Will there be a 1.1.2 ?

[tater]

9 hours ago, fireblade274 said:

Step one: go to google.com

Step two: type in "ksp bug tracker"

Step three: press "I'm feeling lucky"

 

You will be lucky

 

 

The prerelease had a direct link right here on this forum. Why, pray tell, is that not a constant thing in the forum? They have modded and unmoved install forums, but no real link the the actual bug tracker. Yeesh.

[Bedwyr]

Alshain's video is exactly my problem with the gear. Simply unable to fly an early-career plane off the ground because of it. The only difference might be that my experience is the same pre and post 1.1.1. 

Edited April 29, 2016 by Bedwyr

[regex]

16 minutes ago, Alshain said:

I'd settle for being able to just disable the suspension.

I wouldn't.  It's pretty clear to me that some suspension is required but it varies by craft (just like IRL) and I don't trust two discrete settings to handle all situations.

[pandaman]

I'm with the adjustable suspension stiffness idea.  A heavier craft needs more stiffness than a lighter one, irrespective of physical size.

IRL suspension can be adjusted and they can fit stiffer springs etc to otherwise very similar, if not identical, suspension to cater for different circumstances.  An adjustable stiffness  slider would be the best way to simulate this in stock.

[ShadowZone]

Just installed 1.1.1, have more crashes than with 1.1
And maybe even a new Kraken, I call it the Moonwalk Kraken:

I also think I have way higher RAM usage than before. And landing gear LT-2 still attaches with wrong offset center in the editor.

But oh joy, we now can move the navball... /irony

 

So... yeah... 1.1.2 has to happen. Not really happy at the moment.

Edited April 29, 2016 by ShadowZone

[Kiro]

2 minutes ago, ShadowZone said:

Just installed 1.1.1, have more crashes than with 1.1
And maybe even a new Kraken, I call it the Moonwalk Kraken:

I also think I have way higher RAM usage than before. And landing gear LT-2 still attaches with wrong offset center in the editor.

 

So... yeah... 1.1.2 has to happen. Not really happy at the moment.

HAHAH, is this for real? Are you pressing anything or is that thing just sliding on its own?

[TK421d]

i never had any crashes in 1.1.0, and even heavily modded with every mod ive ever used all in one install i never went over about 6.5g ram.

i did however notice that my processor would be at about 80% stress (on a single core) when in the VAB/SPH.  but only about 5-10% in flight...  that bugged me.

 

still havnt gotten a chance to load up 1.1.1 to see if its changed though

[Nightside]

9 hours ago, regex said:

So basically we're stuck with a spring for suspension?  Great, looking forward to driving my old Volvo again.  And again.  And again.

I had that volvo too!

[regex]

Just now, Nightside said:

I had that volvo too!

I even replaced the dampers on mine one afternoon; no change.  A buddy commented on how silly the car looked flying into a turn while I was driving it "fast"; all that body roll...

[sal_vager]

Please remember to stick to the topic of the thread, rather than using this as another gripe thread on individual issues, the developers are aware there are problems with wheels/legs.

A 1.1.2 is more likely if you can provide clear unbiased information on issues on the Squad issues tracker.

Thank you.

[ShadowZone]

Just now, Kiro said:

HAHAH, is this for real? Are you pressing anything or is that thing just sliding on its own?

no input whatsoever was being made.

[allmhuran]

How about developing simplified approximations for friction, wheels, suspension, and so on in order to get a realistic end result rather than attempting to develop realistic simulations that give unrealistic end results? It seemed to me several weeks ago (based on feedback on reddit from the open beta) that there's just too much complexity here to keep trying to go down the simulation rabbit hole. 

Take ground friction as an example. It's always been clear that ground friction is a bit... off... in KSP. Parts (and kerbals!) can slide down hills for miles, rovers slip sideways along inclines, etc.

Does the game try to use realistic calculations, coefficients of static and dynamic friction, and so on? It might seem appealing to do so, because then we can all be glad that KSP is using "realistic physics". But wait... it is also modelling the stickiness of hot rubber? Is it accounting for the fact that hard metal objects will dig into soft terrain? What about the fact that in the real world the ground isn't usually composed of a mesh of perfectly flat triangles, but rather has lots of little bits and bumps sticking out that stuff would get snagged on? And of course stuff also has similar bits and bumps. Or how about the way soft tires in particular will deform to "grab" the slight imperfections of a surface? If it's not modelling all of those things, and is only using coefficients of friction, then the result will be unrealistic. But modelling all of those things would be incredibly difficult, and would kill performance. An "approximulation", or "simpulation" (yes I just made these words up) seems like a far more viable approach than an actual simulation.
 

Edited April 29, 2016 by allmhuran

[numerobis]

6 minutes ago, allmhuran said:

How about developing simplified approximations for friction, wheels, suspension, and so on in order to get a realistic end result rather than attempting to develop realistic simulations that give unrealistic end results? It seemed to me several weeks ago (based on feedback on reddit from the open beta) that there's just too much complexity here to keep trying to go down the simulation rabbit hole. 

Unity wheel physics is a certain level of complexity. Setting the parameters appropriately is something of an art.

[allmhuran]

1 hour ago, numerobis said:

Unity wheel physics is a certain level of complexity. Setting the parameters appropriately is something of an art.

IIRC Squad are using some third party wheel plugin (I've seen " Edy's Vehicle Physics" referenced), not unity wheels as such. But even that seems to be wrong, at least when it comes to interaction with Kerbal terrain.

It's probably the combination of both the wheels and the terrain itself. *Lots* of things behave very strangely when interacting with the ground in KSP, not just wheels.

Edited April 29, 2016 by allmhuran

[regex]

13 minutes ago, allmhuran said:

IIRC Squad are using some third party wheel plugin (I've seen " Edy's Vehicle Physics" referenced), not unity wheels as such. But even that seems to be wrong, at least when it comes to interaction with Kerbal terrain.

The issue is likely that most games in Unity are not based on building things out of multiple parts with multiple joint interaction.  Vehicle Physics Pro is likely built on the assumption that a car is a single "part" rather than a cobbled-together bunch of rigid bodies like KSP uses.  This means that, while the simulation may be accurate, in KSP it is subject to a lot more forces than it would normally expect resulting in strange things.

[Red Iron Crown]

GOG is reporting that Squad asked them to hold off on rolling out 1.1.1, looks like a 1.1.2 patch might be imminent.

[RandomUser]

Yeah, the main issue with the new update for me, and 1.1 in general is how crap all the landing gear have become, plane landing gears and landing legs.  I can't land at the Mun unless the surface is nearly flat, because otherwise my ship will start sliding down, slowly but fast enough to cause problems with a Mun base I'm building.  Would like to have them stick to the ground better. 

[Renegrade]

1 hour ago, regex said:

The issue is likely that most games in Unity are not based on building things out of multiple parts with multiple joint interaction.  Vehicle Physics Pro is likely built on the assumption that a car is a single "part" rather than a cobbled-together bunch of rigid bodies like KSP uses.  This means that, while the simulation may be accurate, in KSP it is subject to a lot more forces than it would normally expect resulting in strange things.

That makes a lot of sense to me.

That being said though, the order of operations is critical in physics simulations, and it's entirely possible that something is being called in the wrong order or some silliness like that.  Or both concepts acting in concert.

That reminds me, I should actually FIND the walls on Pol and bug-report the crap out of 'em.

[Alshain]

1 hour ago, regex said:

The issue is likely that most games in Unity are not based on building things out of multiple parts with multiple joint interaction.  Vehicle Physics Pro is likely built on the assumption that a car is a single "part" rather than a cobbled-together bunch of rigid bodies like KSP uses.  This means that, while the simulation may be accurate, in KSP it is subject to a lot more forces than it would normally expect resulting in strange things.

With KSP's success that is sure to change very quickly.  Look how many Minecraft/Terraria clones there are now.

It's in Unity's best interest to fix it.

[Superfluous J]

1 minute ago, Alshain said:

With KSP's success that is sure to change very quickly.  Look how many Minecraft/Terraria clones there are now.

It's in Unity's best interest to fix it.

I think it's a testament to how FREAKING HARD it is to make a game like KSP that the only thing I've seen that even comes close (spoiler: It doesn't come close) is Simple Rockets on my phone, and that was 2d and didn't actually simulate physics very well.

[regex]

19 minutes ago, 5thHorseman said:

I think it's a testament to how FREAKING HARD it is to make a game like KSP that the only thing I've seen that even comes close (spoiler: It doesn't come close) is Simple Rockets on my phone, and that was 2d and didn't actually simulate physics very well.

^This.

22 minutes ago, Alshain said:

It's in Unity's best interest to fix it.

It's not entirely Unity's, VPP's, or KSP's thing to fix, it's everybody's.  KSP is literally "uncharted territory" for games.

Edited April 29, 2016 by regex

[Glaran K'erman]

Contrary to The Division, at least I feel the people at the heart of this game give a damn. Probably the reason were so far down this magnificently horrible wonderful rabbit hole.

No matter how frustrating it may be sometimes, I didn't take the red pill or the blue pill. I wanted the green pill, side effects and all.

[Gojira1000]

What @regex and @5thHorsemansaid - people talk about No Man's Sky as "pushing the boundaries" but it largely doesn't do anything you couldn't do (and in fact was done) on a C64 30 years ago - it just looks nicer. KSP, OTOH, is doing some serious math.

[numerobis]

2 hours ago, allmhuran said:

IIRC Squad are using some third party wheel plugin (I've seen " Edy's Vehicle Physics" referenced), not unity wheels as such. But even that seems to be wrong, at least when it comes to interaction with Kerbal terrain.

Ah, my bad. Still, wheel dynamics in a game engine is usually physics-based these days -- it's more robust that way than by using a pile of hacks.

Rigid body dynamics can suck sometimes if you allow your system to become overdetermined. Throw in numerical error and you can get all sorts of fun.

[NateDaBeast]

I just want to know if I'm going to need to spend an additional $60 to buy more RAM to stop KSP from crashing from Memory/RAM loss just so I can play the game.