Hamish Willee

@julianoes So now we are in violent agreement, this makes sense as documented right?

@julianoes What I think you are saying is that - when PX4 is put into "Pause" QGC is actually calling [MAV_CMD_DO_REPOSITION](https://mavlink.io/en/messages/common.html#MAV_CMD_DO_REPOSITION) (presumably specifying [MAV_DO_REPOSITION_FLAGS_CHANGE_MODE](https://mavlink.io/en/messages/common.html#MAV_DO_REPOSITION_FLAGS_CHANGE_MODE) flag to set into Hold mode?)...

Thanks @julianoes > This does work for "pause with autocontinue=false" but not for paused to a different mode which makes sense to me. How can that make sense???? The semantics...

> It does "just set the sequence". But when you're in an active mission stuck at autocontinue=false, then changing the sequence (or in this case it could be "manually setting...

@julianoes OK, I have removed the mention of how this might be paused/restarted. I'm still concerned about this, but perhaps we can agree consistent behaviour in https://github.com/mavlink/mavlink/pull/1870 and deprecate SET_MISSION_CURRENT....

@julianoes I have reverted the change to `SET_MISSION_CURRENT` and will move that to be part of the discussion with the command - they should all be agreed together. That leaves...

@The1only Can you please create a PR with the functions in https://github.com/ArduPilot/pymavlink

Thanks @hendjoshsr71 - I like INS too, because then you don't have to discuss the actual location. Hoping to discuss in the next mav call.

Thanks @korigod. I understand that local_frd and body_frd _can_ both be misinterpreted; the difference is that people _might_ misinterpret the local case, but they _often_ misinterpret the body case. I...

We discussed this in the dev call last night. We ended up looking at `MAV_FRAME_BODY_FRD`, which is defined as "FRD _tangent frame_ with origin centered on the vehicle" (so like...