WickedShell

6. It depends. The sequence number is the intended way to tell about it, but if you use batching (allows significantly faster transfers) then sequence failures become much more probable....

What's the value in sticking it in the existing storage destination for mission commands? It can easily be done of course (and a no-op command is easy to handle), but...

@DonLakeFlyer that would be my preference. It also seems to map out better as you can simply ask a vehicle for what the ID is without having to know where...

Depends what way you want to update it. Changing it to be millis instead of micros doesn't break parsers, it can cause some issues, but it doesn't lead to packets...

I'd really like to move to just one GPS_RAW_INT message with a field used to indicate which instance it is. (Basically this allows the simplification and reduction of code as...

@LorenzMeier Yeah, that was expected when I opened the initial issue. The issue was opened to track it, particularly with an aim of starting a discussion for mavlink2.

> 1. I presume you are still interested in this? Somewhat yes. I definitely prefer to see instance based messages rather then have the 2 3 4 variants, as well...

> I think you're saying that if the message is sent by a MAVLink 1 system the field will be received as 0 by the receiving mavlink 2 system? So...

I'm not a major copter person, but I'd argue that the stick check is still valid. The stick check is designed to catch the case when you have nudged trims...

If you are operating without any RC system then why not just disable the check via the RC_OPTIONS parameter? If you are going to use RC at any point then...