WickedShell

> I think we'll need a low pass filter, maybe 0.2Hz cutoff? I'm pretty sure the sensor I'm looking at is internally filtered, but obviously not all sensors are, and...

I'm handling the battery one already. (There is an expected follow up PR to the two stage failsafe one that I'm sending in shortly that consolidates the battery information elsewhere).

The motivator appears to be some oddities with using literals in scripts, aggravated by rejecting negative integer values. Given the need to have useful literals it might be better to...

> we could check if the constructor has 2 arguments we can do: local foo = uint32_t(0xF000, 0x0007) that would create 0xF0000007 Why not accept 0xFFFFFFFF directly? The compiler will...

> I prefer the new aux functions, the param is a `int16` so were not going to run out. There maybe some transport cases where were using `uint8`, but were...

> I could however do Current1 and Battery2 (User Items) to display Current and Voltage in the HUD of Mission planner. But that does not work on Qgrouncontrol Herelink if...

Updated, and SITL tested. MAVLink commit is an easy thing to resolve the conflict on, but I'm actually happy with that being a blocker until the change is merged into...

Updated the MAVLink submodule now that the MAVLink repo has the message, and fixed the conflicts.

> @WickedShell failing lots of tests, looks like mavlink not updated correctly? Sorry, bad rebasing, I've fixed it.

@priseborough Can you test a slowly failing sensor? I've seen multiple MS4525DO based airspeed sensors fail slowly inflight where it drifts off on the scale of 150 pascals a minute...