Holger Steinhaus

Correct, my mistake. However, *args would probably trigger exactly the same problem.

> A keyword argument needs a keyword for naming, where would you expect them to come from From DBus introspection information? Currently thinking about implementig that way - would be...

1. I think, the message and it's encoding is not specific to a particular autopilot implementation. 2. The message is intended to be send periodically, therefore the timestamp is important...

Hi Leonard, > It is normal for motors to regularly hit maximum and minimum PWM outputs so this is not a good measure of the motors health Couldn't disagree more....

After discussing the pros and cons in detail, is there any reason not to add the message? I don't see that there is any cost other than reserving a single...

Yes, I am talking about the zero values on the log shown above. Currently, I am running tests with ArduCopter 3.6-dev, however the problem persists for more than a year...

I think there are at least three aspects of this problem: 1. why are there any params shown as zero, requiring a manually triggered re-read (maybe a race condition between...

> I cannot reproduce 1 or 2. Could it be somehow CRSF related? Well, CRSF is a rather slow link, so this could definitely affect the timing. However, a unrleiable...

ok, will have a look. Building master is not a problem.

Here is the log.txt (all other *.txt.? were empty): https://www.dropbox.com/sh/qvimyt5vdmapskr/AAC8evoZBB63H8XQow3CHel0a