Kraken, will there ever be a fix?

[Jimbodiah]

I had high hopes that 1.x would have solved the Kraken issues, but I am still getting them on large structures (stations).

Will there ever be a fix for this?

[sal_vager]

Well, seeing as the Kraken refers to one specific issue that was solved a long time ago, without knowing what you're talking about I'll hazard a guess of "maybe".

[Jimbodiah]

Large ships that start to oscillate by themselves until they shake so violently that they blow up.

[sal_vager]

Yeah that's not a Kraken, it's a real issue though and mass plays a part, struts will prevent it however (we so need placeable struts and fuel lines...)

[Jimbodiah]

Already using KJR as KSP sucks to the point of being unplayable without it. Mass isn't so much of an issue as I have had heavier ships without problem, it's when you start having long sections in stations that it becomes a nightmare sometimes.

I'm not going to put a ton of struts on parts because KSP will not fix this issue, and besides that; good luck in putting struts on parts when they need to be assembled in orbit.

Joint strength and this Kraken iteration have been problems since I started playing and have never been addressed. I only ever get evasive answers and people saying that it's a feature, not a bug. Kind of getting tired of playing like this, as you are basically limited to childish builds with small parts unless you want to run into issues with KSP.

1.0 with x65 support would fix it all, but I have more crashes with 1.1.2 than I ever did with 1.0.5, even BSODs that I have not had in years. The fun is beginning to leech out of this game.

Edited June 3, 2016 by Jimbodiah

[John FX]

It could be fixed with every joint having a small damper.

Then oscillations would tend to zero, like in real life.

Also if it was wanted to be more like real life the damping could increase heat slightly based on how much energy was dissipated by friction.

[HebaruSan]

 

On 5/31/2016 at 0:23 PM, Jimbodiah said:

I had high hopes that 1.x would have solved

I would recommend reading the dev notes if you're interested in what will or will not be changed in a given version. SQUAD is pretty open/transparent about what they're working on, so there's little need to guess. If you're hoping for change X while the announcements are all about progress on change Y, you're just setting yourself up for disappointment (and frankly I see comments along those lines more often than I'd expect). The focus of 1.0 was aerodynamics and heating, and 1.1 was Unity 5 and wheels, and SQUAD made no representations that there would be improvements in any glitches afflicting large stations.

I'd also make a note in support of @sal_vager's point about how ambiguous the term "Kraken" is on these forums. It's essentially a synonym for "bug" nowadays; the wiki lists no fewer than 26 separate issues that have gone by some variant of that name. If you want to be clear, you pretty much need to spell out exactly the problem you're having.

http://wiki.kerbalspaceprogram.com/wiki/Deep_Space_Kraken

Edited June 3, 2016 by HebaruSan

[Jimbodiah]

Hence my question, as KSP has said nothing about this issue so far.

[Clockwork13]

Most of the krakens are unfixable bugs caused by Unity rather than the game.

[Claw]

It's an issue with the physics engine (which is embedded within Unity). Sometimes there are certain mass and build configurations that cause the physics to enter a feedback state. The migration to Unity 5 was an unknown from the standpoint of how much of a change there would be for this particular issue. I've had plenty of large builds that didn't require crazy amounts of struts to keep from entering an oscillatory state. But you can also cause it to happen by putting a couple of symmetric girders on a FL-T800, with fuel tanks on the end, and watch it walk off the launch pad (that's not a very large ship).

Without sounding more deflective, this is one of the issues with having an open ended physics game such as this. There are configurations that aren't handled well with the default physics and joints. Modifying one area of the game will certainly affect other areas. I'm not saying that it's being ignored, just that I've got no way to provide a "by date X, this issue will be solved!"

In the mean time, placing struts along the axis to create triangle between the axial body and the radial components generally takes care of the issue. Placing them in a web radially is usually less effective and requires many more struts to make stable.

[Jimbodiah]

Thanks for an official answer-ish.

The main problem in that is that it would have to be launched as a whole, as there is no way to add struts while in orbit. My rockets, sometimes as high as the VAB are never an issue with oscillations (I use KJR as well), it's always stations I build that go crazy. I've tried HyperEditing a complete station into orbit without the docking ports between the sections, as I though maybe the ports might be the weak points, but this didn't really help to remove the problem; it seems a bit less prone to oscillate, but still does so at random times.

Spoiler