouldn't be an issue.

Ive been in and out in the last few months but last I heard one of the key goals of colonies and stations is that they would in fact respond to physics and deformation under loads.  

 

Edited July 3, 2023 by Pthigrivi

[Vanamonde]

Overlapping threads have been merged. 

[Lord Aurelius]

The Dev's responses on this topic and focus on the silliness and intentionally keeping bugs like wobbly rockets leave me really concerned that they don't really understand the KSP player base. This is not goat simulator or world of goo. Casual players which might mess around in KSP and are NOT the core player base might have fun with this for a few minutes and move on. The core players who are the evangelists for this game and are speaking up here don't really appreciate this physics wonkiness and want to build realisic-ish rockets which don't wobble.

[PDCWolf]

4 hours ago, Pthigrivi said:

I love Matt but I always find it a bit funny when folks speak on issues of canon as if there is some hard and fast truth there.

I'll first clarify again that anybody is encouraged to play however they want. However, (and to draw a parallel) in Skyrim for example, you play the dragonborn, in nirn. Your dragonborn can be a clumsy dork that gets his/her head bashed in at every turn of a road, however they will always be the dragonborn, the only one able to absorb dragon souls, the only one capable of killing alduin and so on, and the world will always be nirn, a medieval-fantasy setting where people haven't yet discovered gunpowder, but they've already invented most of the advanced medieval machinery we know, and used magic for the rest. No matter how you play or your headcanons, the real canon is set in stone.

The same happens here: You look at Kerbals and they've been able to build an amazing (even more so in KSP2) space center, and have produced parts, like rocket engines, that are also more advanced than what we have. Further on, it's gonna be cannon that they've managed interstellar flight, orbital construction, and interplanetary colonization. Doesn't matter how clumsy you make them in your save, or if you never go past basic tech on the tree, the canon is already written. You can chose to ignore and over-impose your headcanon, but that will never change the actual canon, which is that Kerbals have managed to get more tech than humans, and it looks very properly built as well.

4 hours ago, Pthigrivi said:

Likewise the meaning of the word 'kerbal' used as an adjective is entirely subjective.

This should be the truth, yet we can clearly see how the quality of the game was degraded, or threatened to be degraded by allowing this: the barn, early tier buildings, wobble, useless part descriptions that had to be dialed back to usefulness, and so on. It shouldn't be a problem, but it's become one long ago, and sadly it goes much further than just respecting each other's opinions when the devs take the jank culture and use it as an excuse to propose actual jank.

4 hours ago, Pthigrivi said:

But yes if you have a side booster that's only attached at one point the thing should flex to show the player that they haven't adequately reinforced it. 

How is this not punishing the player for the game's own shortcomings (tree node building not allowing two decouplers to hold a single booster, for example)?

4 hours ago, Pthigrivi said:

Hopefully down the road in KSP2 orbital assembly platforms will help with this too.

Kinda off topic but please try to multi-port dock on KSP2.

4 hours ago, Pthigrivi said:

I do also agree that there should be a bit of training here too. Just because a thing has a strut on it doesn't mean its being strutted in the right direction. Even inside a fairing you can create a lot of stability with regular struts if you know how. The key here is to connect these things in triangles or x-bracing  to create 3d trusses so that you're resisting side-to-side stresses rather than forces that are straight up and down. 

Again, my main point against this is the limited part budged. You're telling people to waste 3 to 10 parts to hold down a payload. Meanwhile in real life, no struts on payload adapters:

Just now, Lord Aurelius said:

This is not goat simulator or world of goo.

Words to live by.

[MechBFP]

40 minutes ago, Lord Aurelius said:

The Dev's responses on this topic and focus on the silliness and intentionally keeping bugs like wobbly rockets leave me really concerned that they don't really understand the KSP player base. This is not goat simulator or world of goo. Casual players which might mess around in KSP and are NOT the core player base might have fun with this for a few minutes and move on. The core players who are the evangelists for this game and are speaking up here don't really appreciate this physics wonkiness and want to build realisic-ish rockets which don't wobble.

Who pays the bills, the former or the later?

[cfds]

The wobbling of rockets is a direct consequence of the use of the Unity chained rigid body system, that Squad introduced because they did not know better and Intercept kept because they prioritize visuals over substance. There is no intend to "teach" or "punish" players for unrealistic designs (otherwise a stack of four 0.5m tanks wold have the same wobble as one 2m tank), there is just an engine that is not suited more than a very basic single stack rocket. Unfortunately, there is far more money to make from a horde of "casuals", who spend two hours crashing Kerbals into the space center, than from a handful of "core player", who want to have a robust game that allows them to explore the system to their heart's content...

[LoSBoL]

8 hours ago, PDCWolf said:

When you ask "how did you learn to play?", I'm 100% sure you expect me to reply "I figured it out". Like yes, I'll definitely take the blame for taking those kinds of arguments to the absurd but at the core, it's what's being said, just with prettier words. Some more examples:

• Me & others: "Your rockets are clearly not as well-engineered as you think they are".
• With stock KSP1 and autostrut there is very little engineering challenge, it's way too simplistic and trivial a game with it, especially for experienced users.
• It just seems rather logic to me that when a rocket wobbles, somethings wrong in the build.
• I have been playing KSP since 2012 and have never had trouble with excessive flexing in my builds.

These arguments do not help the conversation for either side, only attempt to ridicule people who find an issue in wobble by attacking them, and not the wobble or the arguments those people have. Funnily enough, doing it to me is no problem (for me), but Matt literally made a video with a simple vertical rocket, and they're telling the guy that his rockets are bad, lmao.

I'm not attacking anyone, if you feel attacked, that's you. I can't help it that when I share my opinion, you read something else with bad faith in mind of the one sharing that opinion and feel ridiculed. 

As said by someone else before, lower your guns.

[Pthigrivi]

4 hours ago, PDCWolf said:

I'll first clarify again that anybody is encouraged to play however they want. However, (and to draw a parallel) in Skyrim for example, you play the dragonborn, in nirn. Your dragonborn can be a clumsy dork that gets his/her head bashed in at every turn of a road, however they will always be the dragonborn, the only one able to absorb dragon souls, the only one capable of killing alduin and so on, and the world will always be nirn, a medieval-fantasy setting where people haven't yet discovered gunpowder, but they've already invented most of the advanced medieval machinery we know, and used magic for the rest. No matter how you play or your headcanons, the real canon is set in stone.

The same happens here: You look at Kerbals and they've been able to build an amazing (even more so in KSP2) space center, and have produced parts, like rocket engines, that are also more advanced than what we have. Further on, it's gonna be cannon that they've managed interstellar flight, orbital construction, and interplanetary colonization. Doesn't matter how clumsy you make them in your save, or if you never go past basic tech on the tree, the canon is already written. You can chose to ignore and over-impose your headcanon, but that will never change the actual canon, which is that Kerbals have managed to get more tech than humans, and it looks very properly built as well.

Im probably going to quote you a bit here  but only because I think these are quality terms upon which to base discussion. I think you're right, even in the vaguest terms all games create canon and story because that's a big part of how humans understand the world. Some canon is loose, some is hyper specific. KSP is way on the loose end. There's not a lot of fine definition. As far as I'm concerned keeping canon really flexible is a good thing for a game like this because its not really about worldbuilding, it's about physics and chemistry and invention and things that are much more universal. Kerbals are little green versions of some of humanity's core qualities--inventiveness, intelligence, and a willingness to take risks. So it makes sense to give a lot of leeway for players to infill their own stories. Honestly this is a bit why I'm not personally impressed by the wacky ancient-aliens lore building that's going on with anomalies, but, what can I say. Im not writing this thing and my version of KSP2 might have been unnecessarily dry. 

To your point though just because KSP contains drives that are technically speaking advanced doesn't really mean much for how players actually engage. As a matter of gameplay we're not dealing with real fusion physics, all that matters to players are force, efficiency, and spacetime. For any player who's adjusted to the basics of landing on the Mun that's actually plenty to base your game on. All of us who have spent thousands of hours arduously landing on and escaping from Eve or managing clever gravity assists among the Joolean Moons knows there's incredible depth of gameplay there. I have very few worries about depth. The problem is the HUMP--engaging players from 8, 10, 16 years old who love space and find the physics sim fun, drawing them into that deeper exploration challenge, and most importantly giving them the tools to do so effectively. Part of that is being generous about expectations. The game can't expect players to succeed and be stellar engineers day one. It expects them to test and fail and test again and not take failure too harshly.  In some ways the game expects players to be perpetually imperfect. 
 

4 hours ago, PDCWolf said:

This should be the truth, yet we can clearly see how the quality of the game was degraded, or threatened to be degraded by allowing this: the barn, early tier buildings, wobble, useless part descriptions that had to be dialed back to usefulness, and so on. It shouldn't be a problem, but it's become one long ago, and sadly it goes much further than just respecting each other's opinions when the devs take the jank culture and use it as an excuse to propose actual jank.

Here I would try to separate out an air of whimsy from actual programming deficiencies. The two are not dependent or linked. Whimsey and a light-hearted atmosphere for players are fine, fun even, and they do some important work lessening the stakes in a game that could otherwise feel incredibly punishing. Actual bugs and poor mechanics are not fun, and are not fine. 
 

4 hours ago, PDCWolf said:

How is this not punishing the player for the game's own shortcomings (tree node building not allowing two decouplers to hold a single booster, for example)?

Again, lets not conflate KSP2's poor execution with what could be informative, dynamic gameplay. Even in KSP1 I almost always connect boosters at a minimum of 3 points--A decouper just above the booster's dry COM so it will kick out and tilt away under aerodynamic loads, and two struts angled out to counteract lateral and torq forces. Keep in mind real rockets do this: 

https://axm61.files.wordpress.com/2018/02/afc75535-0165-4219-8838-0388ec999962.jpeg

4 hours ago, PDCWolf said:

Again, my main point against this is the limited part budged. You're telling people to waste 3 to 10 parts to hold down a payload. Meanwhile in real life, no struts on payload adapters:

And again KSP2's general performance should really be allowing for multi-hundred and multi-thousand part vessels, and 10 or 30 or even 50 struts really shouldn't be what buries vessels in lag. As far as what KSP1 is capable of I would characterize the satellite in this photo as having a pretty wide wheelbase. I personally would not bother strutting such a squat payload. But I would for something 3-4 times this length to diameter ratio. 

Edited July 4, 2023 by Pthigrivi

[RocketRockington]

The great majority of people on reddit and in this forum and responding to this poll all say "Rockets shouldn't wobble much or at all".

But somehow, a few people who constantly post in thread after thread supporting KSP2 also uniformly seem to think that the least popular thing is the right design decision.

Curious.

Edited July 4, 2023 by RocketRockington

[LoSBoL]

1 hour ago, RocketRockington said:

The great majority of people on reddit a s.in this forum and responding to this poll all say "Rockets shouldn't wobble much or at all".

But somehow, a few people who constantly post in thread after thread supporting KSP2 also uniformly seem to think that the least popular thing is the right design decision.

Curious.

Really? Take a look at the outcome of the pole in this thread.

[Alexoff]

54 minutes ago, LoSBoL said:

Really? Take a look at the outcome of the pole in this thread.

According to the results of the poll the vast majority believes that rockets should wobble weakly (as in real life) or be completely rigid.

[Guest]

If it was me I'd probably take a short cut and make most joints rigid but leave some of them flexible, like radial decouplers for example. 

It would lead to some odd behaviors like letting you build rockets with super skinny waists that remain rigid and might remove some packaging challenges for payloads, but I think the trade-off in simplicity would be worth it — no more noodle rockets and lots of kraken attacks banished. It'd still leave a bit of flexing and depending on which joints you allowed to flex, might leave some engineering challenges as well. 

Tuning joint rigidity with extra attachment points and adjusting values just seems like a lot of work for not much return. Of course I'm not working on the game so I don't know about all the complexities so there may well be a good reason for the approach they've chosen, and I'm sure it's possible to get it to work!

[KUAR]

1 hour ago, Periple said:

If it was me I'd probably take a short cut and make most joints rigid but leave some of them flexible, like radial decouplers for example. 

It would lead to some odd behaviors like letting you build rockets with super skinny waists that remain rigid and might remove some packaging challenges for payloads, but I think the trade-off in simplicity would be worth it — no more noodle rockets and lots of kraken attacks banished. It'd still leave a bit of flexing and depending on which joints you allowed to flex, might leave some engineering challenges as well. 

Tuning joint rigidity with extra attachment points and adjusting values just seems like a lot of work for not much return. Of course I'm not working on the game so I don't know about all the complexities so there may well be a good reason for the approach they've chosen, and I'm sure it's possible to get it to work!

This would also achieve a good reduction in CPU load I reckon and would satisfy most users. Wing attachments should probably flex and perhaps also where node attachments are mismatched or something like that. For inline parts, stress should cause connections to fail rather than flex in my humble opinion. Not a fully fleshed out answer I'm afraid, but part of one!

I bet that some enterprising people would find a way to turn this paradigm (some parts simulate flex and others don't) into a kraken drive or something but I reckon that figuring how to exploit systems, then deciding if you want to use the exploit or not, is part of the KSP DNA. It's engineering, just a discipline that doesn't exist IRL. You could put huge effort in to eliminate all exploit possibilities but somebody will always find a way to beat it, and if they can't somebody will mod to enable it. 

Edited July 4, 2023 by KUAR

[PDCWolf]

6 hours ago, Pthigrivi said:

Here I would try to separate out an air of whimsy from actual programming deficiencies. The two are not dependent or linked. Whimsey and a light-hearted atmosphere for players are fine, fun even, and they do some important work lessening the stakes in a game that could otherwise feel incredibly punishing. Actual bugs and poor mechanics are not fun, and are not fine.

The line between going for an "easy atmosphere" and proposing actual jank is not only very thin, but it's been passed many times. Again, I should not have to remind you of the barn debacle, you were there back then and know how heated up it got. Most, as reflected by the poll, the barn debacle, the part description debacle, and so on, do not want jank to be a Kerbal trait. Jank should be left to the player's own inventions and headcanons. As for the stakes, they're already comically low, with infinite reverts, Kerbals being re-birthed after a while, very forgiving re-entry heat, extremely strong jet engines, and all other complaints you can pile up of the game being too easy, many of which you can straight up take from this thread alone.

7 hours ago, Pthigrivi said:

Again, lets not conflate KSP2's poor execution with what could be informative, dynamic gameplay. Even in KSP1 I almost always connect boosters at a minimum of 3 points--A decouper just above the booster's dry COM so it will kick out and tilt away under aerodynamic loads, and two struts angled out to counteract lateral and torq forces. Keep in mind real rockets do this:

First off, adding those struts neutralize the decoupler's ejection force, so you've also gotta add separatrons, turning what should be at most 2 parts (decoupler on top and bottom, adjust forces to get the separation angle), into at least 4. The games own shortcomings are a thing, a limitation, and having to add struts punishes the player for it, specially lower systems which might already be strapped for part limit. As for realism, I showed some posts above that the SRBs on the SLS/Shuttle were attached by 1 ETA ring on the bottom, and a connection on the top that's also part of the decoupling system. In case of the falcon heavy, those struts are part of the decoupling connection as well, they aren't just "to stop wobble". This once again exposes an obvious shortcoming of the game, that somehow is passed down to the player to deal with instead of being fixed.

7 hours ago, Pthigrivi said:

And again KSP2's general performance should really be allowing for

It doesn't yet, and we don't know if it will. We can't work with maybes or hopefullys

7 hours ago, Pthigrivi said:

 I personally would not bother strutting such a squat payload. But I would for something 3-4 times this length to diameter ratio. 

Still no struts in real life.

 

[Pthigrivi]

3 hours ago, PDCWolf said:

...and all other complaints you can pile up of the game being too easy,

I don't know I've heard a lot of critiques of KSP and its gameplay but it being too easy isn't a common one. 
 

4 hours ago, PDCWolf said:

First off, adding those struts neutralize the decoupler's ejection force, so you've also gotta add separatrons, turning what should be at most 2 parts (decoupler on top and bottom, adjust forces to get the separation angle), into at least 4. The games own shortcomings are a thing, a limitation, and having to add struts punishes the player for it, specially lower systems which might already be strapped for part limit. As for realism, I showed some posts above that the SRBs on the SLS/Shuttle were attached by 1 ETA ring on the bottom, and a connection on the top that's also part of the decoupling system. In case of the falcon heavy, those struts are part of the decoupling connection as well, they aren't just "to stop wobble". This once again exposes an obvious shortcoming of the game, that somehow is passed down to the player to deal with instead of being fixed.

I've never had issues with this? Everything releases at once and the payload floats away, same as without struts. I've never needed to do this in a rush so I've never needed separations. I guess if I had a second stage inside a fairing?  As for the photo that's still pretty squat and shouldn't wobble without struts. Im talking about things like the below (2:00 minutes into the Matt Lown video). I still don't think this should wobble as much as it does in KSP2, but it's something that I would personally strut diagonally in KSP1. Again I think you're conflating separate issues. The game needs to be able to support 1000+ part vessels on decent machines regardless of wobble tuning. A dozen struts shouldn't be a deal breaker either way. I also think folks think they're disagreeing in this thread but they're mostly saying the same thing: there's way to much flex in KSP2 right now, but there also shouldn't be zero.