Benedikt Hallinger

@jmturner Hi, did you found something, or is this mergeable now?

@jmturner I Think this should be ready.

Hi James, thanks. Would be cool if someone could look over this. Also no idea really, how to add unit tests; but this is a purely cosmetic change as by now.

I fixed that in my branch, and it looks good now, I think. The filter is tuned (can be changed at magnetic-compass config and runtime by setting the <fluid-viscosity type="double">8.2</fluid-viscosity> property. it is accepted at runtime at the instruments node, and at config time in instrumentation.xml. it defaults to 8.2 which I guesstimated to the vicsosity and expected damping of kerosine. As I'm no real pilot, we need to work with that, or someone can provide better numbers. But at least it...

And also the disc bounces around a little too much in my opinion. the idea of the filter is to simulate a damping fluid (lowpass is maybe a bad choice for that. Prior that I used exponential filters in XML space and that did the trick. Don't know how to code a exponential filter however. I think I'll ask in the list.

Added another commit to simulate damping due to fluid. However, the result of this PR still looks not good. Probably someone more capable should look over the math and logic. For example, with the current filter setting the disc has roll when the plane clearly is level when standing still on the ground (tested at EGGK rwy 08R)

Tried to reproduce. I got no crash, but also no turbulence, despite the box beeing ticket.

Added something to the wiki, is this correct now?