Rotational Data Held Through Timewarp

[Voyager55]

Probably not everyone would welcome this feature, so it could be optional. But couldn't it be possible for a crafts "rotational data" to be stored while it's in time warp or not in use/out of focus? And once you stop time warping, or take control of a craft from the KSC it's stored rotational data could be reasserted? is this even possible? I am by no means a programmer...

Edited January 14, 2015 by Voyager55

[Myslius]

Thumb up for this one. Personally, i would love to see my station spinning forever. http://forum.kerbalspaceprogram.com/threads/106311-Artificial-Gravity-Station

[Voyager55]

Thanks, and beautiful station (Artificial gravity stations were actually one of my main inspirations for this topic)

[Myslius]

I think it all comes down to performance vs. reality ratio. KSP takes a lot of shortcuts for calculations. But, if there is any possibility to add it (even with some performance drop) i would like to see it. Maybe enabling it thru Alt+F12 window. Or maybe it's not that performance consuming feature?

[vexx32]

Well, if you're not simulating physics, you could just add simple rotation to the vessel without calculating forces on it. I don't think it ought to be the difficult to store xyz axes rotation for the whole craft, nor should it be overly difficult to have xyz 'current rotation' vectors stored with the craft and used for timewarp.

The real issue here is that rotation is modelled improperly. It's deliberately dampened, so that any rotation will eventually stop, even in space where there ought to be nothing to stop it. If you rotate a craft without SAS active, you can see this happening. The rotation gradually slows, for no reason whatsoever. Modelling that in timewarp is the real difficulty. If that were removed, as it probably ought to be, it would become quite simple, I think.

[Myslius]

I think vexx32 is right. It's quite a simple calculation. Hope to see it fixed. It's quite a noticeable bug in KSP physics.

[BudgetHedgehog]

Not until SAS actually makes the craft not move at all. At the moment, timewarp is the only way to completely kill any rotation, which should the job of SAS/RCS.

[Renegrade]

The real issue here is that rotation is modelled improperly. It's deliberately dampened, so that any rotation will eventually stop, even in space where there ought to be nothing to stop it. If you rotate a craft without SAS active, you can see this happening. The rotation gradually slows, for no reason whatsoever. Modelling that in timewarp is the real difficulty. If that were removed, as it probably ought to be, it would become quite simple, I think.

Oh, so that extends down to zero? I noticed that if I set up a violent spin on a capsule in space, it eventually slows to a .. less violent spin..but I assumed that was just some anti-Kraken magic that didn't go all the way to zero....

[vexx32]

It does eventually drop to zero, aye (as far as I know, anyway; it's been quite a while since I tested it). As far as I know, it's just something they added to make things slightly more manageable for players -- having your craft spinning can be a nightmare if you're not sure how to cancel it properly. But then, SAS should be sufficient for such a purpose in any case. So it really doesn't have much purpose in the current game, and its presence is arguably detrimental in some ways.

[sarbian]

uh ? Are you sure about that ? The only limit I have seen related to rotation is the angular velocity limit that's part of Unity.

[Tery215]

I can agree with vexx - I launched a craft with lots of rotational SRB's, it stopped rotating after a while.

Rotational decay might be low-cost to implement along with rotational data during timewarp, I'm not sure.

[vexx32]

Regarding what sarbian said -- I never mentioned a limit to rotation at all. What's happening (as I understand it, having asked people who have looked at the code around it) is that the game actually dampens rotation forces by a certain amount.

I mean, calculating rotational damping probably wouldn't be too tricky over timewarp, but I feel as though it's such an unnecessary thing to have that it should probably go out the window entirely. As a bonus, it would make applying rotation properly over timewarp simpler. What I see as the main problem with this, however, is that you've got basically two solutions:

1. Calculate total rotation for the entire time interval upon dropping out of warp. You won't see the rotation happening once in warp, but you'll see your craft rotated upon dropping out of warp. The issue with this is that it could end up being confusing, as the craft would be abruptly rotated, potentially seriously messing up a planned manoeuvre or the like.
   
2. Calculate rotation on the vessel each step while in timewarp. It would be a simple rotation of the entire craft, no need to simulate forces. The issue with this is that at high warp, even tiny amounts of rotation would be amplified, turning your vessel into a spinning blur. That would make it essentially impossible to interact with any part of the craft and/or check its orientation whilst in high warp. Not a massive loss, but still something to consider, I think.