Hamish Willee

Thanks very much @Davidsastresas ! This looks as we discussed. I'll raise it in the mav call this evening and also try get some eyes on this. Note that in...

> Right, so then can we specify what APIs exactly are determined by this protocol? Is it MANUAL_CONTROL? Is it also RC_CHANNELS_OVERRIDE? Is it sending commands where to go? If...

> @hamishwillee, how should we proceed, are we happy with the current definition? Should I proceed with a PR to mavlink docs? And if so, should I start a new...

Thanks @Davidsastresas Looking good! Some tweaks added as suggestions (FYI for some reason your repo does not allow me to accept these myself). Can you update minimal.xml `MAV_RESULT` with `MAV_RESULT_PERMISSION_DENIED`....

> > * Unless our companion computer is using manual_control or rc_override ( which usually isn't a good idea for a companion computer ), the only way it will control...

> will happen with a Cube Red? A Cube Red has 2 AP's internally each with their own SYSID_THISMAV, a companion computer in a vehicle with a Cube Red needs...

> It does seem to make sense that the two independent (but co-operating) AP's in the Cube Red should have the same SYSID, but a different component Id, but this...

@Davidsastresas @timtuxworth I can see a way forward for this particular issue/protocol assuming: - the two FCs are independent MAVLink components - All **other** components in the system can infer...

If the autopilot loses it controlling GCS then it becomes uncontrolled (is the current discussion). @Davidsastresas We should add that too.

OK, hands off for now. Lets see what general thinking is.