condition_yaw doesn't work without a call to send_global_velocity

[atomictom]

I was testing out the condition_yaw function in the docs and while the command would send a MAVLink message, and the drone (SITL) would respond with success, the drone would not yaw (as indicated by the vehicle.attitude.yaw field). However, if I precede the condition_yaw call with a send_global_velocity(0, 0, 0) call (again, taken from the docs), the drone will rotate successfully.

I am executing these commands in GUIDED mode. As far as I can tell, there doesn't seem to be a mention of this requirement in the docs (I guessed at it since the yaw examples included send_global_velocity).

[hamishwillee]

@atomictom Can you create a gist of your example code please - ideally based off http://python.dronekit.io/examples/guided-set-speed-yaw-demo.html ?

[atomictom]

https://gist.github.com/atomictom/9bae5cb7642045b184da

It uses mostly unmodified versions of the condition_yaw and send_global_velocity functions. If you uncomment either of the lines with # send_global_velocity(0, 0, 0) then it yaws, otherwise it won't.

I load it up using mavproxy.py --master=172.28.128.1:5760 --cmd="module load droneapi.module.api ; api start gist.py" connecting to a vagrant version of SITL. After you see the confirmation Got MAVLink msg: COMMAND_ACK {command : 115, result : 0} the yaw should (when working correctly) change significantly.

[hamishwillee]

The gist didn't work (missing yaw() ) but I modified it and yes, I think you are correct. You can of coruse just move to your current position, but it is a bit annoying. This will be autopilot behaviour, not dronekit as such, but yes we should document it.

@rmackay9 - is this a known issue with Copter, and can we do anything about it?

[atomictom]

Ah, sorry about that. I renamed yaw() to condition_yaw() just before posting it to be consistent with the name in the docs. Here is an updated version for anyone who is interested.

[atomictom]

Also, I'm sorry, I did not mean to close it...

[eliao]

I believe @asetapen encountered this yesterday and has some additional details. Apparently this only happens after using a guided takeoff.

[hamishwillee]

This will be resolved by #307

[asetapen]

Woohoo! Yes I've seen this, thank you @hamishwillee for clarifying.

[hamishwillee]

@asetapen NO worries. I'd hoped @rmackay9 would consider it a bug, but 'twas not to be.