ModuleRCSFX in 1.2

[NecroBones]

 

Has anyone played around with the new ModuleRCSFX in 1.2? I'm having some trouble with the transform orientations. The modules in the stock RCS units reference a new transform "RCSjet". It looks like these still use the +Y direction for the particle effects, like the old RCS transforms. But my custom RCS units are behaving strangely, whether I use +Y or +Z for the thrust transforms. Either way, I get a fairly random selection of thrusters firing, instead of the ones aimed in the direction I'm attempting to go. It looks like some of the "RCSthruster" vectors want +Y and others want +Z, or even -Z, without any clear indication as to which ones will want one or the other. Any thoughts on what I'm overlooking?

 

All I've been able to glean from the stock RCS thrusters is that the RCSjet and RCSthruster transforms are all parallel in the hierarchy, not nested with each other, so I figure it still has to be transform orientation alone that's figuring out which ones to fire.

 

[NecroBones]

 

However, if I make the EFFECTS and ModuleRCSFX both reference the same transforms, instead of duplicate/separate transforms, it looks like it may be working.

 

How are the stock RCS parts correlating those separate transforms, and why are they separating them out, if using the same transforms for both will work fine?

 

EDIT: Yes, with some additional testing, it looks like it works as expected to just reference the same transforms in both EFFECTS and ModuleRCSFX. No idea why the stock RCS block separates them.

 

[Skalou]

On 20/10/2016 at 4:43 PM, NecroBones said:

EDIT: Yes, with some additional testing, it looks like it works as expected to just reference the same transforms in both EFFECTS and ModuleRCSFX. No idea why the stock RCS block separates them.

i think it separates them because the trust is not inline with the effect axis: 90° for the physical thrust vs 105° for the effect, to avoid the front thruster firing if we ask the ship to go starboard for example, the old thrust transform was also too deep in the nozzle for the effect.

i used it for my ( small) mod:

https://kerbal.curseforge.com/projects/rcs-105-180/images

however the new recompiled stock rcs-105 model is also 4% bigger than the old one, mistake?

Edited October 21, 2016 by Skalou

[NecroBones]

27 minutes ago, Skalou said:

i think it separates them because the trust is not inline with the effect axis: 90° for the physical thrust vs 105° for the effect, to avoid the front thruster firing if we ask the ship to go starboard for example, the old thrust transform was also too deep in the nozzle for the effect.

 

That makes a lot of sense, actually. I'm guessing it's correlating the effect vector with the thrust vector based on the order they appear in the list of transforms or something, since my tests seemed so randomized. For my purposes, it's just best to use the same transforms for both anyway.

 

[Skalou]

it's a mystery for too how it works, why and how the effect didn't thrust when the angle should allow it? how it link the trusst to effect? but it worked for me, just put the Y axis ( both for thrustt ransform and effect transform) in the good orientations and go! and yes, matching the config names is a good way to avoid mistake.