numan

Not always. "K" predicts not akp. However "k" predicts akp. I suggest to turn all training data to lowercase, and then only accept lowercase input

"greyed out" generally caused by qmls enabled property which in this case controlled in by ```js QGCTabButton { text: qsTr("Fence") enabled: _geoFenceController.supported } ``` and geoFenceController.supported is at [geofencecontroller.cc](https://github.com/mavlink/qgroundcontrol/blob/43d412cbf11edd7d6577a5bd14bbd3965c7f948c/src/MissionManager/GeoFenceController.cc) ```cpp...

this issue is progresing in the referenced issue (https://github.com/mavlink/mavlink/pull/1771#issuecomment-1040852321) and can be considered duplicate i think

this also happens in docker

please add windows platform to labels

Hi @Jaeyoung-Lim I am interested in this issue. However I am a little confused. What is the purpose of the sanity check `1.0 / real_time_update_rate != max_step_size` ? I am...

in adition i noticed `real_time_update_rate_int` is ignored in merge request.Assuming we set `real_time_update_rate = speed_factor_/max_step_size`; what should it be? `1.0 / max_step_size *250` or `real_time_update_rate_int / 250`.

I have tried, simulation did run, and I also checked value via console messages; but "real time factor" in gazebo GUI did not change according to speed_factor_ so I am...

any news on this ?

thank you :D